Pleural Diseases (Libro) PDF
Pleural Diseases (Libro) PDF
Pleural Diseases (Libro) PDF
SIXTH EDITION
ERRNVPHGLFRVRUJ
PLEURAL DISEASES
SIXTH EDITION
Richard W. Light, MD
Professor of Medicine
Vanderbilt University
Division of Allergy, Pulmonary and Critical Care
Nashville, Tennessee
UnitedVRG
ERRNVPHGLFRVRUJ
• .
Health
I
Wolters Kluwer Lippincott Williams & Wilkins
Philadelphia• Baltimore• New York• London
Buenos Aires· Hong Kong· Sydney· Tokyo
Senior Acquisitions Editor: Sonya Seigafuse
Senior Product Manager: Kristina Oberle
Developmental Editor: Jeff Gunning
Vendor Manager: Alicia Jackson
Senior Manufacturing Manager: Benjamin Rivera
Senior Marketing Manager: Kimberly Schonberger
Design Coordinator: Joan Wendt
Production Service: S4Carlisle Publishing Services
Fifth Edition © 2007 by Lippincott Williams & Wilkins, a Wolters Kluwer business
Fourth Edition © 2001 by Lippincott Williams & Wilkins
Third Edition © 1995 by Williams & Wilkins
Second Edition © by 1990 Lea & Febiger
First Edition © by 1983 Lea & Febiger
All rights reserved. This book is protected by copyright. No part of this book may be reproduced in any form by any means,
including photocopying, or utilized by any information storage and retrieval system without written permission from the copy
right owner, except for brief quotations embodied in critical articles and reviews. Materials appearing in this book prepared by
individuals as part of their official duties as U.S. government employees are not covered by the above-mentioned copyright.
Printed in China
616.2'5--dc23
2012043314
Care has been taken to confirm the accuracy of the information presented and to describe generally accepted practices. How
ever, the authors, editors, and publisher are not responsible for errors or omissions or for any consequences from application
of the information in this book and make no warranty, expressed or implied, with respect to the currency, completeness, or
accuracy of the contents of the publication. Application of the information in a particular situation remains the professional
responsibility of the practitioner.
The authors, editors, and publisher have exerted every effort to ensure that drug selection and dosage set forth in this
text are in accordance with current recommendations and practice at the time of publication. However, in view of ongoing
research, changes in government regulations, and the constant flow of information relating to drug therapy and drug reactions,
the reader is urged to check the package insert for each drug for any change in indications and dosage and for added warnings
and precautions. This is particularly important when the recommended agent is a new or infrequently employed drug.
Some drugs and medical devices presented in the publication have Food and Drug Administration (FDA) clearance for
limited use in restricted research settings. It is the responsibility of the health care provider to ascertain the FDA status of each
drug or device planned for use in their clinical practice.
To purchase additional copies of this book, call our customer service department at (800) 638-3030 or fax orders to (301) 223-
2320. International customers should call UnitedVRG (301) 223-2300.
Visit Lippincott Williams & Wilkins on the Internet: at LW W.com. Lippincott Williams & Wilkins customer service represen
tatives are available from 8:30 am to 6 pm, EST.
This book is dedicated to my wife and best friend, Judi Light.
The first five editions of Pleural Diseases were well received. Since the fifth edition was pub
lished in 2007, there has been a rapid advancement in the knowledge concerning pleural dis
eases. Accordingly, the publishers have requested that I prepare a sixth edition.
Some of the important advances in the knowledge about pleural disease that has become
available since 2007 include the following. It has become apparent that right heart failure
such as occurs with pulmonary hypertension at times leads to pleural effusions. The use of N
terminal-probrain natriuretic factor in the diagnosis of pleural effusions due to heart failure
has assumed a more prominent role. There have been many articles on the use of biomarkers
for the diagnosis of mesothelioma. The recommendations for the surgical treatment of meso
thelioma have been revised. A new definition for the post cardiac injury syndrome has been
proposed. It has been demonstrated that the administration of colchicine will diminish the
incidence of the post cardiac injury syndrome and the incidence of pleural effusion after coro
nary artery bypass surgery. The combination of DNase and tissue plasminogen activator has
been shown to significantly improve the rate of improvement of complicated parapneumonic
effusions compared to either agent alone or to placebo. The use of the indwelling catheter for
the management of malignant pleural effusions has become much more common. Several
articles have demonstrated that small chest tubes (9-12 F) are as effective as larger chest tubes
(>20 F) in most patients who require chest tubes.
However, if the small chest tubes are used for the treatment of parapneumonic effusions,
they should be irrigated every six hours with saline. The importance of adequate training and
the use of ultrasound when thoracenteses are performed has been demonstrated. The blood
patch technique has been shown to be an inexpensive, simple and effective technique for the
management of prolonged airleaks associated with pneumothorax or after thoracic surgery.
The pleural effusions that occur commonly when the anti-leukemic drug dasatinib is admin
istered are discussed. Image guided needle biopsy of the pleural (CT scan or ultrasound) is
more efficient than blind needle biopsy. Thoracentesis in patients on mechanical ventilation
improves oxygenation and decreases the time on the ventilator. Thoracoliths and pleuroparen
chymal fibroelastosis are described for the first time.
Details concerning all the above advances are included in this new edition. Overall, about
10-15% new references have been added.
It is my hope that the sixth edition of this book will continue to provide a practical, updated
reference book for physicians who take care of patients with pleural disease.
Richard W Light, MD
Nashville, Tennessee
vii
Approximately 1 million patients develop a pleural effusion each year. Pleural effusions may
occur with many different infections or as a complication of pulmonary disease. Additionally,
pleural effusions frequently complicate malignant disease, heart disease, liver disease, gastroin
testinal disease, kidney disease, and collagen vascular disease. Yet there are no recent books on
pleural disease to guide the practicing physician in determining the origin of a pleural effusion
or in managing a patient with pleural disease. Moreover, diseases of the pleura receive only
superficial treatment in books on pulmonary disease or internal medicine.
This book is intended primarily as a reference book for physicians who take care of patients
with pleural diseases. Recent advances in the knowledge of pleural disease make publication
of this volume timely. In this one volume, the practicing physician will have a comprehensive
discussion of all aspects of pleural disease.
The first three chapters discuss the anatomy, physiology, and radiology of the pleura. The
next chapter describes the clinical manifestations of pleural disease and discusses in depth the
various diagnostic tests that might be used to establish the etiology of a pleural effusion. In
Chapter 5, I present my recommended approach to the patient with an undiagnosed pleural
effusion. The following 13 chapters contain discussions of the various disease states that can
be associated with a pleural effusion. For each disease, the pathophysiology, clinical manifesta
tions, diagnosis, and management of pleural effusion are outlined. In Chapters 19 through 21,
pneumothorax, hemothorax, and chylothorax are presented, respectively. Pleural thickening
not associated with pleural fluid is covered in Chapter 22. The next two chapters are devoted
to those procedures used most often in managing patients with pleural disease, namely, diag
nostic and therapeutic thoracentesis, pleural biopsy, and tube thoracostomy. The final chapter
includes a description of the various drainage systems used with chest tubes.
It is my hope that publication of this book will result in better and more cost-effective
management of patients with pleural disease.
Richard W Light, MD
ix
There are several people that I would like to acknowledge who helped with the preparation of
this edition. The people that I would like to acknowledge at Wolters Kluwer Health - Lippincott
Williams & Wilkins include Sonya Seigafuse, Senior Acquisitions Editor, Kerry B. Barrett and
Kristina Oberle, Senior Product Managers, and Jeff Gunning, Developmental Editor. Lastly, I
would like to acknowledge Subrahmanyam Katakam of S4Carlisle Publishing Services, who did
a fantastic job preparing the page proofs.
xi
Preface, vii
Preface to the First Edition, ix
Acknowledgments, xi
xiii
Anatomy of the Pleura
The pleura is the serous membrane that covers the The cephalad portion becomes the pericardium and
lung parenchyma, the mediastinum, the diaphragm, communicates bilaterally with the pleural canals,
and the rib cage. This structure is divided into the which, in turn, communicate with the peritoneal ca
visceral pleura and the parietal pleura. The visceral nals. With development, the coelomic cavity becomes
pleura covers the lung parenchyma, not only at its divided into the pericardium, the pleural cavities,
points of contact with the chest wall, diaphragm, and and the peritoneal cavity through the development of
mediastinum but also in the interlobar fissures. The three sets of partitions: (a) the septum transversum,
parietal pleura lines the inside of the thoracic cavities. which serves as an early, partial diaphragm; (b) the
In accordance with the intrathoracic surfaces that it pleuropericardial membranes, which divide the
lines, it is subdivided into the costal, mediastinal, and pericardia! and pleural cavities; and (c) the pleuro
diaphragmatic parietal pleura. The visceral and the peritoneal membranes, which unite with the septum
parietal pleura meet at the lung root. At the pulmo transversum to complete the partition between each
nary hilus, the mediastinal pleura is swept laterally pleural cavity and the peritoneal cavity. This newly
onto the root of the lung. Posterior to the lung root, formed pleural cavity is fully lined by a mesothelial
the pleura is carried downward as a thin double fold membrane, the pleura (1).
called the pulmonary ligament. When the primordial bronchial buds first ap
A film of fluid (pleural fluid) is normally present pear, they and the trachea lie in a median mass of
between the parietal and the visceral pleura. This thin mesenchyme, cranial and dorsal to the peritoneal cav
layer of fluid acts as a lubricant and allows the visceral ity. This mass of mesenchymal tissue is the future me
pleura covering the lung to slide along the parietal diastinum, and it separates the two pleural cavities. In
pleura lining the thoracic cavity during respiratory humans, no communication normally exists between
movements. The space, or potential space, between the two pleural cavities. As the growing primordial
the two layers of pleura is designated as the pleural lung buds bulge into the right and left pleural cavi
space. The mediastinum completely separates the ties, they carry with them a covering of the lining
right pleural space from the left in humans. As previ mesothelium, which becomes the visceral pleura. As
ously mentioned, only a thin layer of fluid is normally the separate lobes evolve, they retain their mesothelial
present in this space, so it is a potential space rather covering. This covering becomes the visceral pleura
than an actual one. Many diseases are associated with in the fissures. The lining mesothelium of the pleural
increased amounts of pleural fluid, however, and cavity becomes the parietal pleura (2).
a large segment of this book is directed toward an
understanding of these diseases.
H I ST O L O G Y O F T H E P L E U RA
lymphatics shaped like flat cisterns and are located and the distance from the microvessels to the pleural
over the intercostal spaces, at least in sheep (3) . space ranges from 1 8 to 56 µm (as compared with
The mean thickness of the parietal pleura in sheep 1 0 to 1 2 µm for the parietal space) (3) .
is 20 to 25 µm, whereas the distance from the mi The connective tissue layer in the visceral pleura
crovessels to the pleural space is 1 0 to 12 µm. Deeper has two important functions: (a) it contributes to the
to the parietal pleura is the endothoracic fascia. This elastic recoil of the lung, which is important in expel
continuous band of dense irregular connective tis ling air from the lung, and (b) it restricts the volume
sue, composed mainly of collagen and elastin, covers to which the lung can be inflated, thereby protecting
the ribs and intercostal spaces and varies in thickness it (5) . In the visceral pleura, fibers of the elastic and
from 75 to 1 50 µm (3) . collagenous systems are clearly interdependent ele
The anatomy of the visceral pleura differs mark ments. Collagenous fibers are interwoven in a pleated
edly from that of the parietal pleura and also varies structure that closely resembles the osiers of a wicker
among species, primarily in its thickness. Dogs, cats, basket, suggesting that collagen fibers allow the lung
and monkeys have a thin visceral pleura, whereas volume to increase up to a point of maximal stretch
humans, sheep, cows, pigs, and horses have a thick ing of the system (5). The pleural contribution to the
visceral pleura (4) . The distinction between lungs elastic recoil pressure of the lung originates from the
with a thick or thin visceral pleura is important physi elastic network, which returns to its resting position
ologically because the blood supply is dependent on when inspiratory pressures are negligible (5) .
the thickness of the pleura. In animals with a thick Both the visceral and the parietal pleura are lined
visceral pleura, the predominant source of blood is with a single layer of flat mesothelial cells. These
the systemic circulation; in those with a thin pleura, mesothelial cells range in size from 6 to 12 µm in
the predominant source of blood is the pulmonary diameter (6) . With scanning electron microscopy (7) ,
circulation (4) . the pleural surface is found to be either flattened or
Histologically, a thick visceral pleura is composed bumpy (Fig. 1 . 1 ) . The bumpy areas include most of
of two layers: the mesothelium and connective tis the visceral pleura and portions of the parietal pleura,
sue. Blood, lymph vessels and nerves are located in including the subcostal regions and the pleural
the connective tissue. Animals with a thick visceral recesses. These areas appear to result from a lack of
pleura have a layer of dense connective tissue of rigidity of the underlying structures (6) .
varying thickness interposed between the mesothe Scanning electron microscopy also demonstrates
lium and the blood vessels (4) . In sheep, the visceral that microvilli are present diffusely over the entire
pleura ranges in thickness from 25 to 83 µm (as pleural surface (Fig. 1 . 1 ) , but the distribution of the
compared with 1 0 to 25 µm for the parietal pleura) microvilli is irregular. The density of the microvilli
A B
ranges from a few to more than 600/ 1 00 µm 2 , with sensmve and responsive to various stimuli. The
a mean of approximately 300 ( 1 ) . The microvilli are mesothelial cells that line the pleural cavity and those
most numerous on the inferior parts of the visceral that line the other body cavities have no recogniz
pleura and the anterior and inferior mediastinum on able cytologic difference ( 1 3) . The cytoplasm always
the parietal pleura ( 1 ) . At corresponding regions in contains a moderate to abundant amount of organ
the thoracic cavity, more microvilli are present on the elles, including mitochondria, rough and smooth
visceral pleura than on the parietal pleura. The micro endoplasmic reticulum, polyribosomes, intermediate
villi are approximately 0. 1 µm in diameter, and their fibrils, Golgi apparatus, and some glycogen granules,
length varies from 0 . 5 to 3.0 µm ( 1 ) . suggesting that the mesothelial cell is a metabolically
Th e exact function o f these numerous microvilli active cell ( 1 4) .
is yet to be defined. At one time, it was believed that Th e mesothelium i s now recognized as a dynamic
their presence increased the capacity of the visceral cellular membrane with many important functions.
pleura to absorb pleural fluid. This is probably incor These include transport and movement of fluid and
rect because recent observations have indicated that particulate matter across the pleural surfaces; leu
the visceral pleura plays a limited role in the absorp kocyte migration in response to inflammatory me
tion of pleural fluid. It is now thought that the most diators; synthesis of cytokines, growth factors, and
important function of the microvilli is to enmesh gly extracellular matrix proteins; release of factors to pro
coproteins that are rich in hyaluronic acid, especially mote both the deposition and clearance of fibrin; and
in the lower thorax, to lessen the friction between antigen presentation ( 1 5) . Mesothelial regeneration
the lung and the chest wall (7) . Moreover, as men involves migration of cells from the wound edge and
tioned earlier, a thin rim of fluid normally separates attachment and incorporation of free-Boating meso
the visceral and parietal pleura. Impingement of the thelial cells from the pleural fluid onto the denuded
microvilli from one pleural surface into the opposing pleural surface ( 1 6) . There is strong evidence that
pleural surface could possibly help maintain this thin mesothelial cells can convert to myofibroblasts. Yang
rim of fluid (8), but this is controversial (9) . et al. ( 1 7) assessed the effects of incubating peritoneal
The mesothelial layer is very fragile. At thoracot mesothelial cells with transforming growth factor
omy in patients without clinical pleural disease, fo beta (TGF-/3) and reported that the mesothelial cells
cal denudation of mesothelial cells is common ( 1 O). took on the characteristic myofibroblastic phenotype.
When the normal layer of mesothelial cells lining the We have observed that the incubation of human
pleura is disrupted, the defect is repaired through mesothelial cells with TGF-/3 results in their morpho
mitosis and migration of the mesothelial cells ( 1 1 ) . logic transformation to cells that look like fibroblasts.
When irritated, they retract but retain continu It has been shown that the intrapleural administra
ity with adjacent cells by projections called cellular tion ofTGF-/3 results in an excellent pleurodesis (18)
bridges. Mesothelial cells are frequently dislodged and the morphologic changes induced by TGF-/3
from the pleural surfaces and are thereby free in the referred to here may be important in producing the
pleural fluid. When free in the pleural space, the cells pleurodesis.
become round or oval ( 1 1 ) . Their cytoplasm is rich In cell culture, mesothelial cells have been shown
in organelles. From this state, they may be trans to produce type I, type II, and type IV collagens, elas
formed into macrophages capable of phagocytosis tin, fibronectin, and laminin, and to express inter
and erythrophagocytosis ( 1 1 ) . Such transformed cells mediate filaments typical of both epithelial cells and
frequently have vacuoles in their cytoplasm. Not all fibroblasts ( 1 9) . Mesothelial cells also express proco
the macrophages in pleural fluid evolve from meso agulant activity because of a tissue factor that binds
thelial cells; some definitely evolve from peripheral factor VII at the cell surface (20) . Mesothelial cells
blood mononuclear cells, and some may evolve from have also been demonstrated to produce nitric oxide
alveolar macrophages ( 1 2) . An immunologic role has (2 1 ) and TGF-/3 1 as well as many other cytokines (see
been suggested for the macrophages derived from the Chapter 5) ( 1 8) .
mesothelial cells ( 1 2) .
Mesothelial cells form a monolayer of specialized Major considerations in the understanding of pleural
pavement-like cells that line the pleural surfaces. fluid are volume, thickness, cellular components, and
The mesothelial cells are active cells, and they are physicochemical factors.
4 PLE U RAL D I S EASES
The blood supply to the visceral pleura is depen ribs. The lymphatic vessels of the mediastinal pleura
dent on whether the animal has a thick or thin pleura. pass to the tracheobronchial and mediastinal nodes,
In general, the blood supply to the visceral pleura in whereas the lymphatic vessels of the diaphragmatic
animals with a thin pleura originates from the pulmo pleura pass to the parasternal, middle phrenic, and
nary circulation, whereas the blood supply in animals posterior mediastinal nodes. When quantum dots
with a thick pleura originates from the systemic circu with a diameter of 1 5 µm are injected into the pleural
lation through the bronchial arteries. Albertine et al. space of pigs, they are first visualized in the superior
have demonstrated in sheep, an animal with a thick mediastinal nodes (27) .
pleura, that the bronchial artery supplies the visceral The lymphatic vessels in the parietal pleura are
pleura completely and exclusively (4) . Humans have in communication with the pleural space by means
a thick visceral pleura, which is probably why it is of stomas that range in diameter from 2 to 6 µm
also supplied by the bronchial artery, but there is still (Fig. 1 .2) (28,29) . When nitric oxide concentra
controversy (26) concerning this statement. All inves tions are increased, the stomas enlarge (30) . In one
tigators agree that the bronchial artery supplies most study, in rabbits the average density of the stomas was
of the visceral pleura facing the mediastinum, the 1 2 1 /mm3 (29) . These stomas have a round or slit
pleura covering the interlobular surfaces, and a part like shape and are found mostly on the mediastinal
of the diaphragmatic surface ( 1 4) . The blood supply pleura and on the intercostal surface, especially in the
for the remaining portions of the visceral pleura is less depressed areas just inferior to the ribs in the lower
understood and is thought by some to be through the thorax. There are more stomas in areas where the me
pulmonary artery ( 1 4) . The venous drainage of the sothelial cells are cuboidal rather than flat (29) . Few
visceral pleura is through the pulmonary veins. stomas are present in other portions of the parietal
pleura (3,28) . The distribution of stomas is similar
P L E U RA L LY M P H AT I C S
to the distribution of particulate matter injected into
the pleural space (Chapter 2) .
The lymphatic plexuses in the costal pleura are mainly The lymphatic vessels in the parietal pleura have
confined to the intercostal spaces and are absent or many branches. Some submesothelial branches have
minimal over the ribs ( 1 4) . The lymphatic vessels of dilated lymphatic spaces called lacunas (Fig. l .2B)
the costal pleura drain ventrally toward nodes along (28) . Stomas are found only over the lacunas. At the
the internal thoracic artery and dorsally toward the stoma, the mesothelial cells with their microvilli are in
internal intercostal lymph nodes near the heads of the continuity with the endothelial cells of the lymphatic
A B
FIGURE 1 .2 • Lym p hatics of the p a r i eta l p l e u ra. A: Sca n n i n g e l ectron m i c roscop i c st udy of the p a r i eta l
p l e u ra i n the ra b b it, d e m o n strat i n g a lym p h at i c sto m a . M i crovi l l i a n d m icro p i n ocyt i c o p e n i n g s on the
mesot he l i a l su rface a re both m u c h s m a l l e r t h a n the sto m a (o rig i n a l m a g n ification: 6,SOOX). B: To l u i d i n e b l u e
sta i n d e m o n strat i n g a red b l o o d c e l l at the sto m a o f a l a c u n a (ori g i n a l m a g n ification: 1,000X). (From Wang NS.
The preformed stomas connecting the pleural cavity and the lymphatics in the parietal pleura. Am Rev Respir Dis. 1 975; 1 1 1 : 12-20, with
permission.)
6 PLE U RAL D I S EASES
vessels. When red blood cells or carbon particles are other mononuclear cells around central lymphatic or
injected into the pleural space, they collect around vascular vessels. It has been suggested that the black
the stomas and in the lacunas and lymphatic vessels spots in patients with parietal anthracosis correspond
(Fig. l .2B) (3,28) . Therefore, these stomas with their to the Kampmeier foci and that the distribution of
associated lacunas and lymphatic vessels are thought asbestos fibers in the pleura is also concentrated in
to be the main pathway for the elimination of par these foci (34) . It has been hypothesized that the high
ticulate matter from the pleural space (3,28) . Occa concentrations of asbestos in these foci leads to the
sionally macrophages can be visible emerging from development of pleural plaques and mesothelioma
lymphatic stoma and entering the pleural cavity (29) . (34) . However, the occurrence of pleural plaques is
The existence of such stomas has been difficult to not related to the location of the black spots (35).
demonstrate in humans. Gaudio et al. (6) were unable
to demonstrate any such stomas in specimens from
30 patients undergoing thoracic surgical procedures. I N N E RVAT I O N O F T H E P L E U RA
Peng et al. ( 1 0) were able to demonstrate stomas in Sensory nerve endings are present in the costal and
only two of their nine human specimens. However, diaphragmatic parietal pleura. The intercostal nerves
subsequently Li (3 1 ) was able to demonstrate pleural supply the costal pleura and the peripheral part of
stoma in the diaphragmatic pleura in human speci the diaphragmatic pleura. When either of these areas
mens. The stoma were usually round or oval in shape is stimulated, pain is perceived in the adjacent chest
and approximately 6.2 µm in diameter. The stoma wall. In contrast, the central portion of the diaphragm
were not present in the visceral pleura or the parietal is innervated by the phrenic nerve, and stimulation
pleura on the chest wall. Most stoma were quite deep, of this pleura causes the pain to be perceived in the
forming channels that seemed to connect the pleural ipsilateral shoulder. The visceral pleura contains no
cavity with the underlying lymphatic lacunae. Inter pain fibers and may be manipulated without causing
estingly, in the golden hamster, there are many stoma unpleasant sensation. Therefore, the presence of pleu
but none in the diaphragmatic pleura (32) . ritic chest pain indicates inflammation or irritation of
The visceral pleura is abundantly endowed with the parietal pleura. However, the visceral pleura does
lymphatic vessels. These lymphatics form a plexus have sensory receptors closely related to elastic fibers
of intercommunicating vessels that run over the (36) . The functional role of these receptors remains
surface of the lung toward the hilum and also pen to be defined (36) .
etrate the lung to join the bronchial lymph vessels
by passing through the interlobular septa. Although
lymph may flow in either direction, all lymph from
the visceral pleura eventually reaches the lung root REFERENCES
either by penetrating the lung or by flowing on the 1 . Wang NS. Anatomy of the pleura. Clin Chest Med.
surface of the lung. The larger lymphatic vessels in 1 998; 1 9:229-240.
the visceral pleura are equipped with one-way valves 2 . Gray SW, Skandalakis JE. Development of the pleura. In:
directing flow toward the hilum of the lung ( 1 4) . No Chretien J, Bignon J, Hirsch A, eds. 1he Pleura in Health
and Disease. Lung Biology in Health and Disease, Vol. 30.
stomas are seen in the visceral pleura, and the lym
New York, NY: Marcel Dekker Inc.; 1 98 5 :3 1 9 .
phatic vessels of the visceral pleura are separated from 3 . Albertine KH, Wiener-Kronish J P, Staub N C . Th e structure
the mesothelial cells by a layer of connective tissue. of the parietal pleura and irs relationship to pleural liquid
The lack of stomas in the visceral pleura explains the dynamics in sheep. Anat Rec. 1 9 84;208 :40 1 -409.
observation that particulate matter injected in the 4. Albertine KH , Wiener-Kronish JP, Roos PJ, et al. Structure,
blood supply, and lymphatic vessels of the sheep visceral
pleural space is removed through the parietal pleura pleura. Am ] Anat. 1 982; 1 6 5 :277-294.
(see Chapter 2) . Fluid from the pleural space does not 5 . Lemos M, Pozo RM, Montes GS, et al. Organization of
enter the lymphatics in the visceral pleura in humans. collagen and elastic fibers studied in screech preparations
of whole mounts of human visceral pleura. Anat Anz.
1 997; 1 79:447-452.
6. Gaudio E, Rendina EA, Pannarale L, et al. Surface morphology
Kampmeier Foci of the human pleura: a scanning electron microscopic study.
Kampmeier (33) in 1 928 described small milky spots Chest. 1 9 88;92: 149- 1 53.
7. Wang NS. The regional difference of pleural mesochelial cells
in the dorsal and caudal portions of the mediastinum in rabbits. Am Rev Respir Dis. 1 974; 1 1 0:623-633.
in rats and humans. Microscopically, the foci are an ag 8 . Miserocchi G, Agostoni E. Pleural liquid and surface pressures
gregate of lymphocytes, histiocytes, plasma cells, and at various lung volumes. Respir Physiol. l 980;39:3 1 5-326.
CHAPT E R 1 I ANATO M Y O F T H E PLE U RA 7
9. Albertine KH, Wiener-Kronish JP, Bastacky ], et al. No 2 5 . Rolf LL, Travis DM. Pleural fl u id-plasma bicarbonate
evidence for mesothelial cell contact across the costal pleural gradients in oxygentoxic and normal rats. Am ] Physiol.
space of sheep. ] Appl Physiol. 1 9 9 1 ;70: 1 23- 14 3 . 1 973;224:857-86 1 .
1 0 . Peng M-J, Wang NS, Vargas FS, e t al. Subclinical surface 26. Bernaudin JF, Fleury J . Anatomy of the blood and lymphatic
alterations of human pleura. Chest. 1 994; 1 06:35 1 -353. circulation of the pleural serosa. In: Chretien J, Bignon J,
1 1 . Efrati P, Nir E. Morphological and cytochemical investigation Hirsch A, eds. The Pleura in Health and disease. Lung Biology
of human mesothelial cells from pleural and peritoneal effu in Health and Disease, Vol. 30. New York, NY: Marcel Dekker
sions. A light and electron microscopy study. Isr J Med Sci. Inc.; 1 98 5 : 1 0 1 - 1 24.
1 976; 1 2 : 662-673. 27. Parungo CP, Colson YL, Kim SW, et al. Sentinel lymph node
12. Bakalos D, Constantakis N, Tsicricas T. Distinction of mono mapping of the pleural space. Chest. 2005; 1 27: 1 799- 1 804.
nuclear macrophages from mesothelial cells in pleural and 28. Wang NS. The preformed stomas connecting the pleural
peritoneal effusions. Acta Cytol. 1 974; 1 8 :20-22. cavity and the lymphatics in the parietal pleura. Am Rev Respir
1 3 . Jones JS. The pleura in health and disease. Lung. Dis. 1 975 ; 1 1 1 : 1 2-20 .
200 1 ; 1 79:397-4 1 3 . 29. L i YY , L i JC. Ultrastructure and three-dimensional study
1 4 . Peng M-J, Wang N-S. Embryology and gross structure. In: of the lymphatic stomata in the costal pleura of the rabbir.
Light RW, Lee YC, eds. Textbook of Pleural Diseases. London, Microsc Res Tech. 2003;62:240-246.
England: Arnold Publishers; 2003:3- 1 6 . 30. Li YY, Li JC. Ultrastructural study of pleural lymphatic
1 5 . Mutsaers S E . Mesochelial cells: their structure, function and drainage unit and effect of nitric oxide on the drainage
role in serosal repair. Respirology. 2002;7: 1 7 1 - 1 9 1 . capacity of pleural lymphatic stomata in the rat. Ann Anat.
1 6 . Mutsaers SE. Th e mesothelial cell. Int ] Biochem Cell Biol. 2004; 1 86:25-3 1 .
2004;36:9-l 6. 3 1 . Li J. Ultrastructural study on the pleural stomata in humans.
1 7 . Yang AH , Chen JY, Lin JK. Myofibroblastic conversion of me Funct Dev Morphol. 1 993;3 :277-280.
sothelial cells. Kidney Int. 2003;63 : 1 530- 1 539. 32. Shinohara H . Distribution oflymphatic stomata on the pleural
18. Lee YC, Lane KB. Cytokines in pleural diseases. In: Light RW, surface of the thoracic cavity and the surface topography of
Lee YC, eds. Textbook of Pleural Diseases. London, England: the pleural mesothelium in the golden hamster. Anat Rec.
Arnold Publishers; 2003:63-89. 1 997;249 : 1 6-23.
19. Antony VB, Sahn SA, Mossman B, et al. Pleural cell biology in 33. Kampmeier OF. Concerning certain mesothelial thickenings
health and disease. Am Rev Respir Dis. 1 992; 1 4 5 : 1 236- 1 239. and vascular plexus of the mediastinal pleura associated with
20. Idell S, Zwieb C, Kumar A, et al. Pathways of fibrin turnover histiocyte and fat cell production in the human newborn.
of human pleural mesothelial cells in vitro. Am J Respir Cell Anat Rec. 1 928;39:20 1 -208.
Mo/ Biol. 1 992;7:4 1 4-426. 34. Boutin C, Dumortier P, Rey F, et al. Black spots concentrate
2 1 . Owens MW, Milligan SA, Grisham MB. Nitric oxide synthe oncogenic asbestos fibers in the parietal pleura. Thoracoscopic
sis by rat pleural mesothelial cells: induction by growth facrors and mineralogic study. Am J Respir Crit Care Med.
and lipopolysaccharide. Exp Lung Res. 1 995;2 1 :73 1 -742. 1 996; 1 5 3:444-449.
22. Noppen M, De Waele M, Li R, et al. Volume and cellular con 3 5 . Mitchev K, Dumortier P, De Vuyst P. 'Black Spots' and hyaline
tent of normal pleural fluid in humans examined by pleural pleural plaques on the parietal pleura of 1 5 0 urban necropsy
lavage. Am J Respir Crit Care Med. 2000; 1 62 : 1 023- 1 026. cases. Am J Surg Pathol. 2002;26: 1 1 98- 1 206.
23. Noppen M . Normal volume and cellular contents of pleural 36. Pintelon I, Brouns I, De Proost I, et al. Sensory recep
fluid. Curr Opin Pulm Med. 200 1 ;7 : 1 80- 1 82. tors in the visceral pleura: neurochemical coding and
24. Miserocchi G , Agostoni E. Contents of the pleural space. J live staining in whole mounts. Am J Respir Cell Mo/ Biol.
Appl Physiol. 1 97 1 ;30 :208-2 1 3 . 2007;36 : 54 1 -5 5 1 .
Physiology of the Pleural Space
The pleural space is the coupling system between Pleural Liquid Pressure versus Pleural
the lung and the chest wall, and, accordingly, it is a Surface Pressure
crucial feature of the breathing apparatus. The pres
There has been a controversy for many years as to
sure within the pleural space (the pleural pressure) is
whether there are two pleural pressures or one (2) .
important in cardiopulmonary physiology because it
The two different pressures had been proposed to
is the pressure at the outer surface of the lung and
explain a discrepancy obtained when the pleural
the heart and the inner surface of the thoracic cavity.
pressure was measured in two different ways. If the
Because the lung, the heart, and the thoracic cavity
pressure was measured using fluid-filled catheters,
are all distensible, and because the volume of a dis
the vertical gradient obtained was approximately 1 .0
tensible object depends on the pressure difference be
cm H 2 0/cm vertical height. This pressure was desig
tween the inside and the outside of the object and its
nated the pleural liquid pressure and was believed to
compliance, pleural pressure plays an important role
represent the pressure that influenced the absorption
in determining the volume of these three important
of fluid. If the pressure was measured using surface
structures.
balloons or suction cups, then a gradient of 0.3 cm
H 2 0/cm vertical height was obtained. This pressure
was designated the pleural surface pressure and rep
P L E U RA L P R E S S U R E resented the balance between the outward pull of
the thoracic cavity and the inward pull of the lung.
I f the thorax i s opened to atmospheric pressure, the It now appears that there is only one pressure, the
lungs decrease in volume because of their elastic re pleural surface pressure, and that the discrepancies in
coil, while at the same time, the thorax enlarges. With the pressures arose because of the distortion from the
the thorax open, the volume of the thoracic cavity catheters (3) . It should be noted, however, that there
is approximately 5 5 % of the vital capacity, whereas is still a school of researchers who believe in the pres
the volume of the lung is below its residual volume. ence of two different pressures (4,5).
With the chest closed and the patient relaxed, the re
spiratory system is at its functional residual capacity
Measurement
(FRC) , which is approximately 35% of the total lung
capacity ( 1 ) . Thus, at FRC, the opposing elastic forces Pleural pressure can be measured directly by inserting
of the chest wall and lung produce a negative pres needles, trocars, catheters, or balloons into the pleural
sure between the visceral and the parietal pleura. This space. Direct measurement of the pleural pressure is
pressure, the pleural pressure, surrounds the lung and not usually made because of the danger of producing
is the primary determinant of the volume of the lung. a pneumothorax or of introducing infection into the
The pleural pressure represents the balance between pleural space. Rather, the pleural pressure is measured
the outward pull of the thoracic cavity and the inward indirectly by a balloon positioned in the esophagus
pull of the lung ( 1 ) . (6,7) . Because the esophagus is a compliant structure
8
CHAPT E R 2 / PHYS I O LOGY OF TH E PLE U RAL SPACE 9
situated between the two pleural spaces, esophageal the pleural pressure gradient causes the alveoli in
pressure measurements provide a close approxima the superior parts of the lung to be larger than those
tion of the pleural pressure at the level of the balloon in the inferior parts. The higher pressure gradient
in the thorax (7,8). Estimation of pleural pressure by at the apex of the lung is thought to be responsible
means of an esophageal balloon is not without diffi for the formation of pleural blebs almost exclusively
culties (8) . The volume of air within the balloon must at the apex of the lung. The pleural pressure gradients
be small so that the balloon is not stretched and the also account for some unevenness in the distribution
esophageal walls are not displaced; otherwise, pleu of ventilation.
ral pressure estimates are falsely elevated. Moreover,
the balloon must be short and must be placed in the P L E U RA L F L U I D F O R M AT I O N
lower part of the esophagus. It has been demonstrated
that reliable measurements of esophageal pressures Fluid that enters the pleural space can originate i n the
can be made with micromanometers (9) . The use of pleural capillaries, the interstitial spaces of the lung,
the micromanometer should circumvent some of the the intrathoracic lymphatics, the intrathoracic blood
problems associated with esophageal balloons. vessels, or the peritoneal cavity.
35 29 35 16
6 0 6 13
- - - - - - - -
29 29 29 29
pressure is approximately - 5 cm H 2 0. The net hy Th e movement o f pleural fluid i s not the same
drostatic pressure is therefore 30 - ( - 5) = 35 cm across all the parietal pleura. Wang and Lai-Fook
H 2 0, and this favors the movement of fluid from the ( 1 3) used Evans blue-dyed albumin to study regional
capillaries in the parietal pleura to the pleural space. pleural filtration of prone anesthetized rabbits. They
Opposing this hydrostatic pressure gradient is the reported that there appeared to be more fluid forma
oncotic pressure gradient. The oncotic pressure in tion across the parietal pleura over the ribs compared
the plasma is approximately 34 cm H 2 0. Normally, with the intercostal spaces. In contrast, pleural liquid
the small amount of pleural fluid contains a small absorption was primarily in the parietal pleura adja
amount of protein and has an oncotic pressure of ap cent to the intercostal space rather than in the parietal
proximately 5 cm H 2 0 ( 1 1 ) , yielding a net oncotic pleura overlying the ribs. There was also more fluid
pressure gradient of34 - 5 = 29 cm H 2 0. Thus, the formation over the caudal ribs than over the cranial
net gradient is 35 - 29 = 6 cm H 2 0, favoring the ribs ( 1 3) . If the breathing frequency was increased,
movement of fluid from the capillaries in the parietal more fluid was formed ( 1 3) .
pleura to the pleural space. Th e transpleural exchange o f fluid i s species de
The net gradient for fluid movement across the pendent. Humans and sheep have a thick visceral
visceral pleura in humans is probably close to zero, pleura and its blood supply is from the bronchial
but this has not been demonstrated (Fig. 2 . 1 ) . The artery rather than from the pulmonary artery ( 1 4) .
pressure in the visceral pleural capillaries is approxi However, many species, such as the rabbit and the
mately 6 cm H 2 0 less than that in the parietal pleu dog, have a thin visceral pleura that receives its blood
ral capillaries because the visceral pleural capillaries supply from the pulmonary circulation. In such a sit
drain into the pulmonary veins. Because this is the uation, as shown in Figure 2.2, the net gradients favor
only pressure that differs from those affecting fluid pleural fluid formation across the parietal pleura and
movement across the parietal pleura and because pleural fluid absorption through the visceral pleura.
the net gradient for the parietal pleura is 6 cm H 2 0,
it follows that the net gradient for fluid movement
Interstitial Origin
across the visceral pleura is approximately zero. It is
also likely that the filtration coefficient (Lp) for the The origin of much of the fluid that enters the pleu
visceral pleura is substantially less than that for the ral space in disease states is the interstitial spaces of
parietal pleura because the capillaries in the visceral the lungs. Either high-pressure or high-permeability
pleura are much farther from the pleural space than pulmonary edema can lead to the accumulation of
those in the parietal pleura ( 1 2) . pleural fluid. When sheep are volume overloaded to
C H APTE R 2 I P H YS I O LOGY O F TH E P L E U RAL S PACE 11
produce high-pressure pulmonary edema, approxi across the visceral pleura appears to be weak, even
mately 25% of all the fluid that enters the interstitial though the visceral pleura is thick (26) . Therefore,
spaces of the lungs is cleared from the lung through once the subpleural interstitial pressure increases, it
the pleural space ( 1 5) . Within 2 hours of starting the follows that fluid will traverse the visceral pleura to
volume overloading, the amount of fluid entering the the pleural space.
pleural space increases, and within 3 hours, the pro
tein concentration in the pleural fluid is the same as
Peritoneal Cavity
that in the interstitial spaces of the lungs ( 1 5) . The
amount of pleural fluid formed is directly related to Pleural fluid accumulation can occur if there is free
the elevation in the wedge pressure. Increases in pleu fluid in the peritoneal cavity and if there are open
ral fluid accumulation occur only after the develop ings in the diaphragm. Under these conditions, the
ment of pulmonary edema ( 1 6) . fluid will flow from the peritoneal space to the pleural
Th e pulmonary interstitial space is the predomi space because the pressure in the pleural cavity is less
nant origin of pleural fluid in patients with congestive than the pressure in the peritoneal cavity. The perito
heart failure. The likelihood of a pleural effusion in neal cavity is the origin of the pleural fluid in hepatic
creases as the severity of pulmonary edema increases hydrothorax (Chapter 9), Meigs' syndrome (Chap
( 1 7) . In addition, the presence of pleural effusions is ter 20) , and peritoneal dialysis (Chapter 9) (27) .
more closely correlated with the pulmonary venous There are no direct lymphatic connections between
pressure than with the systemic venous pressure ( 1 7) . the peritoneal and pleural cavities (28) .
However, patients with right heart failure due to
pulmonary hypertension may have pleural effusions
Thoracic Duct or B lood Vessel Disruption
although their wedge pressures are normal ( 1 8) . The
origin of the pleural fluid in this situation is prob If the thoracic duct is disrupted, lymph will accu
ably the capillaries in the parietal pleura ( 1 8) . The mulate in the pleural space, producing a chylothorax
amount of fluid that enters the pleural space is also (see Chapter 26) . The rate of fluid accumulation with
increased when there is increased interstitial fluid chylothorax can be more than 1 ,000 mL/day. When
due to high-permeability pulmonary edema. When the thoracic duct is lacerated in dogs, sizeable pleural
increased-permeability edema was induced in sheep effusions begin to develop almost immediately (29) .
by the infusion of oleic acid, again, pleural fluid accu In a like manner, when a large blood vessel in the tho
mulated only after pulmonary edema developed ( 1 9) . rax is disrupted owing to trauma or disease, blood can
I n this study, there was n o morphologic evidence of accumulate rapidly in the pleural space, producing a
pleural injury. When pulmonary edema is induced by hemothorax (see Chapter 25) .
xylazine (20) or hyperoxia (2 1 ) in rats, or by ethchlor
vynol in sheep (22) , the high-protein pleural fluid
Origin of Normal Pleural Fluid
appears to originate in the interstitial spaces of the
lungs. The pleural fluid associated with experimen It is believed that the fluid that normally enters the
tal Pseudomonas pneumonia in rabbits originates in pleural space originates in the capillaries in the pari
the lung (23) . It is likely that the origin of the pleu etal pleura (30) . The normal pleural fluid production
ral fluid with many conditions associated with lung is approximately 0.0 1 ml/kg/hour in awake sheep
injury, such as pulmonary embolization and lung and 0.02 ml/kg/hour in rabbits (30) . If these rates are
transplantation, is also the interstitial spaces of the extrapolated to human beings, the amount of pleural
lung (2) . In experimental studies of hydrostatic and fluid formed daily in a 50-kg individual would be ap
increased-permeability edema, a pleural effusion de proximately 1 5 mL (30) . The origin of the fluid does
velops when the extravascular lung water has reached not appear to be the interstitial spaces of the lung be
a critical level in a certain amount of time (24) . The cause the protein level in the interstitial spaces is nor
necessary level of edema appears to be between 5 and mally approximately 4.5 g/dL, whereas the protein
8 g of fluid/gram of dry lung, depending on whether level in normal pleural fluid is only approximately
the edema is secondary to hydrostatic edema, oleic 1 to 1 . 5 g/dL. From Figure 2. 1 , it appears unlikely
acid lung injury, or O'.-naphthyl thiourea lung injury that the fluid originates from the visceral pleura.
(24) . With increasing levels of interstitial fluid, it has Likewise, both a lymphatic origin and a peritoneal
been shown that the subpleural interstitial pressure cavity origin appear unlikely. Supporting evidence for
increases (25) . The barrier to the movement of fluid this theory has been provided by Broaddus et al. (3 1 ) .
12 P L E U RAL D I S EASES
These workers measured the vascular pressures and Experimental work with sheep, a species with a
the pleural fluid protein levels in sheep of different thick visceral pleura similar to that of humans, sug
ages. They found that the systemic vascular pressures gests that most of the fluid that enters the pleural
progressively increased with age, whereas the pleural space in sheep is removed through the lymphatics.
fluid protein levels progressively decreased with age. Broaddus et al. (40) produced artificial hydrothoraces
These findings support a parietal pleural origin for in awake sheep by injecting an autologous protein so
normal pleural fluid because higher vascular pressures lution at a volume of 1 0 mL/kg, with a protein level of
should produce pleural fluid with lower protein levels 1 .0 g/dL. These investigators found that the hydrotho
(3 1 ) . Studies in rabbits with Evans blue-dyed albu rax was removed almost completely by the lymphatics
min have demonstrated that most fluid originates in in a linear manner at a rate of 0.28 ml/kg/hour. The
the parietal pleura over the ribs ( 1 3) . linearity suggests that the lymphatics operate at maxi
mum capacity once the volume of the pleural liquid
exceeds a certain threshold. Note that the capacity for
lymphatic clearance is 28 times as high as the normal
P L E U RA L F L U I D A B S O R PT I O N
rate of pleural fluid formation.
Lymphatic Clearance In the experiments of Broaddus et al. discussed in
the preceding text (40) , the fluid introduced into the
From Figure 2. 1 , one might have the impression
pleural space had an oncotic pressure of approximately
that pleural fluid should continuously accumulate
5 cm H 2 0, and from Figure 2. 1 , one might speculate
because Starling's equation favors fluid formation
that if fluids with oncotic pressures other than 5 cm
through the parietal pleura and there is no gradient
had been introduced, the equilibrium would have
for fluid absorption through the visceral pleura. Fluid
been altered such that fluid would enter the pleural
clearance through the pleural lymphatics is thought
space from the visceral pleura in animals with high
to explain the lack of fluid accumulation in normal
pleural fluid oncotic pressures and would leave the
individuals. The pleural space is in communication
pleural space through the visceral pleura in animals
with the lymphatic vessels in the parietal pleura by
with low oncotic pressures. This does not appear to
means of stomas in the parietal pleura. No such sto
be the case. Aiba et al. produced artificial pleural effu
mas are present in the visceral pleura. Proteins, cells,
sions in dogs with protein levels ranging from 0. 1 to
and all other particulate matter are removed from
9.0 g/dL (4 1 ) . Even when the induced pleural effusion
the pleural space by these lymphatics in the parietal
had a protein level of 0. 1 g/dL, there was no increase
pleura (32-35). When carbon particles are injected
in the concentration of protein with time, indicating
into the pleural space of anesthetized monkeys,
that the low oncotic pressure did not induce a rapid
thoracoscopy demonstrates that the carbon particles
effiux of fluid out of the pleural space. When the pro
go directly to the costal, mediastinal, and diaphrag
tein concentration of the induced effusions was above
matic pleura within 1 5 minutes of injection (36) . The
4 g/dL, the concentration of protein in the pleural
stomas through which the carbon particles exit the
fluid did gradually decrease with time, indicating a
pleural space are in areas where the mesothelial cells
net transfer of protein-free fluid into the pleural space.
are small and not flattened (36) . Increased levels of
However, the net amount of fluid entering the pleural
nitric oxide in the pleura will cause these stomas to
space even with a protein level of 9.0 g/dL was only
increase in diameter (37) .
0.22 ml/kg/hour. This degree of fluid flux is similar to
The amount of fluid that can be cleared through
the lymphatic clearance of 0.22 ml/kg/hour reported
these lymphatics is substantial. Stewart (38) found
in the same studies. These observations strongly sug
that the mean lymphatic flow from one pleural space
gest that most pleural fluid is removed through the
in seven patients was 0.40 ml/kg/hour, whereas
lymphatics in the parietal pleura in species with thick
Leckie and Tothill (39) found that the mean lym
visceral pleura, such as humans.
phatic flow was 0.22 ml/kg/hour in seven patients
with congestive heart failure. In both these studies,
marked variability was noted from one patient to
Clearance through Capillaries in Visceral
another. If these results in patients with congestive
Pleura
heart failure are extrapolated to the normal person,
a 60-kg individual should have a lymphatic drainage Until the mid- l 980s, it was thought that the primary
from each pleural space on the order of 20 mL/hr or route for the exit of fluid from the pleural space was
500 mL/day. through the capillaries in the visceral pleura (42) .
C H APTE R 2 I P H YS I O LOGY OF TH E P L E U RAL S PACE 13
This conclusion was based primarily on experiments Assuming that the 2,000 kDa dextran left the pleural
in animals with thin pleura. It is easily seen from Fig space only through stoma, they concluded that only
ure 2.2 that in animals with thin pleura, there is a 29% of the overall removal of albumin occurred
sizable gradient for the movement of fluid from the through the stoma with small hydrothoraces, while
pleural space into the capillaries in the visceral pleura. 64% of the albumin from large hydrothoraces was
In addition, fluid probably moves across a thin vis removed through the stoma (47) .
ceral pleura more easily than it does across a thick Shinto et al. (48) reported that when the volume
pleural membrane. However, on the basis of the ob of pleural fluid decreased with diuresis in patients
servations cited, it appears that in humans, almost with congestive heart failure, the concentration of the
all the pleural fluid is removed through the lymphat protein and LDH only increased slightly. They took
ics in the parietal pleura. Nevertheless, it should be this as evidence that all pleural fluid was removed by
noted that this view is not accepted by all (43). bulk flow through the lymphatics. However, Romero
The observations mentioned earlier should not et al. (49) reported quite different results in 1 5 pa
be interpreted as indicating that small molecules do tients who had their pleural fluid chemistries mea
not move across the pleural surfaces. Indeed, water sured before and at a mean of 1 1 5 hours after diuresis
and small-sized molecules exchange easily across both was started. They reported that the mean protein level
pleural surfaces (44) . When hydro tho races are induced increased from 2.3 g/dL to 3 . 5 g/dL while the LDH
in dogs, the clearance rate for para-aminohippurate increased from 1 76 IU/L to 262 IU/L (49) . Similar
(PAH) (molecular weight 2 1 6) is approximately percentage increases were seen in the albumin, cho
2 mL/kg/hour (4 1 ) . When urea is injected intrapleu lesterol, and cholinesterase concentrations. Their re
rally into patients with pleural effusions, its concen sults suggest that not all fluid is removed by bulk flow
tration decreases much more rapidly than does that of through the lymphatics.
radiolabeled protein (45). Indeed, the urea clearance If large molecules are removed through lymphat
rate is several hundred milliliters/hour (45). Because ics and smaller molecules are removed by a different
urea and water have comparable molecular weights, mechanism, then there should be a level at which
one can assume that the rates of exchange for urea larger molecules are all removed at one rate and be
and water across the pleural membranes are similar. low which molecules are removed at a different rate.
Therefore, several hundred milliliters of water prob However, Stashenko et al. (50) have shown that
ably traverse the pleural membranes each day, but when dextran molecules of varying sizes are placed
the net movement is of only a few milliliters because in the pleural space of rabbits, there was a continu
the osmolarity is nearly identical on each side of the ous spectrum in the rate of absorption of the dextran
membrane. molecules with the larger molecules being absorbed
more slowly (50) . This latter observation is consistent
Alternative Mechanisms for Pleural Fluid with multiple pore sizes or pores that allow particles
Removal through with a probability dependent on the size of
the particle (50) .
Although the assumption that all pleural fluid is
removed from the pleural space via bulk flow through
the lymphatics is attractive and has a lot of support
PAT H O G E N E S I S O F P L E U RA L
ing evidence, there are some questions about the
EFFUSIONS
validity of this theory. There is some evidence that
transcytosis contributes to the removal of protein Pleural fluid accumulates when the rate o f pleu
from the pleural space. Agostoni et al. (46) studied ral fluid formation exceeds the rate of pleural fluid
the removal of albumin and dextran from the pleural absorption. The main factors that lead to increased
space of anesthetized rabbits with and without the ad pleural fluid formation or decreased pleural fluid ab
ministration of nocodazole, a transcytosis inhibitor. sorption are tabulated in Table 2. 1 . Normally, a small
They reported that the removal of both the albumin amount (0.0 1 mL/kg/hour) of fluid constantly enters
and dextran was significantly greater in the control the pleural space from the capillaries in the parietal
group (46) . They concluded that 0.05 mL/hour of pleura. Almost all of this fluid is removed by the lym
liquid was removed by transcytosis (46) . These same phatics in the parietal pleura, which have a capacity
researchers subsequently conducted a study (47) in to remove at least 0.20 mL/kg/hour. Note that the
which they assessed the removal of labeled albumin capacity of the lymphatics to remove fluid exceeds the
and labeled dextran from the pleural space of rabbits. normal rate of fluid formation by a factor of 20.
14 P L E U RAL D I S EASES
not too important because when a pleural effusion pleural Buid formation ( 1 8) . Pleural effusions also
is induced in sheep with a protein level of 9.0 g/dL, develop in sheep when the pressure in the superior
the rate of Buid entry into the pleural space is only vena cava is increased. Allen et al. (55) found that
0 .22 mL/kg/hour (4 1 ) . This rate of Buid formation is pleural Buid accumulated over a 24-hour period
approximately equal to the capacity of the lymphatics when the pressure in the superior vena cava exceeded
to remove pleural Buid. Moreover, hypoproteinemia 1 5 mm Hg. The amount of pleural fluid that accu
is thought to be a very uncommon cause of pleural mulated increased exponentially as the pressure was
effusion (54) . increased. These workers reported that the larger the
pleural effusion, the higher the protein level. They
Presence of Free Peritoneal Fluid, or
concluded that the pleural effusions developed be
Disruption of the Thoracic Duct or an cause of (a) lymph leakage out of the lymphatics that
lntrathoracic Blood Vessel pass through the chest (these include the thoracic
duct and the diaphragmatic and pulmonary lymphat
If there is free Buid in the peritoneal cavity, it will lead ics) or (b) obstruction of lung or chest wall lymphat
to pleural fluid accumulation if there is a hole in the ics with subsequent leakage of interstitial fluid into
diaphragm (27) . In a similar manner, chyle will ac the pleural space (5 5 ) .
cumulate in the pleural space if there is a disruption
in the thoracic duct, and blood will accumulate in the Role of Aquaporins in Pleural Fluid
pleural space if there is a disruption of a blood vessel Exchange
in the thorax.
The aquaporins (AQPs) are a family of proteins that
transport water across membranes (56) . A deficiency
Decreased Pleural Fluid Absorption of an AQP in certain organs has produced significant
Obstruction of lymphatics abnormalities. For example, deletion of AQP 1 in
mice results in a severe defect in the ability to con
The most common cause of a decrease in pleural fluid centrate urine and the mice become profoundly de
absorption is obstruction of the lymphatics draining hydrated when deprived of water (56) .
the parietal pleura. Normally, the lymphatic flow There are at least four AQPs present in the lung
from the pleural space is approximately 0 . 0 1 mL/kg/ (57) . Transgenic mouse models with AQP deletion
hour or 1 5 mL/day because this is the amount of have provided information about their physiologic
pleural Buid formed. However, the capacity of the role. In the lung, AQP l and AQP5 provide the prin
lymphatics is approximately 0.20 mL/kg/hour or cipal route for osmotically driven water transport;
300 mL/day. Lymphatic blockade is an important however, neither alveolar Buid clearance in the neo
factor that contributes to the development of a ma natal and adult lungs nor Buid accumulation in ex
lignant pleural effusion. Leckie and Tothill (39) stud perimental models of lung injury is affected by AQP
ied the lymphatic Bow in eight patients with lung deletion (57) .
carcinoma and six patients with metastatic breast Immunostaining of the pleura has revealed the
carcinoma and found that the mean lymphatic Bow presence of AQP 1 in microvascular endothelia near
was only 0.08 mL/kg/hour. Obviously, pleural ef the visceral and parietal pleura and in mesothelial
fusions would not have developed in these patients cells in the visceral pleura (58) . In AQP l knockout
unless excess Buid had also been entering the pleural mice, osmotic equilibration of either hypertonic or
space. Unless the lymphatic Bow is markedly im hypotonic pleural fluid was slowed by a factor of four
paired, another factor must be present in addition to compared with wild-type mice (58) .
lymphatic disease to produce a pleural effusion given However, in a Buid overload model produced by
the excess reserve capacity of the lymphatics. intraperitoneal saline administration and renal ar
tery ligation, the accumulation of pleural Buid was
Elevation of Systemic Venous Pressures
not affected by AQP l deletion (58) . Moreover, in a
There is high incidence of pleural effusions in pa thiourea toxicity model of acute endothelial injury
tients with pulmonary hypertension ( 1 8) . Most of causing pleural effusions and lung interstitial edema,
the patients with pulmonary hypertension who have AQP l deletion did not affect pleural fluid accumu
pleural effusions also have right heart failure ( 1 8) . lation (58) . These results suggest that AQP l does
I t i s thought that pleural Buid accumulates because not play a role in clinically relevant mechanisms of
the elevated systemic venous pressure leads to more pleural Buid accumulation or clearance.
16 P L E U RAL D I S EASES
containing the plasmid for placental alkaline phos 8. Milic-Emili J, Mead J, Turner JM, et al. Improved technique
phatase are injected into the pleural space of rabbits, for estimating pleural pressure from esophageal balloons.
] Appl Physiol. 1 964; 1 9:207-2 1 1 .
the levels of placental alkaline phosphatase increase 9 . Hartford CG, Rogers GG, Turner MJ. Correctly selecting a
in both the pleural fluid and the serum (65) . Another liquid-filled nasogastric infant feeding catheter to measure
group (66) administered adenoviruses containing intraesophageal pressure. Pediatr Pu/mono!. 1 997;23 :362-369.
an antiangiogenesis vector expressing a soluble, se I 0. Parameswaran S, Brown LV, Ibbott GS, et al. Hydraulic con
ductivity, albumin reflection and diffusion coefficients of pig
creted, extracellular portion of the Flt- 1 receptor
mediastinal pleura. Microvasc Res. 1 999;58: 1 1 4- 1 27.
for VEGF intrapleurally in mice that had lung tu 1 1 . Miserocchi G, Agostoni E. Contents of the pleural space.
mors. Treatment of mice with established lung me ] Appl Physiol. 1 97 1 ;30:208-2 1 3 .
tastases significantly improved survival as compared 1 2 . Albertine KH, Wiener-Kronish JP, Staub N C . The structure
with control animals (66) . This group also demon of the parietal pleura and its relationship to pleural liquid dy
namics in sheep. Anat Rec. 1 984;208:40 1 -409.
strated in mice that unilateral intrapleural adminis 13. Wang PM, Lai-Fook SJ. Regional pleural filtration and ab
tration was sufficient to transfer genes bilaterally to sorption measured by fluorescent tracers in rabbits. Lung.
the pleura (66) . There are a few other studies that 1 999; 1 77:289-309.
have demonstrated the feasibility of using the pleu 14. Albertine KH, Wiener-Kronish JP, Roos PJ, et al. Structure,
ral space for gene transfer in animals, but the utility blood supply, and lymphatic vessels of the sheep's visceral
pleura. Am ] Anat. 1 9 82; 1 6 5 :277-294.
of this approach in humans with disease is yet to be 1 5 . Broaddus VC, Wiener-Kronish JP, Staub NC. Clearance of
demonstrated. lung edema into the pleural space of volume-loaded anesthe
A second therapeutic use of the pleural space is tized sheep. ] Appl Physiol. l 990;68 :2623-2630.
to warm individuals with accidental hypothermia. 16. Allen S, Gabel J, Drake R. Left atrial hypertension causes
pleural effusion formation in unanesthetized sheep. Am J
Kjaergaard and Bach (67) reported that they had suc
Physiol. 1 9 89;257(2 Pt 2):H690-H692.
cessfully warmed five patients with accidental hypo 1 7 . Wiener-Kronish JP, Matthay MA, Callen PW, et al. Rela
thermia, who were unconscious but who had a stable tionship of pleural effusions to pulmonary hemodynamics
heart rhythm with pleural lavage. They inserted bi in patients with congestive heart failure. Am Rev Respir Dis.
lateral chest tubes and then injected 500 mL isotonic 1 9 8 5 ; 1 32: 1 253- 1 256.
18. Brixey AG, Light RW. Pleural effusions occurring with right
saline at 40°C in one pleural space followed by clamp
heart failure. Curr Opin Pulm Med. 2 0 1 1 ; 1 7:226-23 1 .
ing of the chest tube for approximately 2 minutes. 1 9 . Wiener-Kronish J P, Broaddus VC, Albertine KH, et al.
After the tube was undamped, the procedure was re Relationship of pleural effusions to increased permeabil
peated on the other side. The pleural lavage was con ity pulmonary edema in anesthetized sheep. J Clin Invest.
tinued until the bladder temperature was above 40°C. 1 9 88;82: 1 422- 1 429.
20. Amouzadeh HR, Sangiah S, Qualls CW Jr, et al. Xylazine
All five patients survived and were discharged. The induced pulmonary edema in rats. Toxicol Appl Pharmacol.
amount of lavage varied between 32 and 1 02 L (67) . 1 99 1 ; 1 08 :4 1 7-427.
2 1 . Bernaudin JF, Theven D, Pinchon MC, et al. Protein transfer
in hyperoxic induced pleural effusion in the rat. Exp Lung Res.
REFERENCES 1 986; I 0:23-3 8 .
2 2 . Miller KS, Harley RA , Sahn SA. Pleural effusions associated
I . Light RW. Mechanics of Respiration. In: George RB, Light with ethchlorvynol lung injury result from visceral pleural
RW, Matthay MA, et al. , eds. Chest medicine: Essentials ofPul leak. Am Rev Respir Dis. 1 9 8 9 ; 1 40:764-768.
monary and Critical Care Medicine, 5 th ed. Philadelphia, PA: 23. Wiener-Kronish JP, Sakuma T, Kudoh I, et al. Alveolar
Lippincott Williams & Wilkins; 2005:24-3 8 . epithelial injury and pleural empyema in acute P aeru
2 . Feller-Kopman D, Ernst A . Pleural manometty. I n : Light ginosa pneumonia in anesthetized rabbits. J Appl Physiol.
RW, Lee YC (Gary), eds. Textbook ofPleural Diseases, 2nd ed. 1 993;7 5 : 1 6 6 1 - 1 669.
London: Holder Arnold; 2008:227-232. 24. Wiener-Kronish JP, Broaddus VC. Interrelationship of
3 . Lai-Fook SJ. Pleural mechanics and fluid exchange. Physiol pleural and pulmonary interstitial liquid. Ann Rev Physiol.
Rev. 2004;84:385-4 1 0 . 1 993;55:209-226.
4. Agostoni E, Zocchi L. Pleural liquid and its exchanges. Respir 25. Bhattacharya J, Gropper MA, Staub NC. Interstitial fluid
Physiol Neurobiol. 2007; 1 59:3 1 1-323. pressure gradient measured by micropuncture in excised dog
5 . Del Fabbro M . An improved technique for studying pleural lung. J Appl Physiol. 1 9 84;56:27 1 -277.
fluid pressure and composition in rabbits. Exp Physiol. 26. Payne DK, Kinasewitz GT, Gonzalez E. Comparative
1 998;83:43 5-448 . permeability of canine visceral and parietal pleura. J Appl
6. Benditt JO. Esophageal and gastric pressure measurements. Physiol. 1 9 88;65:25 5 8-2564.
Respir Care. 2005;50:68-77. 27. Kirschner PA. Porous diaphragm syndromes. Chest Surg Clin
7. Loring SH, O'Donnell CR, Behazin N, et al. Esophageal N Am. 1 998;8 :449-472.
pressures in acute lung injury-do they represent artifact or 28. Grimaldi A, Moriondo A, Sciacca L, et al. Functional arrange
useful information about transpulmonary pressure, chest wall ment of the rat diaphragmatic initial lymphatic network. Am J
mechanics, and lung stress? JAppl Physiol. 20 I O ; I 08:3 1 5-322. Physiol Heart Circ Physiol. 2006;29 l :H 876-H 8 8 5 .
18 P L E U RAL D I S EASES
29. Hodges CC, Fossum T'W, Evering W. Evaluation of thoracic 49. Romero-Candeira S, Fernandez C, Martin C, e t al. Infl u ence
duct healing after experimental laceration and transection. Vet of diuretics on the concentration of proteins and other com
Surg. 1 993;22:43 1--435. ponents of pleural transudates in patients with heart failure.
30. Nahid P, Broaddus VC. Liquid and protein exchange. In: Am ] Med. 200 1 ; 1 1 0:68 1 -686.
Light RW, Lee YC (Gary) , eds . : Textbook of Pleural Diseases. 5 0 . Stashenko GJ, Robichaux A, Lee YC, (Gary) et al. Pleural fluid
London: Arnold Publishers; 2003:33--44. exchange in rabbits. Respirology. 2007; 1 2 :495--499.
3 1 . Broaddus VC, Araya M, Carlton DP, et al. Developmental 5 1 . Light RW, Jenkinson SG, Minh V, et al. Observations on pleu
changes in pleural liquid protein concentration in sheep. Am ral pressures as fluid is withdrawn during thoracentesis. Am
Rev Respir Dis. 1 9 9 1 ; 1 43:38-4 1 . Rev Respir Dis. 1 980; 1 2 1 :799-804.
32. Cooray GH. Defensive mechanisms i n the mediastinum, 52. Cheng C-S, Rodriguez RM , Perkett EA. Vascular endo
with special reference to the mechanics of pleural absorption. thelial growth factor in pleural fluid. Chest. 1 9 99 ; 1 1 5 :
J Pathol Bacteriol. 1 949;6 1 : 5 5 1-567. 760-76 5 .
33. Courtice FC, Simmonds WJ. Absorption of fluids from the 5 3 . Thicken DR, Armstrong L , Millar AB. Vascular endothelial
pleural cavities of rabbits and cats.]Physiol. 1 949; 1 09: 1 1 7- 1 30. growth factor (VEGF) in inflammatory and malignant pleural
34. Burke H . The lymphatics which drain the potential space effusions. Thorax. 1 999;54:707-7 1 0 .
between the visceral and the parietal pleura. Am Rev Tuberc 5 4 . Eid AA , Keddissi J I , Kinasewitz G T. Hypoalbuminemia a s a
Pulmon Dis. 1 9 5 9;79: 52-6 5 . cause of pleural effusions. Chest. 1 999; 1 1 5 : 1 066- 1 069.
3 5 . Wang NS. Th e preformed stomas connecting the pleural cav 55. Allen SJ, Laine GA, Drake RE, et al. Superior vena caval pres
ity and the lymphatics in the parietal pleura. Am Rev Respir sure elevation causes pleural effusion formation in sheep. Am J
Dis. 1 975 ; 1 1 1 : 1 2-20. Physiol. 1 988;255:H492-H49 5 .
36. Miura T, Shimada T, Tanaka K, et al. Lymphatic drainage of 56. Verkman AS, Matthay MA, Song Y. Aquaporin water chan
carbon particles injected into the pleural cavity of the monkey, nels and lung physiology. Am J Physiol Lung Cell Mo! Physiol.
as studied by video-assisted thoracoscopy and electron micros 2000;278:L867-L879.
copy. J Thorac Cardiovasc Surg. 2000; 1 20:437--44 7. 57. Borok Z, Verkman AS . Lung edema clearance: 20 years
3 7. Li YY, Li JC. Ultrastructural study of pleural lymphatic drainage of progress: invited review: role of aquaporin water chan
unit and effect of nitric oxide on the drainage capacity of pleu nels in fluid transport in lung and airways. J Appl Physiol.
ral lymphatic stomata in the rat. Ann Anat. 2004; 1 86:25-3 1 . 2002;93:2 1 99-2206.
3 8 . Stewart PB. Th e rate o f formation and lymphatic removal of 58. Song Y, Yang B, Matthay MA, et al. Role of aquaporin water
fluid in pleural effusions. ] Clin Invest. 1 963;42:25 8-262. channels in pleural fluid dynamics. Am J Physiol Cell Physiol.
39. Leckie WJH, Tothill P. Albumin turnover in pleural effusions. 2000;279 : C l 744-C l 750.
Clin Sci. 1 965 ;29:339-352. 5 9 . Magnussen H, Perry SF, Willmer H , et al. Transpleural dif
40. Broaddus VC, Wiener-Kronish JP, Berthiauma Y, et al. fusion of inert gases in excised lung lobes of the dog. Respir
Removal of pleural liquid and protein by lymphatics in awake Physiol. 1 974;20 : 1 - 1 5 .
sheep. ] Appl Physiol. l 988;64:384-390. 6 0 . Northfield TC. Oxygen therapy fo r spontaneous pneumotho
4 1 . Aiba M, lnatomi K, Homma H . Lymphatic system or rax. Br Med] 1 9 7 1 ;4: 86-88.
hydro-oncotic forces. Which is more significant in drainage of 61. Gaensler EA. Parietal pleurectomy for recurrent spontaneous
pleural fl u id? Jpn ] Med. 1 984;23 :27-33. pneumothorax. Surg Gynecol Obstet. 1 9 56; 1 02:293-308.
42. Agostoni E. Mechanics of the pleural space. Physiol Rev. 62. Fleetham JA, Forkert L, Clarke H, et al. Regional lung func
l 972; 52:57- 1 2 8 . tion in the presence of pleural symphysis. Am Rev Respir Dis.
4 3 . Agostoni E, Zocchi L. Starling forces and lymphatic drain 1 980; 1 22:33-3 8 .
age in pleural liquid and protein exchanges. Respir Physiol. 63. West JB. Why doesn't the elephant have a pleural space? News
1 9 9 1 ;86 :27 1 -28 1 . Physiol Sci. 2002 ; 1 7:47-5 0.
44. Pistolesi M , Miniati M , Giuntini C . Pleural liquid and solute 64. Albelda S, Sterman DH. Gene therapy in pleural disease. In:
exchange. Am Rev Respir Dis. 1 9 8 9 ; 1 40:825-847. Light RW, Lee YC (Gary) , eds . : Textbook of Pleural Diseases.
45. Nakamura T, Iwasaki Y, Tanaka Y, et al. Dynamics of pleu London: Arnold Publishers; 2003 : 526-5 3 5 .
ral effusion estimated through urea clearance. Jpn J Med. 6 5 . Devin CJ, Lee YC, Light RW, e t al. Pleural space a s a site
l 987;26:3 1 9-322. of ectopic gene delivery: transfection of pleural mesothelial
46. Agostoni E, Bodega F, Zocchi L. Albumin transcytosis from cells with systemic distribution of gene product. Chest.
the pleural space. J Appl Physiol. 2002;93: 1 806- 1 8 1 2. 2003 ; 1 23:202-208.
47. Bodega F, Agostoni E. Contribution of lymphatic drainage 66. Mae M, Crystal RG . Gene transfer to the pleural mesothelium
through stomata to albumin removal from pleural space. as a strategy to deliver proteins to the lung parenchyma. Hum
Respir Physiol Neurobiol. 2004; 1 42:25 1 -263. Gene Ther. 2002; 1 3 : 1 47 1 - 1 482.
48. Shinto RA, Light RW The effects of diuresis upon the charac 67. Kjaergaard B, Bach P. Warming of patients with accidental
teristics of pleural fluid in patients with congestive heart fail hypothermia using warm water pleural lavage. Resuscitation.
ure. Am J Med. l 990;88:230-233 . 2006;68:203-207.
Physiologi cal E ffects of Pneumothorax
and Pl eural E ffusi on
In this chapter, the effects of pleural air or pleural to the hydrostatic column of fluid. Accordingly, the
fluid on pleural pressures, pulmonary function and pleural pressure with a pleural effusion in the depen
gas exchange, the diaphragm, the heart, and exercise dent part of the hemithorax is much greater than it is
tolerance will be discussed. in the superior part of the hemithorax. The net result
is that with a pneumothorax, the upper lobe is af
E F F E CTS OF P N E U M OT H O RAX O N
fected more than the lower lobe whereas with a pleu
P L E U RA L P R E S S U R E
ral effusion the lower lobe is affected more than the
upper lobes. The upper lobes are affected more with
Normally, the pressure i n the pleural space is negative pneumothorax because the pressure in the apices is
with reference to the atmospheric pressure during the normally much more negative than that at the bases.
entire respiratory cycle. The negative pressure is due With a pneumothorax the pleural pressures are only
to the inherent tendency of the lungs to collapse and slightly negative so there are much greater changes in
of the chest wall to expand. The resting volume of the pleural pressure at the apex of the lung. One way to
lung, the functional residual capacity (FRC) , is the conceptualize the difference between air and liquid
volume at which the outward pull of the chest wall is to understand that with a pneumothorax the lung
is equal, but opposite in direction, to the inward pull sinks to the bottom of the hemithorax because it is
of the lung with the respiratory muscles relaxed. In heavier than air, whereas with a pleural effusion, the
Figure. 3 . 1 , the FRC is at 36% of the vital capacity. lung rises to the top of the hemithorax because it is
The pleural pressure is always less than the alveolar lighter than the fluid and is floating in the fluid ( 1 ) .
pressure and the atmospheric pressure owing to the
elastic recoil of the lung. Therefore, if a communica
E F F E CTS O F P N E U M OT H O RAX O N
tion develops between the pleural space and an alveo
P U L M O N A RY F U N CT I O N
lus or between the pleural space and the atmosphere,
air will flow into the pleural space until a pressure gra When there i s a communication between the alveoli
dient no longer exists or until the communication is and the pleural space or between the ambient air and
sealed. Because the thoracic cavity is below its resting the pleural space, air will enter the pleural space be
volume and the lung is above its resting volume, with cause the pleural pressure is normally negative. As air
a pneumothorax, the thoracic cavity enlarges and the enters the pleural space, the pleural pressure gradually
lung becomes smaller. increases. Air will continue to enter the pleural space
When a pneumothorax is present, the pleural until the pleural pressure becomes zero or the com
pressure increases as it does with the presence of a munication is closed.
pleural effusion. However, with a pneumothorax the The influence of a pneumothorax on the volumes
pressure is the same throughout the entire pleural of the hemithorax and lung is illustrated in Figure 3. 1 .
space if it is not loculated. In contrast, with a pleural In the example, enough air entered the pleural space
effusion there is a gradient in the pleural pressure due to increase the pleural pressure from - 5 to - 2.5 cm
19
20 P L E U RAL D I S EASES
60
l
I
� I
> 20 fJ Lung volume
Pressure volume curve 1 (25% VC)
thoracic cavity l
I
I B
10 I
I
I
I
l
I
-20 -1 5 -1 0 -5 0 5 10 15 20
Pleural pressure
cm H20
H 20 at end expiration. The end-expiratory volume the pneumothorax, however, the decrease in the vital
of the lung (point B) decreased from 36% to 1 1 % capacity may lead to respiratory insufficiency with al
of the vital capacity, whereas the end-expiratory vol veolar hypoventilation and respiratory acidosis.
ume of the hemithorax (point C) increased from 36% Most patients with a pneumothorax have a re
to 44% of the vital capacity. The total volume of the duced Pa02 and an increased alveolar-arterial oxygen
pneumothorax is equal to 33% of the vital capacity, difference [P(A - a) O 2] . In one series of 1 2 patients,
of which 25% represents a decrease in lung volume the Pa0 2 was below 80 mm Hg in 9 patients (75%)
and 8% represents an increase in the volume of the and was below 55 mm Hg in 2 patients (2) . In the
hemithorax. There is essentially no information avail same series, 1 0 of the 1 2 patients (83%) had an in
able on the results of the pulmonary function tests of creased P(A - a)0 2 • As one would expect, patients
patients with pneumothorax since they rarely undergo with secondary spontaneous pneumothorax and those
pulmonary function testing while the pneumothorax with larger pneumothoraces tended to have a greater
is present. decrease in the Pa0 2 (2) . In the Veteran's Adminis
tration (VA) cooperative pneumothorax study, blood
gases were obtained in 1 1 8 patients with spontaneous
E F F E CTS O F P N E U M OT H O RAX O N
pneumothorax; the mean PaO 2 was below 5 5 mm Hg
B L O O D GAS E S
in 20 ( 1 7%) and below 45 in 5 (4%), and the mean
Th e main physiologic consequences o f a pneumo PaC0 2 exceeded 50 mm Hg in 1 9 ( 1 6%) and 60 mm
thorax are a decrease in the vital capacity and a de Hg in 5 (4%) (3) . Of course, the abnormalities in
crease in Pa0 2 • In the otherwise healthy individual, the blood gases may have been due at least in part to
the decrease in the vital capacity is well tolerated. If the underlying lung disease in this study (3) . Similar
the patient's lung function is compromised before findings are present in animals with pneumothoraces.
C HAPTE R 3 / PH YS I O LOG I CAL E F F E CTS OF P N E U M OT H O RAX A N D PLE U RAL E F F U S I O N 21
When a pneumothorax was induced in awake, stand tension pneumothorax, the oxygenation status re
ing dogs by the intrapleural injection of 50 mL/kg turns to normal almost immediately (6) .
N 2 , the mean Pa0 2 fell from 86 to 5 1 mm Hg (4) .
The reduction in PaO 2 appears to be due to E F F E CTS O F P N E U M OT H O RAX O N
both anatomic shunts and areas of low ventilation D I A P H RAG M AT I C F U N CT I O N
perfusion ratios in the partially atelectatic lung.
When Norris et al. (2) gave 1 00% oxygen to their 1 2 To my knowledge there have been n o studies evaluat
patients, the average anatomic shunt was more than ing the effects of a pneumothorax on diaphragmatic
1 0 % . The larger pneumothoraces were associated function. I would anticipate that the presence of a
with greater shunts (2) . Pneumothoraces occupying pneumothorax would have less effect on the dia
less than 25% of the hemithorax are not associated phragmatic function than would a pleural effusion of
with increased shunts. comparable volume, since the pleural pressure would
In the study on dogs conducted by Moran et al. increase much more with the pleural fluid. The dia
(4) , the relative perfusion of the lungs was not altered phragmatic inversion that is seen relatively frequently
when pneumothorax was induced, but the ventila with a pleural effusion is not seen with pneumotho
tion to the ipsilateral lung was reduced, resulting in rax. With a tension pneumothorax, the diaphragm
low ventilation-perfusion ratios on the side with the may be displaced inferiorly because of the increased
pneumothorax. Anthonisen (5) reported that lungs pleural pressure but the functional significance of this
with pneumothorax demonstrated uniform airway displacement is not known.
closure at low lung volumes, and he suggested that
airway closure is the chief cause of ventilation maldis
E F F E CTS O F P N E U M OT H O RAX O N
tribution in spontaneous pneumothorax.
E X E R C I S E TO L E RA N C E
The Pa0 2 usually improves with treatment of the
pneumothorax. In the animal study of Moran et al. There have been n o studies o n the effects o f a pneu
(4) in which the mean Pa0 2 dropped from 86 to mothorax on the exercise tolerance of either animals
5 1 mm Hg with the introduction of a pneumotho or man. However, it would be anticipated that the
rax, the Pa0 2 returned to baseline immediately after exercise tolerance would be markedly impaired since
reexpansion. In humans treated for pneumothorax, many patients are dyspneic at rest.
the normalization of the Pa0 2 takes longer. Three
patients with an initial anatomic shunt above 20%
E F F E CTS OF P N E U M OT H O RAX O N
had a reduction of at least 1 0% in their shunt 30 to
CA R D IAC F U N CT I O N
90 minutes after the removal of intrapleural air, but
it still remained above 5% in all patients (2) . Three Th e presence o f a small-to-moderate pneumothorax
additional patients with anatomic shunts of 1 0% to has very little influence on cardiac function. When
20% had no change in their shunts when the air was Moran et al. (4) introduced 50 mL/kg N 2 into the
removed (2) . The delay in improvement in humans as pleural spaces of dogs, the cardiac output was not
compared with animals may be related to the dura significantly affected. However, the presence of a ten
tion of the pneumothorax. sion pneumothorax in an animal can cause a marked
When a tension pneumothorax is produced in reduction in cardiac output. Carvalho et al. (7) pro
animals spontaneously breathing room air , there is duced right-sided tension pneumothoraces with
a profound deterioration in the oxygenation status. mean pleural pressures of + 1 0 and + 25 cm H2 0 in
In one study in goats, the mean Pa0 2 fell from 85 to 1 0 mechanically ventilated adult sheep. In these ani
28 mm Hg, whereas in monkeys the PaO 2 fell from mals, the mean cardiac output fell from 3.5 LIminute
90 to 22 mm Hg before the animals became apneic to approximately 1 .2 L/minute and the mean blood
(6) . There was a linear reduction in the Pa0 2 as the pressure fell from 80 mm Hg to less than 40 mm Hg
volume of pleural air was increased (6) . The reduc with a pleural pressure of + 2 5 cm H 2 0.
tion in the Pa0 2 appeared to be due to the continued The development of a tension pneumothorax in
perfusion of the side with the pneumothorax despite humans is also associated with impaired hemodynam
decreased ventilation (6) . The cardiac output was ics. Beards and Lipman (8) recorded the hemodynamics
relatively well preserved in the animals with a ten of three patients who developed a tension pneumotho
sion pneumothorax (6) . When the air is evacuated rax while on mechanical ventilation. With the develop
from the pleural space in experimental animals with ment of the tension pneumothorax, the mean cardiac
22 P L E U RAL D I S EASES
outputs that were 7.3, 4.8, and 3.6 L/minute/m2 the pleural pressure will lead to a decreased lung vol
at baseline fell to 3.0, 3. 1 , and 1 .4 L/minute/m2, re ume. Since the distending pressure of the heart is the
spectively. The mean arterial pressures that were 97, intracardiac pressure minus the pleural pressure, an
96, and 68 mm Hg fell to 33, 68, and 57 mm Hg, increase in the pleural pressure will lead to a decrease
respectively. The probable mechanism for the decreased in the size of the heart.
cardiac output is decreased venous return due to the The pleural pressure is normally negative. However,
increased pleural pressures. Moderate increases in the when more than minimal pleural fluid accumulates,
pleural pressure with a pneumothorax in conjunction the pleural pressure becomes positive. When there is
with thoracoscopy have little influence on the cardiac sufficient pleural fluid such that the lung is separated
output. Ohtsuka et al. (9) studied the hemodynamics from the chest wall, there is a vertical gradient of 1
while the lung was hemicollapsed and CO 2 was infused cm H 2 0/cm vertical height due to the weight of the
at a pressure of 8 to 1 0 mm Hg. The mean cardiac in fluid ( 1 0) . If there is a hydrostatic column 40 cm high
dex was virtually the same before and after 30 minutes in a hemithorax, then the pressure at the bottom of
of co2 infusion ( 1 .98 vs. 1 .95 L/minute/m2) (9) . the column would be expected to be approximately 40
cm H 2 0. When pleural pressures are measured in pa
tients with pleural effusions, the mean pressure is not
E F F E CTS O F E F F U S I O N O N T H E
particularly high. We measured the pleural pressure in
P L E U RA L P R E S S U R E
52 patients with significant pleural effusions (median
When pleural fluid i s present, its volume must b e com amount of fluid greater than 1 ,000 mL) . Overall, the
pensated for by an increase in the size of the thoracic mean pleural pressure was approximately zero, but
cavity, a decrease in the size of the lung, a decrease in there was a wide range in the pleural pressures from
the size of the heart, or a combination of these changes - 2 1 to + 8 cm Hp (Fig. 3.2) ( 1 1 ) . Pleural pressures
( 1 ) . Since the thoracic cavity, lungs, and heart are all of - 5 cm H 2 0 and less were seen only with a trapped
distensible objects, the volume of each is dependent lung or with malignancy. Villena et al. ( 1 2) measured
on the pressure inside minus the pressure outside. The the pleural pressure in 6 1 patients and reported that
presence of pleural fluid increases the pleural pressure. the initial pleural pressure ranged from - 1 2 to + 25
Since the distending pressure of the thoracic wall is the cm H 2 0. The mean pressure in the patients was ap
atmospheric pressure minus the pleural pressure, an proximately + 5 cm H 2 0 ( 1 2) . The probable reason
increase in the pleural pressure will lead to an increase that the pleural pressures were not more positive in
in the distending pressure of the thoracic cavity and the two studies is that the thoracentesis needle was in
an increase in the volume of the thoracic cavity. The serted closer to the superior than the inferior aspect of
distending pressure of the lungs is the alveolar pressure the hydrostatic column produced by the pleural effu
minus the pleural pressure. Therefore, an increase in sion. With a pleural effusion, pleural pressures can at
+1 0 -
0
-• •
•
0 •
0C\J
I
+5 -
- •
•
-
t.
E
.:..
+O - •
.2-
•
T
• •
Cii • 0
E •
FIGURE 3.2 • I n it i a l p l e u ra l p ressures for .s -5 - '•
0
•
•
�
::i
52 patie nts at the t i m e of t h o racentesis.
(/) •
Each patient i s represented by a s i n g l e poi nt. (/) -1 0 -
� • •
The o p e n c i rcles i n the category of t ra n su a.
•
d ates represent the patie nts with h e patic � -1 5 -
::i
hydrot h o rax. The c l osed c i rc l es i n the cat Q)
egory of m i sce l l a neous exud ates represent a:
-20 -
patients with p l e u ra l i nfect i o n . (Reprinted with •
permission from Light RW, Jenkinson SG, Minh VD, et
_;::
al. Observations on pleural fluid pressures as fluid is
withdrawn during thoracentesis. Am Rev Respir Dis. Transudates Malignancies Miscellaneous Trapped lung
1 980; 1 2 1 : 799-804.) exudates
C HAPTE R 3 / PH YS I O LOG I CAL E F F E CTS OF P N E U M OT H O RAX A N D PLE U RAL E F F U S I O N 23
times be quite positive. Neff and Buchanan ( 1 3) re pleural space elastances. If one looks at the plot of the
ported that the initial pleural pressure was 76 cm H 2 0 pleural pressure versus the volume of fluid removed
in a patient with a pleural effusion secondary to pneu (Fig. 3.3), the elastance tends to be higher during the
mothorax therapy for tuberculosis many years earlier. latter part of the thoracentesis ( 1 1 , 1 2) .
When pleural fluid is removed with thoracente Clinically, i t is useful to measure the pleural
sis, the volume removed is compensated for by an pressure and calculate the pleural elastance during
increase in the volume of the lung, an increase in the a thoracentesis. The demonstration that the pleural
volume of the heart, and/or a decrease in the volume elastance is greater than 25 cm H 2 0/L establishes the
of the hemithorax. When the volume of these organs diagnosis of trapped lung ( 1 1 , 1 2) . Thoracentesis can
changes in this manner, the pleural pressure must de be continued safely as long as the pleural pressure
crease. When the pleural pressure is monitored during remains above - 20 cm H 2 0 and the patient does
pleural fluid removal, there is tremendous variability not develop chest tightness or pernicious coughing
in its changes from patient to patient ( 1 1 , 1 2) . ( 1 1 , 1 2) . Indeed, on several occasions I have removed
Th e elastance o f the pleural space has been defined more than 5,000 mL pleural fluid from patients when
as the change in pleural pressure (cm H 2 0) divided the pleural pressure remained above - 20 cm H 2 0
by the amount of fluid removed (liters) ( 1 1 ) . The and the patients suffered no ill consequences.
larger this number the greater the pleural pressure Measurements of the pleural space elastance ap
change per unit volume change. In our original series pear to be useful in predicting whether a pleurodesis
of 52 patients, the pleural space elastance varied from will be successful ( 1 4) . The theory is that if the pleural
2 to more than 1 50 cm H 2 O/L with a mean elastance pressure falls rapidly when fluid is removed from the
of approximately 1 5 cm H2 0/L ( 1 1 ) . Patients with pleural space, then the negative pleural pressure will
trapped lungs due to malignancy or benign disease make it difficult to create a pleurodesis because the two
had pleural space elastances that exceeded 25 cm pleural surfaces will be difficult to be kept together
H 2 0/L. Villena et al. ( 1 2) reported similar values for (which is necessary to create a pleurodesis) . Lan et al.
+1 0
....
'<'.\.
0
I
�
"'
-1 0
E
�
� I
::::> I
(/) -20
(/) I
� I
a. I
I
� -30
::::>
Q)
a:
---4 0
t
I
I
I
I
I
\
I
I
I
-50
400 800 1 ,200 1 ,600 2,000 2,400 2,800 3,200 3,600
P l e u ral fluid withdrawal (ml)
( 1 4) measured the change in pleural pressure after 500 pulmonary dysfunction is due to the pleural effusion
mL of pleural fluid had been withdrawn in 65 patients and what part is due to the underlying disease.
with a pleural malignancy. They then inserted a chest There have been a few studies on the effects of
tube and continued to drain the pleural space until (a) a pleural effusion on the pulmonary function of
the drainage was less than 1 50 mL/day, (b) the drain animals. Krell and Rodarte ( 1 5) studied the volume
age was less than 250 mL/day for four consecutive changes in the lung and thorax of dogs after 200 to
days, or (c) the drainage had continued for 1 0 days. 1 ,200 mL fluid was added to the right hemithorax.
After one of the three criteria was met, they attempted They found that the decrease in lung volume at FRC
pleurodesis if the lung had expanded. They reported was approximately one third of the added saline vol
that the lung did not reexpand (trapped lung) in 1 1 ume, whereas the decrease in the lung volume at to
of the 1 4 patients who had a pleural elastance greater tal lung capacity (TLC) was one fifth of the added
than 1 9 cm H2 0/L ( 1 4) . Pleurodesis was attempted saline volume. Consequently, the chest wall volume
in the other three patients with a high pleural elas increased by two thirds of the added saline volume at
tance and it failed in all three. In contrast, only 3 of FRC and by four fifths of the added saline volume at
5 1 patients with pleural elastance less than 1 9 cm TLC ( 1 5) . The decrease in the upper lobe volume was
H 2 0/L had a trapped lung, and pleurodesis was suc less than that of the lower lobe volume ( 1 5) .
cessful in 42 of 43 patients (98%) who returned for There have been several studies concerning the
evaluation ( 1 4) . pulmonary function of patients with pleural effusions.
We measured the pulmonary function in 1 5 patients
with moderate to large pleural effusions and found
E F F E CTS O F E F F U S I O N O N
that the mean forced expiratory volume in 1 second
P U L M O N A RY F U N CT I O N
(FEV) and the forced vital capacity (FVC) were only
Th e effects o f a pleural effusion o n pulmonary func 43% ± 1 7% and 49% ± 1 7% of the predicted val
tion are difficult to determine. Many diseases that ues respectively ( 1 6) . Seven of the 1 5 patients had
cause pleural effusions, such as congestive heart failure, obstructive lung disease as reflected by an FEV/ FVC
malignancy, pneumonia and pulmonary embolism, ratio less than 0.70. Estenne et al. ( 1 7) reported that
also affect the pulmonary parenchyma. Therefore, it the FVC was less than 50% of the predicted value in
is frequently difficult to determine what part of the all nine patients with a large pleural effusion.
TABLE 3 . 1 • Resu lts of M axi m a l Exe rcise Tests B efore a n d After a T h e ra peutic T h o racentesis
in 1 5 Pat i e nts from W h o m a Mean of 1 , 61 2 ml P l e u r a l F l u i d was Rem oved
FEV1 , forced expi ratory vol u m e i n 1 seco n d ; FVC , forced vita l capac ity; V o2 m ax, m a xi m u m oxyge n consu m pt i o n ;
V E max, m a x i m u m exp i red vo l u m e p e r u n it t i m e ; Vco2, c a r b o n d i oxide cons u m pt i o n p e r u n i t t i m e ; H R, h e a rt rate; SW,
,
s a m e workloa d .
C HAPTE R 3 / PH YS I O LOG I CAL E F F E CTS OF P N E U M OT H O RAX A N D PLE U RAL E F F U S I O N 25
The pulmonary function abnormalities are not Although on the average there is a 20-mL in
completely corrected by performing a therapeutic crease in the FVC for every 1 00 mL of pleural fluid
thoracentesis. We performed pulmonary function withdrawn, much interindividual variability exists
tests before and 24 hours after thoracentesis in 1 5 pa (Fig. 3.4) . Changes in the FVC are related to pressure
tients from whom a mean of 1 ,6 1 2 mL pleural fluid measurements during thoracentesis ( 1 8) . Patients
was withdrawn (Table 3 . 1 ) ( 1 6) . In this study, the with higher initial pleural pressures and patients with
mean FEV1 and FVC, expressed as a percent of pre smaller changes in the pleural pressure as fluid was
dicted value, increased from 43% to 47% and from removed tended to have larger improvements in their
49% to 56%, respectively ( 1 5) . In a second study, we vital capacity. Nevertheless, when multiple regres
measured the pulmonary function in 26 patients be sion was performed with the change in the FVC as
fore and 24 hours after thoracentesis during which the dependent variable and the initial pleural pres
a mean of 1 ,740 mL pleural fluid was removed. In sure, the pleural elastance, and the amount of fluid
this study, the FVC and the FEV 1 each increased by removed as the independent variables, the multiple
approximately 400 mL while the TLC improved by regression coefficient ( r) never exceeded 0.60 ( 1 8) .
approximately 700 ml ( 1 8) . In other words, for ev This indicates that less than 40% o f the variance in
ery 1 ,000 mL pleural fluid removed, FVC and FEV1 the change in the FVC was related to the amount of
improved by approximately 200 mL and the TLC fluid removed and the measures of pleural pressure.
improved more than the FVC or FEV 1 • Our results Possible explanations for the poor correlation are
are in contrast to the dog study outlined here, where that (a) the pulmonary function testing was not per
the TLC was impacted less than the FRC. A possible formed until 24 hours after the thoracentesis and the
explanation for the varying results in humans and in fluid might have returned to a variable degree at this
dogs is that in humans, the lower lobe is frequently time, and (b) the pleural pressure changes recorded
completely atelectatic when a large pleural effusion is reflected the elastance of the pleural space during the
present. When fluid is then removed, the lower lobe thoracentesis. If the lung had been atelectatic for a
reexpands and the residual volume in the reexpanded prolonged period, it may take several hours or days
lobe increases the TLC more than just the vital capac for the lung to reexpand, and (c) in some patients
ity of the lobe. with large effusions, there is mediastinal shift to the
1 ,000
:::J
s
0
>
LL
.5
E
Q)
E
Q) 500
>
e
c.
E
opposite side, but this does not always occur with A therapeutic thoracentesis has essentially no ef
malignancy because of fixation of the mediastinum fect on the diffusion capacity of the lung (DLCO)
by the malignant process ( 1 8) . ( 1 8) or the specific airway conductance ( 1 7, 1 8) .
Estenne et al. ( 1 7) measured the changes i n respi
ratory mechanics in nine patients 2 hours after the
E F F E CTS O F E F F U S I O N O N B LO O D
removal of a mean 1 , 8 1 8 mL pleural fluid. They re
GAS E S
ported a mean increase of 300 mL in the vital ca
pacity, which was similar to what we have reported. Although patients with pleural effusions frequently
They also found that the TLC increased approximately have abnormal arterial blood gas results, the perfor
twice as much as did the FVC ( 1 7) . In addition, they mance of a therapeutic thoracentesis has relatively
studied the maximal pressures generated by the in little effect on the arterial blood gas results.
spiratory muscles and found that postthoracentesis In animals, the experimental induction of a
the pressures generated were much greater at a given pleural effusion can have a profound effect on the
lung volume. The greater inspiratory pressures were blood gases if the effusion is large enough. Nishida et
attributed to a decrease in the thoracic cage volume al. (20) induced bilateral pleural effusions in pigs and
( 1 7) . Although a mean of 1 ,8 1 8 mL pleural fluid was reported that there was a mild decrease in Pa0 2 (525
removed, the mean TLC increased by only 640 mL, to 375 mm Hg on 1 00% 0 2) , while the PaC0 2 re
indicating that the thoracic cage volume decreased by mained stable as the total amount of pleural fluid was
1 ,8 1 8 - 640 = 1 , 1 78 mL. The changes in the in increased to 30 mL/kg. However, when additional
spiratory pressures were significant. For example, the fluid was added, Pa0 2 dropped precipitously and
maximal inspiratory pressure (MIP) at TLC was - 1 6 PaC0 2 started to increase (20) . When the amount of
cm H 20 before thoracentesis and increased to - 25 pleural fluid reached 80 mL/kg, the Pa0 2 (on 1 00%
cm H 2 0 after thoracentesis, whereas the highest MIP oxygen) was less than 80 mm Hg, whereas the PaC0 2
went from - 4 1 cm H 2 0 before thoracentesis to - 52 had increased from 34 to 5 1 mm Hg. When the pleu
cm H 20 postthoracentesis ( 1 7) . I believe that one ral fluid was removed in these normal pigs, PaO 2 and
factor that allows the patient to generate more pres PaC02 rapidly returned to normal (20) .
sure postthoracentesis is that the thoracentesis can Things seem more complicated in humans prob
relieve the downward displacement of the diaphragm ably because of disease in the underlying lung and
by the pleural fluid. the fact that the pleural effusion has been present for
Wang and coworkers ( 1 9) have provided support much longer. Brandstetter and Cohen (2 1 ) obtained
for the last hypothesis. These researchers compared blood gases on 1 6 patients before, and then 20 min
the changes in pulmonary function before and 24 utes, 2 hours, and 24 hours after a thoracentesis in
hours after a therapeutic thoracentesis in 2 1 patients which 1 50 to 1 ,600 mL pleural fluid was removed.
with pleural effusion and paradoxical movement of They reported that the mean PaO 2 at baseline was
a diaphragm and 41 patients without paradoxical 70.4 mm Hg, and this decreased significantly to
movement of a diaphragm. Although the amount of 6 1 .2 mm Hg 20 minutes following thoracentesis.
pleural fluid removed did not differ in the two groups, In every patient there was a decrease in Pa0 2 over
post thoracentesis the paradoxical movement group this 20-minute period. Pa0 2 remained significantly
had significant improvements in the FEV1 , the FVC, reduced at 2 hours (64.4 mm Hg) but returned to
the A-a 0 2 gradient, and the Borg score while the baseline 24 hours later (2 1 ) . In this study there were
group without paradoxical movement did not have no significant changes in the pH or the PaC0 2 (2 1 ) .
a significant change in any parameter. They found Conflicting results were reported i n a study (22) of
that the mean FVC increased by 260 mL, whereas 1 9 patients with adult respiratory distress syndrome
the mean FEV1 increased by 2 1 0 mL in the paradoxi (ARDS) on mechanical ventilation who had pleu
cal movement group, which were changes similar to ral effusions and refractory hypoxemia. When chest
those observed by Estenne et al. ( 1 7) and our group tubes were inserted to drain the fluid, Pa0 / Fi0 2
( 1 6, 1 8) . Interestingly, the patients in this study with improved from 1 5 1 ± 1 3 to 245 ± 29 mm Hg (22) .
paradoxical diaphragmatic movement were very In addition, the dynamic compliance of the lung in
dyspneic before thoracentesis and their dyspnea im creased from 27. 1 to 35. 7 mL/cm H 2 0 immediately
proved markedly following the thoracentesis. They after the insertion of the chest tubes (22) . However,
attributed the decreased dyspnea to the fact that the in another study on eight patients with pleural ef
diaphragm was no longer inverted. fusions of various etiologies receiving mechanical
C HAPTE R 3 / PH YS I O LOG I CAL E F F E CTS OF P N E U M OT H O RAX A N D PLE U RAL E F F U S I O N 27
I j n •• • I I ••
ventilation, there was no improvement in the alveo i n this category are usually asymptomatic although
lar-arterial oxygen gradient [P(A-a) 0 2] after a thera the effusion may be large. In the second category,
peutic thoracentesis (23) . the diaphragm is flattened and does not move with
Why do patients with pleural effusions have hy respiration. Patients in this category frequently com
poxemia? Agusti et al. (24) used the multiple inert plain of dyspnea, which is likely to be relieved with
gas technique in an attempt to determine the mecha a therapeutic thoracentesis. In the third category, the
nisms for hypoxemia with pleural effusions. They diaphragm is inverted, and there may be paradoxical
studied the nine patients before and immediately af movements on respiration. Patients in this category
ter a thoracentesis (mean 693 ± 424 mL) (24) . They usually have severe dyspnea that is markedly relieved
reported that the mean P(A-a)0 2 was 29 mm Hg be with a therapeutic thoracentesis.
fore thoracentesis and remained at 29 mm Hg post The percentage of patients in each of the three cat
thoracentesis (24) . These investigators also showed egories has not been studied carefully. Inversion of the
that the main mechanism underlying arterial hypox diaphragm may be more common than is generally
emia in patients with pleural effusion is an intrapul realized. Wang and coworkers ( 1 9) were able to docu
monary shunt, which does not change significantly ment diaphragmatic inversion in 2 1 patients over a
after thoracentesis. It is likely that the intrapulmo 3-year period. Interestingly, when these patients un
nary shunt results from blood flowing through an derwent therapeutic thoracentesis, they experienced
atelectatic lung. marked relief of their dyspnea.
There have been four studies that examined the
effects of position on the oxygenation status of pa E F F E CTS OF E F F U S I O N ON T H E H E A RT
tients with pleural effusions (25-28) . In each study,
the oxygenation status was slightly better when the The presence of pleural fluid may also adversely in
patients were positioned with the side of the effusion fluence cardiac function. Vaska et al. (30) studied
superior (Table 3.2) . The improvement was thought seven spontaneously breathing dogs with a two-di
to be due to the effect of gravity distributing more mensional echocardiograph during the infusion of
blood to the lung that was not partially compressed saline into both pleural spaces. In this study, right
by the pleural effusion. However, the differences in ventricular diastolic collapse began when the mean
the mean levels of the oxygenation were not statisti pleural pressure increased by 5 mm Hg. By the time
cally significant in any of the studies and are probably the mean pleural pressure increased by 1 5 mm Hg,
not clinically significant either. the stroke volume had fallen by nearly 50%, and the
cardiac output had fallen by 33% (30) . In contrast,
Nishida et al. reported that the infusion of 20 mL/kg
E F F E CTS O F E F F U S I O N O N T H E
saline into each pleural space had no effect on the car
D I A P H RAG M
diac output in anesthetized pigs (20) . In the Nishida
The presence of fluid in the pleural space can pro study, the pleural pressure increased only by 3 mm
foundly affect the function of the ipsilateral dia Hg (20) as compared with 1 5 mm Hg in the Vaska
phragm because of the weight of the fluid on the study (30) . It should be noted that Vaska et al. in
diaphragm. The changes in the diaphragm have been fused more than 50 mL/kg saline into the pleural
classified into three categories by Mulvey (29) on the spaces of their dogs (30) .
basis of the findings on the plain film and fluoros The frequency with which large pleural effusions
copy. In the first or least severe category, the hemidia adversely affect cardiac function is not clear. Traylor
phragm is domed and it functions normally. Patients et al. (3 1 ) studied 27 patients who had more than a
28 P L E U RAL D I S EASES
hemithorax occupied by pleural fluid. Before thora (35). Kopterides et al. (36) reported two patients who
centesis, some of the patients had clinical or echocar had hemodynamic compromise with large left-sided
diographic signs of cardiac tamponade. Eight subjects effusions in whom transthoracic echocardiography
had elevated jugular venous pressure, 8 had pulsus demonstrated left ventricular diastolic collapse. It is
paradoxus, 6 had right ventricular diastolic collapse probable that the increased pleural pressure resulting
and 23 had flow velocity paradoxus. After thoracen from the pleural fluid is responsible for the ventricular
tesis, these abnormalities resolved in all but one of the collapse and the decreased cardiac output.
patients (3 1 ) . In most instances, however, the presence
of a moderate or large pleural effusion occupying less E F F E CTS O F E F F U S I O N O N E X E R C I S E
than an entire hemithorax does not adversely affect TO L E RA N C E
the cardiac output. Sadaniantz et al. (32) reviewed the
echocardiograms of 1 1 6 patients with pleural effusion There has been limited research o n the effects o f a
and reported that 21 ( 1 8%) had right atrial collapse. pleural effusion on the exercise tolerance of patients.
Of those with right atrial collapse, one had concomi We obtained maximum exercise tests on 1 5 patients
tant right ventricular collapse, four had left atrial col with moderate to large pleural effusions before and
lapse and none had left ventricular collapse (32) . Of after they underwent a therapeutic thoracentesis. The
the 21 patients with chamber collapse, 13 had large, symptom-limited exercise tests were conducted on
3 had moderate and 2 had small left pleural effusions. a bicycle ergometer with 1 5-watt increments every
Chamber collapse was unusual with right-sided pleu minute ( 1 6) . The mean age of the patients was 64.7
ral effusion (32) . It was not clear whether the cham and most of them had malignant pleural effusions.
ber collapse compromised the hemodynamics in any The exercise tolerance of these elderly patients was
of these patients (32) . Ahmed et al. (33) studied the significantly reduced before the thoracentesis. The
hemodynamics and the oxygen delivery in 22 me mean maximum workload was only 79 watts (43%
chanically ventilated patients who had moderate or of predicted) while the mean maximum Vo 2 was
large pleural effusions before and after they under only 907 mL/minute (37% of predicted) (Table 3 . 1 ) .
went drainage (mean 1 ,262 mL) of the effusions with When the individual exercise tests were examined, the
a pigtail catheter. They reported that the mean car explanation for the reduced exercise tolerance was not
diac output increased from 7.7 to 8.4 L/minute, but obvious. Eight of the patients appeared to be ventila
this change was not statistically significant (33) . The tory limited (VEmax greater than 80% of predicted
pulmonary capillary wedge pressure and the central maximum at exhaustion) , and four of these also ap
venous pressure both decreased significantly after the peared to be cardiac limited (maximum heart rate
pleural fluid was drained (33) . greater than 80% of predicted at exhaustion) . There
There have been three reports of patients with life were two additional patients who appeared to be only
threatening compromised cardiac output attributed cardiac limited. At the maximum tolerated work
to large pleural effusions (34-36) . Negus et al. (34) load (Emax) the remaining five patients appeared to
reported a 60-year-old woman who presented with a be neither ventilatory nor cardiac limited. In gen
large left-sided pleural effusion with marked media eral, the patients' ventilation was inefficient as evi
stinal shift to the right. Shortly after presentation, her d� nc� d by the high ventilatory � quiyalents for oxygen
blood pressure became unobtainable, and her carotid (VE/Vo2) and carbon dioxide (VE/Vco 2) (Table 3 . 1 ) .
and femoral pulses were very weak. When a chest tube I n addition, the patients' cardiac function appeared
was placed and 1 , 1 25 mL of pleural fluid was with to be impaired as evidenced by the high resting pulse
drawn, the blood pressure rose to 1 40/86 mm Hg and (Table 3 . 1 ) and the reduced oxygen pulse (o 2 pulse) ,
the pulses became bounding (34) . Kisanuki et al. (35) which is a reflection of the stroke volume.
reported a 68-year-old man who presented with a blood The exercise tolerance of these 1 5 patients did not
pressure of 90160 and a large left encapsulated pleural increase impressively after the performance of thera
effusion. A two-dimensional echocardiogram revealed peutic thoracentesis (mean, 1 , 6 1 2 mL) (Table 3 . 1 ) .
that the pleural effusion compressed the lateral wall of Although the mean FEV1 and FVC both improved sig
the left ventricle. With M-mode echocardiogram, the nificantly (Table 3 . 1 ) , there was no significant change
left ventricular collapse was observed throughout di in the maximum workload or the Vo 2max. Overall,
astole. After drainage of 500 mL of pleural fluid, the after thoracentesis five patients had an improvement
blood pressure rose from 90/60 to 1 20/80 mm Hg and in their workload, five patients had a decrease in the
the left ventricular collapse during diastole resolved workload, and five patients had no change in the
C HAPTE R 3 / PH YS I O LOG I CAL E F F E CTS OF P N E U M OT H O RAX A N D PLE U RAL E F F U S I O N 29
workload. The change in exercise capacity was not 1 1 . Light RW, Jenkinson SG, Minh VD, et al. Observations on
significantly correlated with the amount of fluid re pleural fluid pressures as fluid is withdrawn during thoracen
tesis. Am Rev Respir Dis. 1980; 1 2 1 :799-804.
moved or with the changes in pleural pressure. How 12. Villena V, Lopez-Encuentra A, Pozo F, et al. Measurement of
ever, there was a significant correlation with changes pleural pressure during therapeutic thoracentesis. Am J Respir
in the FEV 1 ( r = 0. 576, p < 0.05), changes in the Crit Care Med. 2000; 1 62: 1 534- 1 5 3 8 .
FVC ( r = 0.6 1 0, p < 0.05), and changes in the maxi 1 3 . Neff TA, Buchanan BO. Tension pleural effusion: a delayed
mum 02 pulse ( r = 0.78, p < 0.05). complication of pneumothorax therapy in tuberculosis. Am
Rev Respir Dis. 1 973; 1 1 1 : 543-548.
In summary, based on this series, elderly patients 1 4 . Lan RS, Lo SK, Chuang ML, et al. Elastance of the pleural space:
with moderate to large pleural effusions have a marked a predictor for the outcome of pleurodesis in patients with ma
reduction in their exercise capacity. Lung function, as lignant pleural effusion. Ann Intern Med 1 997; 1 26:768-774.
reflected by the FEV1 and the FVC, and cardiac func 1 5 . Krell WS, Rodarte JR. Effects of acute pleural effusion on respira
rorysystemmechanics indogs. ]ApplPhysiol. 1 985;59: 1458-1463.
tion, as reflected by the 0 2 pulse, are both reduced 16. Shinto RA, Stansbury OW, Brown SE, et al. Does therapeutic
and contribute to the exercise limitation. However, thoracentesis improve the exercise capaciry of patients with
the performance of a therapeutic thoracentesis results pleural effusion? Am Rev Respir Dis. 1 987; 1 3 5 :A244.
in no significant improvement in the mean exercise 17. Estenne M, Yernault J-C, De Troyer A. Mechanism of relief
tolerance of the patients. of dyspnea after thoracocentesis in patients with large pleural
effusions. Am ] Med. 1 9 83;74: 8 1 3-8 1 9 .
In a second study, Cartaxo and coworkers (37) 1 8 . Light RW, Stansbury OW, Brown SE. The relationship be
measured the 6-minute walking distance before and tween pleural pressures and changes in pulmonary func
48 hours following a thoracentesis in 25 patients tion following therapeutic thoracentesis. Am Rev Respir Dis.
from whom a mean of 1 ,564 ml pleural fluid was 1 986; 1 33:658-66 1 .
removed. The 6-minute walking distance increased 1 9 . Wang L-M, Cherng J-M, Wang J-W Improved lung function
after thoracocentesis in patients with paradoxical movement
from 432 m to 495 m (p < 0.00 1 ) after the thora of a hemidiaphragm secondary to a large pleural effusion. Res
centesis. The mean Borg score at the end of the walk pirol. 2007; 1 2 : 7 1 9-723.
decreased from 5 . 1 to 2.4. The fluid removal had no 20. Nishida 0, Arellano R, Cheng DC, et al. Gas exchange and
significant effect on the Spo 2 (37) . hemodynamics in experimental pleural effusion. Crit Care
Med. 1 999;27:5 83-5 87.
2 1 . Brandstetter RD, Cohen RP. Hypoxemia after thora
REFERENCES centesis: a predictable and treatable condition. JAMA.
1 979;242: 1 060- 1 06 1 .
I . Light RW. Physiological effects of pleural air or fluid. In: Light 22. Talmor M , Hydo L , Gershenwald JG, e t al. Beneficial effects
RW, Lee YCG, eds. Textbook of Pleural Diseases. London, of chest tube drainage of pleural effusion in acute respirarory
England: Arnold Publishers; 2003:45-5 5 . failure refractory to positive end-expiratory pressure ventila
2 . Norris RM, Jones J G , Bishop J M . Respiratory gas ex tion. Surgery. 1 998; 1 23 : 1 37- 1 43 .
change in patients with spontaneous pneumothorax. Thorax. 23. Doelken P, Abreu R , Sahn SA, e t al. Effect of thoracentesis o n
1 968;23 :427-433. respiratory mechanics and gas exchange in the patient receiv
3 . Light RW, O'Hara VS , Moritz TE, et al. Intrapleural tetracy ing mechanical ventilation. Chest. 2006; 1 30: 1 354-1 36 1 .
cline for the prevention of recurrent spontaneous pneumotho 24. Agusti AG, Cardus J , Roca J , et al. Ventilation-perfusion mis
rax. JAMA. l 990;264:2224-2230. match in patients with pleural effusion: effects of thoracente
4. Moran JF, Jones RH, Wolfe WG . Regional pulmonary func sis. Am ] Respir Crit Care Med. 1 997; 1 56 : 1 205-1 209.
tion during experimental unilateral pneumothorax in the 2 5 . Sonnenblick M , Melzer E, Rosin AJ. Body positional ef
awake state. J Thorac Surg. 1 977;74:396--402. fect on gas exchange in unilateral pleural effusion. Chest.
5 . Anthonisen NR. Regional function in spontaneous pneumo 1 9 83;83:784-786.
thorax. Am Rev Respir Dis. 1 977; 1 1 5 : 873-876. 26. Neagley SR, Zwillich CW The effect of positional changes
6. Rurherford RB, Hurt HH, Brickman RD, et al. The patho on oxygenation in patients with pleural effusions. Chest.
physiology of progressive, tension pneumothorax. J Trauma. 1 9 8 5 ; 8 8 : 7 1 4-7 1 7 .
1 968;8:2 1 2-227. 2 7 . Chang SC, Shiao GM, Perng RP. Postural effect on gas ex
7. Carvalho P, Hilderbrandt J, Charan NB. Changes in bronchial change in patients with unilateral pleural effusions. Chest.
and pulmonary arterial blood flow with progressive tension l 989;96:60-63.
pneumothorax. ] Appl Physiol. 1 996;8 1 : 1 664- 1 669. 28. Romero S, Martin C, Hernandez L, et al. Effect of body posi
8 . Beards SC, Lipman J. Decreased cardiac index as an indicator tion on gas exchange in patients with unilateral pleural effusion:
of tension pneumothorax in the ventilated patient. Anaesthe influence of effusion volume. Respir Med. 1 995;89:297-30 1 .
sia. 1 994;49 : 1 37- 1 4 1 . 29. Mulvey RB. The effect o f pleural fluid o n the diaphragm. Ra
9 . Ohtsuka T, Nakajima J , Kotsuka Y, e t al. Hemodynamic re diology. 1 965;84: 1 080- 1 086.
sponses to intrapleural insuffiation with hemipulmonary col 30. Vaska K, Wann LS, Sagar K, et al. Pleural effusion as a cause of
lapse. Surg Endosc. 200 1 ; 1 5 : 1 327-1 330. right ventricular diastolic collapse. Circulation. l 992;86:609---0 17.
10. Agostoni E, D'Angelo E. Thickness and pressure of the pleu 3 1 . Traylor JJ, Chan K, Wong I, et al. Large pleural effusions pro
ral liquid at various heights and with various hydrothoraces. ducing signs of cardiac tamponade resolved by thoracentesis.
Respir Physiol. I 969;6:330-342 . Am J Cardiol. 2002;89: I 06- 1 0 8 .
30 P L E U RAL D I S EASES
32. Sadaniantz A, Anastacio R, Verma V, et al. The incidence of 35. Kisanuki A, Shono H, Kiyonaga K, et al. Two-dimensional
diastolic right atrial collapse in patients with pleural effu echocardiographic demonstration of left ventricular diastolic
sion in the absence of pericardia! effusion. Echocardiography. collapse due to compression by pleural effusion. Am Heart].
2003;20:2 1 1 -2 1 5 . 1 9 9 1 ; 1 22: 1 1 73-1 1 7 5 .
3 3 . Ahmed S H , Ouzounian S P, Dirusso S, e t al. Hemodynamic 36. Kopterides P, Lignos M, Papanikolaou S, e t al. Pleural effusion
and pulmonary changes after drainage of significant pleural causing cardiac tamponade. Report of two cases and review of
effusions in critically ill, mechanically ventilated surgical pa the literature. Heart Lung. 2006;3 5 :66-67.
tients. J Trauma. 2004; 57: 1 1 84- 1 1 8 8 . 37. Carraxo AM, Vargas FS, Salge JM, et al. Improvements in
3 4 . Negus RA , Chachkes J S , Wrenn K . Tension hydrothorax the six-minute walk rest and spirometry following thora
and shock in a patient with a malignant pleural effusion. Am centesis for symptomatic pleural effusions. Chest. 2 0 1 O; 1 39 :
] Emerg Med. 1 990;8:20 5-207. 1 424- 1 429.
An i mal M od els i n Pl eural I nvestigati on
The pleural space is involved in many different dis an attempt is made to outline the more common
ease processes. Animal models of pleural disease offer models that have been developed and to give exam
the opportunity to answer many questions regard ples of the type of information that has been obtained
ing the pathophysiology and treatment of pleural with their use.
diseases. There are several advantages to studying
animal models rather than human disease. Animals
M A L I G N A N CY
are certainly easier to recruit. Variables can be more
closely controlled in animal models than in humans. In general, there are three types of experimental mod
This reduces variance in the results and allows con els of malignant pleural effusion. In the first type, cells
clusions to be made with a smaller number of study derived from spontaneously occurring tumors are
participants. Investigational agents can be instilled in injected into the species from which they are derived
animals with much less bureaucracy than when they in an attempt to produce a malignant pleural effusion.
are instilled in humans. The animals are immunocompetent. In the second
When performing experiments in animals, it is type, tumor cells from different species are injected
important to select the appropriate species. Investi intravenously or intrapleurally. Most animals in which
gators should be aware of two important anatomical malignancies have been induced have some type of
characteristics of the species they use. First, many ani immunologic deficiency. In the third type, malig
mals, such as mice, have incomplete mediastinal sepa nant pleural mesothelioma develops after animals are
rations with the left and right pleural spaces being in exposed to asbestos particles (intrapleurally, inhala
free communication ( 1 ) . This eliminates the possibil tionally, intratracheally, or intraperitoneally) (2) .
ity of using the contralateral pleural space as a con
trol. Second, some animals such as sheep have a thick
Adenocarcinoma
visceral pleura resembling that of humans, whereas
other animals such as dogs, rabbits, and mice have a In Mice. The best model malignant pleural effu
thin visceral pleura. This difference in the thickness sion in mice has been developed by Stathopoulos et al.
of the pleura has large influences on the formation (3) . In this model, 1 . 5 X 1 05 Lewis lung cancer cells,
and reabsorption of pleural fluid. Another important derived from a spontaneously arising lung adenocar
factor to consider is the availability of commercial cinoma in C57B/6 mice, are injected directly into the
enzyme-linked immunosorbent assays (ELISAs) for pleural space ofC57B/6 mice. This results in multiple
the proposed species. These kits are usually available discrete pleural tumors and a malignant pleural effu
only for humans and mice and one must determine sion with a mean volume of 667 ,uL by day 1 4 and a
if there is sufficient cross-reactivity with other species 1 00% mortality by day 1 7 (3) . The unique aspect of
such as rabbits, guinea pigs, and sheep ( 1 ) . this model is that the animals are immunocompetent.
A large number o f animal models have been devel This model has been used to study several aspects
oped to study diseases of the pleura. In this chapter, of malignant pleural effusion including the following.
31
32 P L E U RAL D I S EASES
(a) Nuclear factor (NF)-KB is important for the pro The level of expression of VEGF messenger ribonu
gression of the tumor. If Lewis lung cancer cells cleic acid (mRNA) and protein by the cell lines directly
expressing a dominant NF-KB inhibitor are injected, correlated with the amount of pleural effusion. When
there is decreased tumor burden and decreased pleu the PC l 4 cells were transfected with the antisense
ral effusion volume (3) . (b) Administration of a vas VEGF- 1 6 5 gene, the tumor invasion of the pleura was
cular endothelial growth factor (VEGF) inhibitor not altered, but the amount of pleural effusion was
decreases the bulk of the tumor and the amount of decreased. When the H226 cells were transfected with
pleural Buid (4) . (c) The administration of recom either sense VEGF- 1 65 or sense VEGF- 1 2 1 genes,
binant adenovirus-mediated delivery of human their direct implantation, but not their intravenous
endostatin resulted in significant reduction in pleu injection, led to localized vascular hyperpermeabil
ral effusion volume, the number of pleural tumor ity and pleural effusion. This study suggested that in
foci, microvessel density, and vascular permeability, order for a metastatic malignancy to produce a pleural
while it significantly prolonged the survival time (5) . effusion, the tumor must both invade the pleura and
(d) Host-derived IL-5 promotes experimental MPE express high levels ofVEGF ( 1 4) .
and may be involved in the pathogenesis of human These same researchers then attempted t o block
MPE (6) . (e) A sundilac derivative with antiangio the formation of malignant pleural effusion in their
genic capabilities halted experimental pleural effusion model by administering VEGF inhibitors ( 1 3) . The
formation and intrapleural tumor dissemination, inhibitor studied, PTK 787, is a specific inhibitor of
through down-regulation of pleural vascular perme VEGF receptor tyrosine kinase phosphorylation. The
ability (7) . (f) Angiopoietin/Tie2 axis blockade sig administration of PTK 787 at a dose of 1 00 mg/kg
nificantly reduced pleural Buid volume and pleural nearly eliminated the formation of pleural effusions,
tumor foci (8) . (g) Zoledronic acid, an aminobiphos but doses of 10 and 50 mg/kg had no significant
phonate, treatment resulted in significant reductions effect. PTK 787 appeared to decrease the formation
in pleural Buid accumulation and tumor dissemina of pleural Buid by reducing vascular permeability
tion, while it significantly prolonged survival (9) . rather than by inhibiting tumor growth or pleural
(h) The knockout of osteopontin in the lung cancer invasion. This experiment suggested that PTK 787
cells significantly reduces the formation of MPE, but or other inhibitors of VEGF could be useful for the
does not inhibit in vivo tumor growth of the cancer cells treatment of malignant pleural effusions.
in mice (1 O). (i) Elevated expressions of aquaporin- 1 Nagamachi et al. ( 1 5) assessed the effects of inject
and VEGF-protein are associated with increased vol ing various human non-small cell lung carcinoma
ume of malignant effusion ( 1 1 ) . Studies such as those (NSCLC) cell lines into nude mice. They injected
outlined above can provide insights into possible 2 X 1 06 cells from eight different cell lines into the
treatment for malignant effusions in humans. left pleural space. Seven of the eight cell lines were
In another model of malignant pleural effusion in tumorigenic and growth rates were relatively high in
mice, male thymic BALL/ c nude mice (Animal Produc four of the cell lines (PC 14, Lu65, Lu99A, and H l 57)
tion Area of the National Cancer Institute, Frederick ( 1 5) . Pleural effusions were not mentioned with these
Cancer Research Facility, Frederick, Maryland) are cell types. However, when mice were injected with the
used ( 1 2) . The tumor cells (1 X 1 06/300 ,uL of HBSS PC9 cells, they developed malignant pleural effusions
lung adenocarcinoma cell line PC1 4PE6 or the squa ( 1 5) . It was concluded that the PC9 cells could be
mous cell line H226) are injected into the lateral tail used for a model of malignant pleural effusion ( 1 5) .
vein of unanesthetized nude mice. The mice develop
colonies of the adenocarcinoma in the lungs as early In Rats. Ohta e t al. ( 1 6) have developed a model
as 4 weeks after tumor inoculation and all recipients of malignant pleural effusion in immune deficient
of the adenocarcinoma line, but not the squamous cell rats that also uses the PC- 1 4 adenocarcinoma cell
line, develop bloody pleural effusions ( 1 3) . line. These researchers reported that the injection of
Th e role o f VEGF i n the production o f malig 1 X 1 07 cells into the subpleural space of the parietal
nant pleural effusion has been investigated using the pleura or into the pneumonectomy space resulted
model in the above paragraph ( 1 3) . Yano et al. ( 1 4) in disseminated malignancy in 8/8 animals and the
found that the intravenous injection of PC 1 4, the development of a pleural effusion in 5/8 animals with
adenocarcinoma cell line, resulted in multiple lung a positive pleural Buid cytology in two ( 1 6) . Interest
lesions with invasion of the pleura, and produced ingly, these workers showed that the intraperitoneal
a pleural effusion containing high levels of VEGF. injection of monoclonal antihuman VEGF antibody
CHAPT E R 4 I A N I MAL M O D E LS I N PLE U RAL I NVESTIGAT I O N 33
at a dose of 250 µg twice weekly prevented the devel a pleural effusion in 9/ 1 0 mice with a mean volume
opment of the pleural effusion but did not affect the of 320 µL. If the melanoma cells were administered
dissemination of the tumor ( 1 6) . to syngeneic mice lacking the gene for nitric oxide
Ebright et al. ( 1 7) injected 1 X 1 07 A549 cells of synthase (NOS) II, only approximately 50% of the
the human lung adenocarcinoma into nude male rats animals developed pleural effusions and those that
and reported that by 5 weeks, three of nine rats had did had a lower volume (2 1 ) . lmmunohistochemical
developed pleural effusions. All the rats had tumor analyses indicated that absence of NOS II expression
nodules on the lung and pericardium ( 1 7) . In this was correlated with decreased VEGF expression and
model the intrapleural administration of NVl 020, tumor-associated vascular formation.
which is a novel, multimutated, replication-restricted
herpes simplex virus, significantly inhibited the Fibrosarcoma
growth of the tumor ( 1 7) .
Prosst et al. ( 1 8) used a similar model to assess the The intrapleural injection of fibrosarcoma can lead to
usefulness of thoracoscopic fluorescence in the diag a pleural fibrosarcoma with pleural effusion. Yasutake
nosis of pleural malignancies. These workers injected et al. (22) injected 1 X 1 05 Meth A fibrosarcoma cells
human lung cancer cells (adenocarcinoma 82/5) into intrapleurally in syngeneic BALBI c mice. All the injected
the right pleural cavity through the lateral diaphragm mice in this model developed pleural fibrosarcoma and
and found that after 5 to 7 weeks the entire pleural all developed pleural effusions with a mean volume of
cavity was affected with the malignancy ( 1 8) . They 733 µL. The mean survival of the injected mice was only
did not mention whether the animals developed pleu 8.7 days after injection. Interestingly, when heat-killed
ral effusions. They used this model to demonstrate cells of Lactobacillus casei were injected on day 3 or day
that fluorescence enhanced the diagnostic capabilities 6, the mean survival was increased beyond 40 days and
of thoracoscopy ( 1 8) . the amount of pleural fluid was only 1 4 µL (22). Intra
pleural injection of antibodies against tumor necrosis
In Rabbits. Th e intravenous Injection o f VX2 factor alpha (TNF-a) completely eliminated the anti
tumor cells in rabbits can lead to the development tumor effect of the heat-killed cells of L. casei (22) . The
of a pleural effusion ( 1 9,20) . The VX2 is rabbit pap results of this study suggest that TNF-a can act as an
illoma tumor. Tumors are created in rabbits by the antitumor drug in this model.
intravenous injection of approximately 25 ,000 VX2
cells obtained from an alive donor rabbit ( 1 9,20) .
A S B E STO S A N D M E S O T H E L I O M A
When the rabbits are sacrificed 28 days after the
injection, approximately 1 5% have a pleural effusion There have been many papers published o n the rela
(20) . The problem with this model is that the pro tionship between asbestos and mesothelioma. Wagner
duction of each pleural effusion is expensive because (23) first reported in 1 962 that in rats the intrapleural
pleural effusions develop in only 1 5% of the animals. administration of chrysotile and crocidolite led to the
We have tried to create malignant effusions in rabbits production of mesothelioma. Then in 1 965, Smith
by injecting the VX2 cells directly into the pleural et al. (24) demonstrated that the intrapleural injec
space but were unsuccessful. tion of amosite in hamsters led to the development of
mesothelioma. Subsequently, there have been many
papers written on the induction of mesotheliomas
Melanoma
in animals. Asbestos has been administered intra
Melanoma metastatic to the lung can be produced by pleurally, intratracheally, intraperitoneally, and by
injecting murine melanoma cell lines into the lateral inhalation. In the following section, the results will
tail vein of syngeneic mice (2 1 ) . In this model, syn be briefly described when asbestos is administered
geneic mice receive 5 X 1 0 4 B l 6-BL6 or B l 6-F l 0 through these different routes.
melanoma cells in a total volume of 0.2 mL. At sacri
fice 2 1 days after the cells were injected, Wang et al.
lntrapleural Injections
(2 1 ) reported that all 20 mice injected with the
two different cell lines developed lung metastases. The intrapleural injection of asbestos leads to the
The injection of B l 6-BL6 cells resulted in a pleural development of mesothelioma of the pleural space.
effusion in 8/ 1 0 with a median volume of 250 µL Most studies have been performed in rats (23, 25-30) ,
whereas the injection of the B 1 6-F l O cells resulted in although there has been at least one report in hamsters
34 P L E U RAL D I S EASES
(24) . When asbestos fibers are injected intrapleurally, effects of inhaled asbestos on rats (35-4 1 ) , and one
approximately 30% to 60% of the animals develop on baboons (42) . Wagner et al. (40) exposed rats to
a pleural mesothelioma (26,30) . The tumors develop asbestos clouds 7 hours/day, 5 days/week for 1 year.
approximately 1 8 months after the intrapleural injec Even when asbestos is administered with such inten
tion (26) . There is a dose response, with larger doses sive regimens, the incidence of mesothelioma is less
being associated with higher incidence of tumor (30) . than it is after one intrapleural (4 1 ) or one intraperi
When different types of asbestos are compared, the toneal injection (38) of asbestos. When rats are given
intrapleural injection of 25 mg chrysotile or crocido inhaled asbestos for 1 year, less than 1 0% develop
lite produce mesotheliomas in approximately 60% mesothelioma, but approximately 30% develop lung
whereas the same dose of amosite produces tumors in tumors (40) . Multivariate fibers analysis of multiple
35% (3 1 ) . The main difference in these types of asbes inhalational experiments has shown that the mea
tos is that mesotheliomas occur approximately 200 sure most highly correlated with tumor incidence
days later for amosite than for the other two types (3 1 ) . is the concentration of structures 20 µm or more in
Th e propensity o f a fiber type t o produce a meso length (35 ) . Potency appears to increase with increas
thelioma is dependent to a large extent on its geomet ing length, with fibers longer than 40 µm being
ric configuration. Long, thin fibers appear particularly approximately 500 times more potent than fibers
likely to produce mesothelioma in man (27) . The between 5 and 40 µm (35). When different types of
chemical composition of the asbestos fiber also influ asbestos are compared, chrysotile is less potent than
ences whether its intrapleural injection will induce amphiboles in inducing mesothelioma, but compa
mesothelioma. If more than 80% of the magnesium rable in producing lung tumors (35). The most effi
is leached from chrysotile fibers, the proportion of cient agent at inducing mesothelioma after inhalation
animals developing mesotheliomas is dramatically is erionite (32) . Wagner et al. (32) demonstrated that
lower than with untreated chrysotile (28) . The rate the inhalation of Oregon erionite induced mesothe
of pleural mesothelioma will be more than doubled liomas in 27 of 28 (96%) rats and decreased their
if pleural inflammation is induced by the intrapleural mean survival time from 738 days to 504 days.
injection of carrageenan several months after the ini It has also been demonstrated that inhaled asbestos
tial injection of asbestos (3 1 ) . can induce mesotheliomas in primates. Webster et al.
O f all fibers tested, the most potent inducer of (42) exposed baboons 6 hours daily for 5 days a week
pleural mesothelioma is erionite (25,32) . Wagner et al. to amosite asbestos, except for 3 weeks of the year
(32) injected 20 mg of Oregon erionite, Karain Rock when the chamber had to be serviced. The animals
fiber, chrysotile, or nonfibrous zeolite intrapleurally were exposed for up to 900 days. Five of 1 0 animals
to rats, and the incidence of mesothelioma was as developed malignant mesothelioma between 6 and
follows: erionite 40/40 ( 1 00%), Karain Rock fiber 1 0 years after the initial exposure, and the remaining
38/40 (95%), chrysotile 1 9/40 (48%) , and nonfi animals all developed asbestosis (42) .
brous zeolite 2/40 (5%) (32) . Erionite in the air has
been implicated in the epidemic of mesotheliomas in lntratracheal I njections
Turkey (33) .
If normal rat mesothelial cells are treated in vitro There have also been a few reports on the effects of
with chrysotile fibers, mesotheliomas will develop asbestos administered intratracheally in hamsters
rapidly when these cells are injected subcutaneously (43,44) , dogs (45), and monkeys (45). It is less burden
into nude mice. Fleury-Feith et al. (34) demonstrated some to administer the asbestos intratracheally than
that when untreated cells were injected, tumors devel to do inhalational treatments nearly daily for a year
oped in 3/5 animals at a median of 22 weeks after or more. However, in general it is difficult to induce
injection. However, if the cells were treated repeat mesotheliomas with intratracheal asbestos. Pylev (43)
edly with chrysotile, tumors developed in 5/5 animals administered 10 mg of chrysotile twice a month to
at a median of only 1 week after injection (34) . hamsters and reported that only 1 of 4 1 animals devel
oped a malignant mesothelioma, although 25% of
the animals developed a malignant lung tumor. The
Inhaled Asbestos
same researcher also tried to induce tumors in three
It is more difficult in general to produce tumors by monkeys with the intratracheal administration of
the inhalation of asbestos than by intrapleural admin 400 to 600 mg asbestos and reported that no tumors
istration. There have been numerous reports of the developed during 1 7 to 22 months of follow-up (43).
CHAPT E R 4 I A N I MAL M O D E LS I N PLE U RAL I NVESTIGAT I O N 35
dependent upon the organisms injected. None of empyema. One question that was investigated was
the animals injected with B. .fragilis developed empy whether therapeutic thoracentesis was a reasonable
ema, whereas empyema developed in 33% of those alternative to tube thoracostomy in the management
injected with B. .fragilis plus S. aureus and 3 1 % of of rabbits with empyema (64) . Rabbits were random
those injected with E. coli plus B .fragilis (62) . These
. ized to receive daily therapeutic thoracentesis start
investigators also used this model to demonstrate that ing at 48 hours, chest tubes at 48 hours, or neither
the more organisms injected into the pleural space, therapeutic thoracentesis nor tube thoracostomy
the more likely the animal was to develop an empy (controls) . The animals in the chest tube group had
ema. For example, if 1 X 1 08 E. coli plus B. .fragilis their chest tubes attached to a Heimlich valve and
were injected, 1 0/ 1 0 animals developed an empyema, had their chest tubes aspirated at 1 2-hour intervals.
whereas if 1 X 1 06 of these organisms were injected, In this study, the mortality rate in the therapeutic
only 5 / 1 0 animals developed an empyema (5 5). If thoracentesis group (01 1 6) was significantly less (p =
blood were injected in conjunction with the bacte 0.02) than the mortality rate in the other two groups
ria, the likelihood of developing an empyema was not combined (9/33) . When the surviving animals were
increased. Lastly, guinea pigs that developed pneu sacrificed at 1 0 days, the gross empyema score in
monia were more likely to have an empyema. the therapeutic thoracentesis group was significantly
The primary problem with this model is that it lower (p < 0.05) than that in the chest tube group
requires a foreign body in the pleural space that does or the control group (64) . This study suggests that
not mimic the clinical situation in humans. therapeutic thoracentesis possibly has a role in the
Rabbits with Bacteria in Agar. The fourth model management of patients with empyema.
was developed in the late 1 990s by Sasse et al. (63). In A second question addressed was whether the tim
this rabbit model, 1 X 1 09 Pasteurel/,a multocida cul ing of the chest tube insertion is important in the
tured in agar (rather than broth) were injected into the treatment of empyema. Rabbits were randomized to
pleural space of rabbits. Agar, rather than broth, was receive no chest tube or a chest tube 24, 48, or 72
used so that the mixture would remain in the pleural hours after receiving the P. multocida intrapleurally
space longer. Procaine penicillin G was administered (Fig. 4. 1 ) (65). The rabbits that received the chest
once per day starting 24 hours after the initial injection tube at 24 or 48 hours had significantly better results
to prevent death from sepsis. With this model, the rab
bits developed an empyema; 24 hours postinjection, the
mean pleural fluid pH was 7.0 1 , the mean glucose was
p < 0 . 05
1 0 mg/ dL, the mean lactic acid dehydrogenase (LD H) p < 0 . 05
was 70 times higher than the upper normal limit for
serum, and the Gram's stain and culture of the pleural 4 15
fluid were positive (63) . At 96 hours postinjection, the
Gram's stain and culture of the pleural fluid were usually � 14
8 3 n 11
(/)
=
antibiotics alone. The pleural fluid becomes infected, pus, 2 =m i n i m a l p l e u ra l peel, 1 adhesions
=
then loculated, and a rind is formed over the visceral betwee n the viscera l a n d parieta l p l e u ra, a n d
pleura. The loculation in this model is less than that 0 = n o r m a l p l e u r a l space. (Reprinted with permission
from Sasse S, Nguyen T, Teixeira LR, et al. The utility of daily
seen in human empyema.
therapeutic thoracentesis for the treatment of early empyema.
This model has been used in an attempt to answer C hest. 1 999; 1 1 6: 1 703- 1 708.)
several questions concerning the management of
38 P L E U RAL D I S EASES
than did the rabbits that received the chest tube at 72 recently evaluated the usefulness of tissue plasmino
hours or those that did not receive a chest tube at all gen activator (tPA) , or human recombinant DNase
(65) . This study demonstrates that a relatively short (rhDNase) , or their combination in facilitating the
delay in initiating tube thoracostomy adversely affects drainage in this rabbit model of empyema (69) . The
the outcome in these animals with an empyema. combination of tPA (4 mg) and rhDNase (1 mg)
The third question addressed with this model was administered intrapleurally twice a day for a total of six
whether all antibiotics penetrate empyemic pleural doses led to significantly lower empyema scores when
fluid similarly (66) . Twenty-four hours after the intra the rabbits were sacrificed on day 1 0 (69) . Neither
pleural injection of P. multocida, Teixeira et al. (66) agent by itself significantly reduced the empyema
injected either penicillin 24,000 units/kg, clindamy score (69) . However, the mean amount of fluid aspi
cin 9 mg/kg, gentamicin 1 mg/kg, metronidazole rated from the rabbits that received tPA either alone
37 mg/kg, vancomycin 1 5 mg/kg, or ceftriaxone or in combination with rhDNase ( 1 86 mL) was
30 mg/kg intravenously. Antibiotic levels in samples much greater than that in the rabbits that received
of pleural fluid and serum, collected serially for up to rhDNase alone (0.7 mL) or saline (5 .8 mL) (69) .
480 minutes, were determined using a bioassay. In These results demonstrate that the administration of
this study, the degree to which antibiotics penetrated a combination of a fibrinolytic agent and a DNAase
the pleural fluid was highly variable (Figure 1 2.2) may be more effective than the administration of
(66) . Metronidazole penetrated most easily, followed either agent by itself. These results also demonstrate
by penicillin, clindamycin, vancomycin, ceftriaxone, that the administration of tPA leads to the produc
and gentamicin. This variance in the penetration of tion of large amounts of pleural fluid, but does not
antibiotics into empyemic pleural fluid should be improve the empyema score. A possible explanation
taken into consideration when antibiotic therapy is for the lack of benefit in this model with tPA alone is
chosen in patients with empyema. the observation that at autopsy, the animals have very
Sasse et al. (67) investigated the role of transform few adhesions between the visceral and parietal pleura
ing growth factor beta 1 (TGF-/3 ) in the production and the pleural fluid is not loculated in any group.
of pleural fibrosis that sometimes occurs in this exper Our group has also shown that in this model the intra
imental model. These investigators reported that the pleural administration of 1 ,000 IU heparin alone or
levels of TGF-/3 1 closely correlated with microscopic in combination with 1 mg of rhDNase every 1 2 hours
pleural thickness (r = 0.7, p < 0.00 1 ) and number for 3 days is no more effective than saline in the treat
of fibroblasts present in the visceral pleura (r 0.68,
= ment of empyema in rabbits (70) .
p < 0.00 1 ) . The level of TGF-/3 1 increased during Na et al. (7 1 ) have subsequently shown that the
the 8 days of observation (67) . This study suggests excess fluid induced by the tPA is very inflammatory.
that inhibition ofTGF-/3 1 in patients with empyema The pleural fluid WBC count is above 1 00,000, the pH
might decrease the amount of residual fibrosis. These is below 7. 1 0, and the LDH is more than 20 times the
workers subsequently administered monoclonal upper limit of normal. The characteristics of the fluid
antibodies to TGF-/3 1 to determine if the antibodies change little with time. This study (7 1 ) suggests that
would decrease the degree of fibrosis associated with the intrapleural administration of tPA creates a large
the empyema (68) . When the animals were sacri inflammatory process in the pleural space associated
ficed on day 6, immunohistochemistry revealed that with the production of large amounts of pleural fluid.
the TGF-/3 1 was localized to the macrophages in the
exudative material and the visceral pleura. The ani Rats with Bacteria in Agar or Broth. Empy
mals that received antibody to TGF-/3 1 had markedly ema can also be induced in rats by the intrapleural
decreased amounts of exudative material in the pleu instillation of bacteria in agar (72) or broth (73). An
ral space relative to control animals. All markers of empyema was induced in 25 rats by the intrapleural
empyema and pleural fibrosis were also significantly instillation of 1 X 1 0 1 0 S. aureus in agar and 5 died of
decreased in the rabbits receiving intrapleural anti sepsis within the first 24 hours. Nine rats were sacri
TGF-/3 1 (68) . ficed on the third day and all had pleural fluid (vol
The technique reported by Sasse was modified ume 0 . 5 to 3.8 mL) and bacteria in the pleural fluid.
so that chest tubes were implanted and drawn out Eight rats were sacrificed on the fifth day and only
between the scapulae to facilitate the intrapleural three had pleural fluid (72) . The amount of pus with
administration of therapeutic agents and the aspiration this rat model appears to be less than when rabbits are
of pleural fluid (69) . With this modification we have injected with P. multocida in agar. Luo and associates
CHAPT E R 4 I A N I MAL M O D E LS I N PLE U RAL I NVESTIGAT I O N 39
evaluated the effect of LP l 7, a synthetic inhibitor This hypersensitivity and the capacity to form pleural
of triggering receptor expressed on myeloid cells- 1 effusions can be transferred passively with cells from
(TREM- 1 ) on empyemas induced by the intrapleural a sensitized animal (75). This experiment was the first
injection of Pseudomonas aeroginosa or Staphylococcus conclusive evidence that the pleural effusion with
aureus (73) . They reported that there was a marked tuberculosis could be due to hypersensitivity.
reduction in neutrophil numbers in the LP l 7 treated Leibowitz et al. (76) subsequently used this same
rats due to the reduction of both pleural effusion vol model to demonstrate that neutrophils were the
ume and total cell numbers. The LP l 7 treated rats predominant cells in the early stages of the effusion
also had a significantly higher survival rate. (<6 hours) , whereas at 24 hours, lymphocytes and
macrophages were the predominant cells. They also
Mice with Pneumonia Induced by Intranasal demonstrated that antilymphocyte serum (ALS)
Inoculation of Bacteria. None of the above mod could completely suppress the reaction (76) . More
els mimic parapneumonic effusions because none of over, after treatment with ALS, the ability to produce
the animals have pneumonia as the primary cause of a pleural effusion with the intrapleural injection of
the effusion. Wilkosz et al (74) developed a model PPD paralleled the return of the positive skin reac
of empyema in mice via the intranasal inoculation tion. These studies provided additional evidence that
of 0.5 to 1 .0 X 1 07 Streptococcus pneumonia. The delayed hypersensitivity is responsible for the effusion.
animals developed pneumonia. By 48 hours, all the
mice had developed macroscopic pus in their pleu BCG Model. Widstrom and Nilson (77) first
ral spaces. By 4 hours after inoculation, bacteria were described this model in which guinea pigs are immu
detected in the pleural space of 33% of animals and nized by the intracutaneous injection of 0.4 mg BCG
by 48 hours all animals had large numbers of bacteria into the left thigh that causes a small necrotizing skin
in their pleural spaces. reaction. Within 3 weeks of the injection, the ani
mals have a positive skin test to PPD. Then if 2 mg
of BCG is injected into the pleural space, a pleural
T U B E R C U LO S I S
effusion develops. The effusion is larger (mean 8 mL)
Three types o f animal models have been developed that than with the hypersensitivity model, reaches its
relate to tuberculosis and pleural disease. In the hyper maximal size at approximately 1 4 days and has, for
sensitivity model, an animal is sensitized to tuberculous the most part, disappeared by 21 days (77) . Again
protein. Subsequently when tuberculous protein is in this model neutrophils were the predominant
injected into the pleural space, a pleural effusion devel cells in the first 24 hours, but lymphocytes were the
ops. In the Bacillus Calmette-Guerinrin (BCG) model, predominant cell after 5 days (77) . When the lym
BCG is injected directly into the pleural space and a phocytes first appear in the pleural fluid on approx
pleural effusion develops. In the direct inoculation imately day 3, they do not respond to PPD . From
model, M. tuberculosis is injected directly into the pleu day 5 onward, however, reactivity to PPD is found
ral space in either sensitized or nonsensitized animals. in most cases (78) . The reactivity of the lymphocytes
in the peripheral blood parallels that of the pleural
Hypersensitivity Model. With this model, which lymphocytes (78) . A similar model in which guinea
was first developed by Allen and Apicella (75) , guinea pigs were immunized with a low-virulent strain of
pigs are immunized with a footpad injection of0.2 mL Mycobacterium tuberculosis was described by Paterson
of complete Freund's adjuvant containing 1 mg killed in 1 9 1 7 (79) .
tubercle bacilli/ml, emulsified with an equal volume A rabbit model using BCG has been described by
of isotonic saline. Three to 5 weeks after immuniza Antony et al. (80) . The rabbits are sensitized by the
tion, tuberculin purified protein derivative (PPD) in injection of BCG ( 1 X 1 06 organisms) intradermally
a total volume of 0.5 mL is injected intrapleurally. on the leg. Three weeks later the rabbits have a posi
A pleural effusion then develops with a maximal tive skin test to PPD. Then BCG (4 X 1 06 organisms)
volume of 4 mL at 24 hours and a protein concen is instilled into the pleural space (80) . This results in a
tration of 3.8 g/dL. In nonimmunized animals, the pleural effusion with a maximal volume of 5 . 9 :±: 0.6
intrapleural injection of PPD leads to effusions with mL at 24 hours that decreases to 1 . 8 :±: 0.4 mL by
much lower volumes. There is no clear relationship 1 20 hours (80) . Neutrophils are the predominant
between the amount of PPD injected and the size of cells in the first 24 hours, whereas macrophages are
the effusion once a threshold amount of PPD is given. the predominant cells after 48 hours (80) . If the
40 P L E U RAL D I S EASES
animals are made neutropenic, then the accumulation is injected into the pleural space of rats, the initial
of pleural fluid and inflammatory cells is decreased inflammatory response is in the immediate subpleu
(80) . The intrapleural injection of neutrophils in the ral tissue that contains the blood vessels (9 1 ) . Then
neutropenic animals restores the response to control there is a rapid release ofWBCs into the pleural space
levels. The neutrophils in the pleural space appear to along with the development of an exudative pleural
secrete a monocyte chemotaxin that recruits mono effusion (9 1 ) . The peak amount of exudative pleural
cytes to the pleural space and thereby contributes to fluid ( 1 .75 mL) and the peak number ofWBCs occur
the formation of granulomas (80) . approximately 1 6 hours after injection (9 1 ) . Initially,
most of the WBCs are neutrophils. The number of
Direct Inoculation Model. If nonsensitized guinea mononuclear cells peaks at 24 hours when the num
pigs are given M. tuberculosis intrapleurally, they die of bers of mononuclear cells and neutrophils are compa
generalized tuberculosis within 4 to 6 weeks (79). Ly rable. By 72 hours postinjection, mononuclear cells
and associated studied the role ofTNF alpha in guinea account for 90% of the cells. By 96 hours, negligible
pig tuberculous pleuritis in guinea pigs sensitized to numbers of neutrophils and pleural fluid remain
BCG (8 1 ) . In this study, M tuberculosis was directly (92) . Pleural inflammation has also been studied in
injected into the pleural space. They reported that the mice after the intrapleural administration of carra
neutralization of TNF alpha hastened the transition geenan. They develop inflammatory exudates 4 hours
to an anti-inflammatory cytokine response in guinea postinjection (93) .
pig pleural granulomas and exudate cells (8 1 ) . Du and This model has been used to investigate the influ
coworkers injected M. tuberculosis intrapleurally to rats ence of many compounds on this inflammatory reac
sensitized to BCG (82) . They reported that all injected tion. The following are a few examples of the many
rats developed bilateral pleural effusions within 5 days. investigations carried out. The early ( <48 hours)
The pleural fluid had predominantly neutrophils at day inflammatory response is depressed by cyclooxygen
1 but lymphocytes by day 5. When an adenovirus car ase inhibitors (aspirin and indomethacin) (9 1 ) , but
rying the vector for aquaporin- 1 was injected intrapleu the response at 72 hours is influenced much less (92) .
rally 7 days prior to the injection ofM. tuberculosis, the Corticosteroids decrease the inflammatory response
volume of pleural fluid was significantly increased (82) . throughout the entire period (9 1 ) . The selective
cyclooxygenase inhibitors celecoxib and rofecoxib
P L E U RA L I N F LA M M AT I O N
inhibit the inflammatory process less than aspirin or
Inflammation o f the pleural space has been studied indomethacin (94) .
by means of the intrapleural injection of several dif There have also been studies at a more basic level.
ferent inflammation-producing agents. Advantages D'Acquisto et al. (95) demonstrated that there was
of the pleural space for the study of inflammation in NF-KB deoxyribonucleic acid (DNA) binding activ
comparison to the paw edema and cutaneous mod ity in the inflammatory cells that migrated into the
els include the following (a) inflammatory exudates pleural cavity at 3 and 6 hours. This activity was
can be harvested relatively simply from the pleural markedly increased at 24 hours and then decreased
cavity and (b) inflammatory irritants can be directly at 48 hours (95). Frode et al. (96) demonstrated that
injected into the pleural cavity (83) . Although pleural the intraperitoneal administration of NF-KB inhibi
inflammation following the injection of carrageenan tors before the intrapleural injection of carrageenan
has been studied most intensively, the inflammatory inhibited the inflammatory response. Sautebin et al.
reaction following the injection of other agents such (97) demonstrated that there was a close correlation
as zymosan (84,85); endotoxin, or lipopolysaccharide between the levels of nitric oxide and prostaglandin
(LPS) (86,87) ; monosodium urate (89) ; and micon £2 4 hours after the intrapleural injection of carra
azole (90) , and following antigen-antibody reactions geenan (97) . Scavengers of nitric oxide such as hemo
(88) has also been studied. globin reduced the inflammation and the amount of
prostaglandin E2 in the pleural fluid (97) .
The intrapleural injection of IL-6 5 minutes
Carrageenan
before the injection of carrageenan reduced in a dose
Carrageenan-induced pleurisy in rats or mice has dependent and significant manner the exudation and
been widely used for many years for the study of total and differential leukocyte migration in both the
inflammation in general and pleural inflammation early and late phases of the inflammatory response
in particular. When 1 .2 mL of 0.25% carrageenan (98) . The intrapleural injection of IL- 1 0 resulted in
CHAPT E R 4 I A N I MAL M O D E LS I N PLE U RAL I NVESTIGAT I O N 41
a significant inhibition of the early but not the late permeability) accumulates in the pleural cavity than
response (98) . The administration of anti-IL-6 anti when carrageenan is administered ( 1 03) . Kikuchi
bodies caused a significant decrease in both total et al. ( 1 02) have demonstrated that 5-lipoxygen
and differential leukocyte influx, but significantly ase inhibitors, but not cyclooxygenase inhibitors
increased the exudation at 4 hours. The adminis (e.g., indomethacin) , inhibit the infiltration of leuko
tration of anti-IL- 1 0 antibodies caused graded and cytes into the pleural fluid 3 hours after the intrapleu
marked increases of both total and differential leuko ral injection of zymosan in rats. However, another
cyte influx and fluid exudation (98) . study in mice ( 1 04) demonstrated that the adminis
This model has also been used in experiments tration of indomethacin reduced exudate formation
with knockout mice (mice deficient in certain genes) . to almost the extent that it was reduced in knockout
When IL-6 knockout mice are injected with car mice for prostaglandin I 2 • The degree of exudation
rageenan, both the amount of pleural fluid and the was similar in the wild type mice and in the knockout
pleural fluid WBCs fall by 50%, as compared with mice for the prostaglandin E 2 •
wild-type mice (99) . The same reduction in the
inflammatory response is obtained if the mice are
Lipopolysaccharide or Endotoxin
given anti-IL-6 before the carrageenan treatment
(99) . There is a marked reduction in the inflamma The injection of endotoxin into the pleural space of
tory response to carrageenan intrapleurally in mice mice (88), rats ( 1 05), guinea pigs ( 1 05), and rabbits
deficient for the inducible NOS gene, as compared (87, 1 06, 1 07) leads to the accumulation of neutro
with mice that are not deficient in this gene (93) . phils in the pleural space at 4 hours and the accu
Yuhki et al. ( 1 00) induced pleurisy with intra mulation of eosinophils and mononuclear cells at
pleural carrageenan in knockout mice for the pros 24 hours, which persist for at least 96 hours ( 1 05).
taglandin receptors (IP, EP l , EP2, EP3, or EP4) and This model has been used to study pleural fluid eosin
demonstrated that the pleural exudation at 1 to 5 ophilia ( 1 05).
hours was significantly reduced in the IP, EP2, and Although the intrapleural injection of platelet
EP3 knockout mice, but not in the EP l or EP4 activating factor (PAF) ( 1 08), leukotriene B 4 ( 1 08) or
knockout mice. Leukocyte migration into the pleu bradykinin ( 1 08) all induce delayed and long-lasting
ral cavity was not influenced by any of the receptor eosinophil infiltration in the rat pleural cavity, none
deficiencies ( 1 00) . This study demonstrated that fluid of these appears to be involved in the eosinophil accu
exudation and leukocyte migration are not coupled. mulation resulting from endotoxin. Treatment with
Knockout mice for 5-lipoxygenase have a signifi inhibitors of any of these three cytokines does not
cantly reduced inflammatory reaction to the intra prevent the eosinophil accumulation after endotoxin
pleural injection of carrageenan ( 1 0 1 ) . intrapleurally ( 1 05). In contrast, dexamethasone
and the protein synthesis inhibitor, cycloheximide,
abolish endotoxin-induced eosinophil accumulation
Zymosan
( 1 09) . It has also been shown that the blockade of
The intrapleural injection of zymosan in rats also pro nitric oxide biosynthesis prevents endotoxin-induced
duces an inflammatory exudate. When 0 . 1 mL of2% eosinophil accumulation ( 1 1 0) . Inhibitors of IL-5
zymosan is injected into the pleural space, the reac do not prevent eosinophil accumulation after endo
tion is similar to that induced by carrageenan, with toxin is given intrapleurally ( 1 05), but they do pre
an early neutrophil influx and the maximal accu vent eosinophil accumulation after antigen-induced
mulation of pleural fluid at 24 hours. The maximal inflammation ( 1 1 1 ) .
amount of fluid following zymosan and carrageenan I n rabbits, the intrapleural injection o f endo
in the rat are comparable (� 1 . 5 mL) (85,9 1 , 1 02) . toxin leads to neutrophil influx into the pleural space
The levels of TNF-a, IL- 1 , IL-6, and cytokine that peaks at 6 hours (87, 1 06) . There is also a bipha
induced neutrophil chemoattractant (CINC) in the sic increase in the vascular permeability with the first
pleural fluid all begin to increase after 1 to 2 hours, increase 1 5 minutes after the injection and the second
preceding the influx of neutrophils, and peak after 4 increase 2 hours after the injection (87, 1 06) . Broaddus
to 5 hours (85). The inflammatory response to carra et al. ( 1 07) have demonstrated that the neutrophil
geenan and zymosan are very similar. However, after influx is profoundly inhibited by neutralization ofIL-8.
the intrapleural administration of zymosan in the Edamitsu et al. (87) have demonstrated that the early
mouse, more Evans blue dye (a marker of vascular neutrophil influx is partly inhibited by anti-TNF-a
42 P L E U RAL D I S EASES
but not IL- 1 receptor antagonists. In contrast, the receptor B, inhibits the antigen-induced eosinophil
late phase is inhibited by both anti-TNF-a and IL- I and mononuclear cell migration ( 1 1 3) . The adminis
receptor antagonists (87) . These latter investigators tration of a bradykinin receptor antagonist blocks the
also demonstrated that the immediate increase in per entire response in a dose-dependent manner ( 1 1 4) .
meability was inhibited by antihistamines but not by This model has also been used to show that lipoxin
anti-TNF-a or IL- I receptor antagonists (87) . The A4 and aspirin-triggered 1 5-epi-LXA dramatically
late increase in vascular permeability was completely block the influx of eosinophils into the pleural fluid,
inhibited by either the depletion of neutrophils or by while concurrently inhibiting the earlier edema and
anti-TNF-a but not by IL- 1 receptor antagonists (87) . neutrophil influx ( 1 1 5) . The intrapleural injection of
Fukumoto et al. ( 1 06) confirmed these results and also CCL22 by itself induces a dose- and time-dependent
demonstrated that the delayed increase in vascular per recruitment of eosinophils into the pleural cavity of
meability was due to TNF-a-induced increases in IL-8. mice ( 1 1 6) . However, the administration of anti
CCL22 polyclonal antibodies during sensitization or
before challenge had no significant effect on eosino
Miconazole
phil recruitment (1 I 6) .
There has been one study that reported that the intra A model o f allergic pleurisy has also been developed
pleural injection of I , 2, or 4 mg of the synthetic anti in mice. In this model, mice are injected subcutane
fungal imidazole, miconazole, resulted in an exudative ously on days 1 and 8 with 1 00 µg of ovalbumin and
pleural effusion in rats (90) . The peak amount of fluid 70 µg of aluminum hydroxide. Then 0. 1 to 1 0 µg
(�3 mL) occurs at 9 hours and is dose dependent. The of ovalbumin are injected into the pleural space 7 to
peak cellular influx is at 7 hours and tends to persist 1 0 days after the last subcutaneous injection. In this
for at least 24 hours. At 7 hours, most of the cells are model, there is a dose-dependent recruitment of eosin
neutrophils, whereas at 24 hours there are approxi ophils after 48 hours ( 1 1 7) . As in the rat model, leu
mately 40% neutrophils, 40% mononuclear cells, kocytes also accumulate early in the pleural cavity. This
and 20% eosinophils. Dexamethasone, and to a lesser model has been used to show that both stem cell factor
degree phenylbutazone, but not chlorpheniramine or (SCF) ( 1 1 7) and eotaxin ( 1 1 8) play a major role in the
methysergide, markedly attenuate the fluid formation recruitment of eosinophils in allergic pleurisy. In addi
and leukocyte accumulation. tion, this model has been used to show that the admin
istration of IL-5 monoclonal antibodies prevents the
accumulation of eosinophils in the pleural space ( 1 1 9) .
H Y P E R S E N S I T I V ITY R EACTI O N S
hydrochloric acid (0.01 N), quinacrine (10 mg/kg), and doxycycline preparations that are passed through
nitrogen mustard (0.2 mg/kg), bleomycin ( 1 . 5 mg/kg), a sterile membrane filter are as effective as the paren
and NaOH (5%) each in a total volume of2 mL. These teral preparations in producing pleurodesis in rabbits.
animals did not have chest tubes but did undergo five Ors Kaya et al. ( 1 25) have shown that administra
diagnostic thoracenteses over the first 6 days so the tion of systemic diclofenac sodium, an antiinflamma
pleural fluid could be sampled. The intrapleural injec tory agent, decreases the effectiveness of tetracycline
tion of any of these agents led to an exudative pleural 35 mg/kg in producing pleurodesis. Zhu et al. ( 1 26)
effusion with very similar characteristics. However, have shown that the pleurodesis score at autopsy cor
when the animals were sacrificed at 30 days, the pleural relates very well with the gliding sign on ultrasound.
space was essentially normal in all the animals except They concluded that ultrasound was the best way to
those that received 35 mg/kg tetracycline ( 1 20). Pleu noninvasively assess pleurodesis ( 1 26) .
ral symphysis involving more than 75% of the pleural Komissarov et al. ( 1 27) showed that the locula
space occurred in 9 of the 1 0 animals that received this tions that occur after the intrapleural administra
dose of tetracycline. On the basis of this study, tetracy tion of tetracycline are better broken down with the
cline became the most popular agent for pleurodesis in proenzyme single chain urokinase plasminogen acti
the 1 9 80s. vator (scuPA) than with urokinase or alteplase. The
In the early 1 990s, the company that manufac profibrinolytic activity in the pleural fluid is greater
tured parenteral tetracycline in the United States after scuPA administration than after the other two
ceased to produce the product because of more agents ( 1 27) . Lastly, these researchers concluded
stringent manufacturing requirements. Accordingly, that scuPA promotes durable intrapleural fibri
animal studies were performed to evaluate possible nolysis via formation of alphaM/uPA complexes.
alternative agents. Light et al. ( 1 2 1 ) demonstrated These complexes promote uPA-mediated plasmino
that minocycline at doses of 7 mg/kg and doxy gen activation in scuPA-treated rabbits with TCN
cycline at doses of 20 mg/kg and above produced induced pleural injury ( 1 27) .
pleurodesis comparable to tetracycline 35 mg/ There have been multiple studies evaluating the
kg. These investigators did not aspirate the pleural effectiveness of talc for producing a pleurodesis in
fluid. They noted that hemothorax occurred with rabbits. The intrapleural administration of talc slurry
doses that produced effective pleurodesis ( 1 2 1 ) . produces a pleurodesis in rabbits, but the dose neces
Subsequently, Wu e t al. ( 1 22) demonstrated that if sary to produce a satisfactory pleurodesis in rabbits
small chest tubes were implanted in the rabbits at (�400 mg/kg) is much larger than the dose that is
the time that they received the minocycline intra normally recommended for humans (� 1 00 mg/kg)
pleurally and if the chest tubes were aspirated on a ( 1 28). The pleurodesis that results from talc develops
daily basis, the hemothoraces did not develop and slowly and the pleurodesis at 28 days is much better
the excess mortality was prevented. In their initial than the pleurodesis at 14 days ( 1 29). The pleurodesis
publication concerning chest tubes, the chest tubes that results from this large dose of talc is less complete
were kept in place with thoracic vests. Later inves than the pleurodesis that results from tetracycline
tigators demonstrated that a better way to keep the derivatives ( 1 28) or silver nitrate ( 1 30) . The addition
chest tubes in place was to tunnel the tubing under of thymol iodide to the talc does not lead to a bet
the skin and draw the proximal end out through ter pleurodesis ( 1 3 1 ) . If the rabbits are given systemic
the skin posteriorly and superiorly between the two corticosteroids concomitantly, the effectiveness of the
scapulae ( 1 23). The exterior end of the catheter is pleurodesis is decreased ( 1 32) .
plugged with a stub adaptor fitted with a Luer lock. When humans are given talc intrapleurally in
Air or liquid can be aspirated through a self-sealing an attempt to create a pleurodesis, a small percent
injection site fitting with a Luer lock ( 1 23). The site age develop the acute respiratory distress syndrome
of the exit wound posteriorly between the scapulae ( 1 33) . There have been several studies in rabbits that
is very important. If the rabbit can reach up to this have investigated the pathogenesis of this syndrome.
site, it will destroy the chest tube and the animal will Ferrer et al. ( 1 34) investigated the influence of talc
end up with a pneumothorax. particle size on extra pleural talc dissemination after
There have been several studies that attempted to talc slurry. They reported that although normal talc
answer simple questions concerning pleurodesis with and large talc were both effective in producing a
the tetracycline derivatives. Bilaceroglu et al. ( 1 24) pleurodesis, the intrapleural injection of normal talc
have demonstrated that solutions of oral tetracycline elicited a greater pulmonary and systemic talc particle
44 P L E U RAL D I S EASES
deposition than did the large size talc ( 1 34) . A sec mustard (0.8 mg/kg) ( 1 45). Mitoxantrone is also
ond study by the same group attempted to determine effective in producing a pleurodesis, but at the doses
if the dose of the talc had any effect on its systemic necessary to produce a pleurodesis, cardiac toxicity
distribution ( 1 35). They found that pleurodesis with develops and the animals develop congestive heart
the high talc dose (200 mg/kg) was associated with failure ( 1 46, 1 47) . Interestingly, the intrapleural injec
an increased risk of extra pleural talc deposition tion of mitoxantrone leads to much more pleural
compared with the low talc dose (50 mg/kg) ( 1 35). inflammation than does the intrapleural injection of
Kennedy et al. ( 1 36) demonstrated talc in the medi the tetracycline derivatives or talc ( 1 46) . The intra
astinal lymph nodes, kidney, and spleen in some of pleural injection of dacarbazine or cytarabine does
the animals that received 70 mg/kg talc intrapleurally. not produce a pleurodesis ( 1 45).
Marchi et al. ( 1 37) have shown that the intrapleural The mechanism of pleurodesis is not definitely
injection of 400 mg/kg talc produces a significant known and it may vary from agent to agent ( 1 48) . It
increase in the blood WBC count and percentage has been thought that the initial event in the produc
of neutrophils and an increase in the blood VEGF tion of a pleurodesis is an injury to the pleura because
levels. Montes et al. ( 1 38) have demonstrated nerves the intrapleural injection of all the agents listed in
in adhesions between the visceral and parietal pleural the preceding text produces an acute exudative pleu
seven days after the intrapleural administration of talc ral effusion. However, the factors that dictate whether
slurry. the inflammatory response will resolve or proceed to
Recent studies have demonstrated that the intra pleural fibrosis are not known. It has been shown that
pleural administration of 2 mL of 0.5% silver nitrate the intrapleural administration of tetracycline results
produces an excellent pleurodesis in rabbits ( 1 39) . in the local elaboration of IL-8 and monocyte che
The pleurodesis resulting from this dose of silver motactic protein 1 (MCP- 1 ) ( 1 49). The pleurodesis
nitrate is equivalent to that resulting from 35 mg/kg resulting from talc slurry but not that resulting from
of tetracycline ( 1 30) and better than that resulting tetracycline derivatives is partially inhibited by block
from 400 mg/kg of talc ( 1 39, 1 40) . The pleurodesis ing antibodies to TNF-a ( 1 23) and corticosteroids
following silver nitrate intrapleurally persists for at ( 1 50).
least 1 year ( 1 40) . Marchi et al. ( 1 37) have shown Cytokines are without a doubt involved in the
that there is a systemic inflammatory reaction after production of a pleurodesis. In our first experiment
silver nitrate is given intrapleurally manifested by an to assess this assertion, we injected IL-8 alone or
increase in the serum LDH and IL-8 levels along with with talc to see if it would facilitate the creation of a
an increase in the VEGF level in the serum. This same pleurodesis. We found that IL-8 had no effect on the
group has demonstrated that if 0 . 1 % silver nitrate is production of the pleurodesis in either circumstance
given three times rather than 0. 5% silver nitrate once, (unpublished data) . We next turned our attention to
the systemic inflammatory is markedly reduced, but TGF-/3 . We demonstrated that the intrapleural injec
the degree of pleurodesis is not altered. ( 1 4 1 ) . Trem tion of one dose ofTGF-/3 2 could produce a pleurode
blay ( 1 42) and associates have demonstrated that the sis in a dose-dependent manner with a dose of 1 .67
administration of 2 ml of 0.05% silver nitrate daily µg/kg producing an excellent pleurodesis ( 1 5 1 ) . The
for 1 4 days results in an adequate pleurodesis. intrapleural injection of TGF-/3 2 also led to the pro
Guo and associates ( 1 43) have shown that the duction of large amounts of exudative pleural fluid,
intrapleural instillation of2 ml of2% or 4% iodopovi but the WBC count and the LDH level in the fluid
done (Betadine) produces a pleurodesis equivalent to were much lower than in the fluid resulting from the
that produced by 1 0 mg/kg doxycycline. If 0.8 mg/kg intrapleural administration of either talc slurry or tet
triamcinolone was administered weekly intramuscu racycline derivatives ( 1 5 1 ) . These observations sug
larly to the rabbits that received 4% iodopovidone, gested that TGF-/3 2 stimulated the mesothelial cells
the degree of pleurodesis was markedly reduced ( 1 43) . to produce collagen even in the absence of injury
The efficacy of various antineoplastic agents to to the pleura. We subsequently demonstrated that
induce a pleurodesis has also been investigated. the pleurodesis that resulted from TGF-/3 2 occurred
Although bleomycin is used in humans with malig much faster than that which occurred following talc
nant pleural effusions to produce a pleurodesis, administration ( 1 52) and that the pleurodesis that
it is ineffective in rabbits at doses up to 3.0 IU/kg occurs following intrapleural TGF-/3 2 is not inhibited
( 1 20, 1 44) . In rabbits, the best antineoplastic agent by corticosteroids ( 1 53). We have also shown that
for producing a pleurodesis appears to be nitrogen TGF-/3 produces a pleurodesis comparable to that
3
CHAPT E R 4 I A N I MAL M O D E LS I N PLE U RAL I NVESTIGAT I O N 45
Q)
0
(.) 6
(/)
(/)
"iii
Q)
"C
e
:::J 4
Q)
c:::
0 2 3 4 5 6 7
P l e u ral vascular density (percent of area occupied by blood vessels)
thorne M, et al. Pleurodesis is inhibited by anti vascular endothelial growth factor antibody C hest.
2005; 7 28: 7 790-7 797.)
produced by TGF-,8 2 ( 1 54) . I feel confident that the the angiogenesis in the pleural tissue and the pleurode
future for pleurodesis lies in its production by the sis score ( 1 57) (Fig. 4.2) .
intrapleural injection of cytokines. Teixeira et al ( 1 58) have shown that anti-VEGF
This rabbit model has been used to investigate antibodies also reduce the pleurodesis that occurs
the mechanisms involved in the production of a after the intrapleural administration of talc or silver
pleurodesis. Idell et al. ( 1 55) have demonstrated that nitrate. These studies demonstrate that pleural angio
the intrapleural administration of single-chain uro genesis is very important in producing a pleurodesis.
kinase plasminogen activator 24 and 48 hours after
intrapleural tetracycline prevents the formation of Dog Model. There have been several studies of
adhesions at 72 hours indicating that fibrinogenesis pleurodesis that utilized dogs. However, dogs, as
is important in producing a pleurodesis. Marchi et al do rabbits, have a thin visceral pleura, so extrapolat
( 1 56) measured the pleural fluid levels of IL-8, VEGF ing results in dogs to humans may not be possible.
and TGF-,8 1 in rabbits 6. 24 and 48 hour after they Bresticker et al. ( 1 59) performed bilateral thoracoto
received 200 mg/kg talc. They found that the pleural mies on mongrel dogs and subjected the animals to
fluid IL-8 concentration peaked at 6 hours whereas the mechanical dry gauze abrasion, chemical sclerosis
VEGF and TGF-,8 1 concentrations increased steadily with 500 mg of tetracycline, talc poudrage with 1 g,
over 24 hours. Guo et al. ( 1 5 7) administered anti Nd:yttrium aluminium garnet (YAG) laser photoco
VEGF antibodies before rabbits were given TGF-,8 2 agulation, or argon beam electrocoagulation of the
intrapleurally with the hypothesis that the TGF-,8 2 parietal pleura ( 1 59) . When the animals were sac
was inducing VEGF that was responsible for the large rificed at 30 days, the mechanical abrasion and the
amounts of pleural fluid seen with intrapleural TGF-,8 2 . talc were found to produce the best pleurodesis and
They were surprised when they observed that the rab were virtually equivalent. Jerram et al. ( 1 60) reported
bits that received the anti-VEGF antibodies had much that mechanical abrasion was significantly better than
less effective pleurodesis ( 1 57) . When all the rabbits 1 g talc slurry at producing a pleurodesis ( 1 60) . Colt
were analyzed, there was a close relationship between et al. ( 1 6 1 ) compared the pleurodesis resulting from
46 P L E U RAL D I S EASES
dry gauze abrasion, thoracoscopic mechanical abrasion absorbable polyglactin (Vicryl) or nonabsorbable
using a commercially available stainless-steel grooved polypropylene (Marlex) meshes in rats. Marlex alone
burr abrader, thoracoscopic talc insuffiation (4 g), and incited an extensive and dense pleural reaction not
instillation of talc slurry (5 g) at 30 days. They reported suitable for clinical use because it induces a fibrotho
that the talc insuffiation was the best, followed by rax ( 1 ) . However, the use of Vicryl in conjunction
mechanical abrasion ( 1 6 1 ) . The differing results in the with pleural abrasion produced a better pleurodesis
studies mentioned in the preceding text may be related than the pleural abrasion alone 3 to 4 months after
to the vigor with which mechanical abrasion was per the procedure ( 1 66) . Cetin et al. ( 1 67) demonstrated
formed or the amount of talc used. that the intrapleural administration of 2.5 mg of
0.5% polidocanol, a sclerosing agent used in the treat
Pig Model. There have also been several articles in ment of extremity and esophageal varices, was more
which pleurodesis has been studied in pigs, a species effective than 35 mg/kg of tetracycline in producing
which has a thick pleura similar to humans. Whitlow a pleurodesis. The same group ( 1 68) subsequently
et al. compared the pleurodesis that results from 3 g of demonstrated that fibrin tissue adhesive with fibrino
insuffiated talc and 300 mg of minocycline adminis gen and thrombin concentrations of 30 mg/mL and
tered during thoracoscopy and reported that the insuf 10 U/mL was more effective for pleurodesis than tet
flated talc produced superior pleurodesis ( 1 62) . Cohen racycline 35 mg/kg. Werebe et al. ( 1 69) administered
et al. reported that the pleurodesis resulting from the 1 0 or 20 mg of talc intrapleurally to 40 rats and found
insuffiation of 5 g of talc or the instillation of 5 g of talc crystals in every organ examined 24 to 48 hours
talc as a slurry was virtually identical ( 1 63). Lardinois after the intrapleural injection. Fraticelli et al. ( 1 70)
et al. ( 1 64) reported that the systemic administration came to the opposite conclusion after they injected
of diclofenac (2 mg/kg) orally for 3 weeks after surgery 33 rats with 40 mg of talc slurry. They reported that
significantly reduced the pleurodesis scores in pigs. a few talc particles were found in the liver of two rats,
Sheep Model. There has been one pleurodesis
in the spleen of one rat, and on the brain surface of
study in sheep, another species with a thick pleura like another rat ( 1 70) . Unfortunately, there were no con
that in humans. Lee et al. ( 1 65) demonstrated that trol animals in either of these studies.
the intrapleural administration of TGF-/3 2 produced
Mouse model. It would be very useful to have a
an excellent pleurodesis in sheep as it had in rabbits
( 1 5 1 ) . There were two significant differences in the mouse model of pleurodesis. Mice are inexpensive and
results with sheep and those with rabbits. First, the the availability of knockout mice would allow one to
dose ofTGF-/3 2 necessary for pleurodesis in the rabbits study the mechanisms of pleurodesis. However, it is dif
( 1 .67 µg/kg) was much larger than the dose necessary ficult to produce pleurodesis in mice. Kalomenidis et al.
for pleurodesis in sheep (0.25 µg/kg) . One possible ( 1 7 1 ) attempted to produce pleurodesis in mice by the
explanation for this observation is that the dose is injection of up to 1 ,360 µg/kg ofTGF-{3 2, 1 00 mg/kg
more dependent on the body surface area than on the of doxycycline or 4 g/kg of talc, and none of the mice
weight. This is a reasonable supposition because the had a pleurodesis in which there was more than 25%
surface of the lung is more directly correlated with symphysis between the visceral and parietal pleura.
the body surface area than it is with the body weight.
Indeed, if the dose of the TGF-/3 2 is adjusted for body Pleural Fibrosis. Animal models have also been
surface area, the dose in each species is approximately used to study pleural fibrosis. Decologne and asso
the same, 9.2 µglm2 in rabbits and 6.3 µglm2 in sheep. ciates ( 1 72) used adenoviral gene transfer of TGF
This relationship of dose to body surface area should beta 1 in rats to demonstrate that local and transient
be considered when results in animals are extrapolated TGF-beta 1 overexpression induces homogenous,
to humans. Secondly, in sheep, the pleurodesis could prolonged, and progressive pleural fibrosis without
be accomplished without the side effect of producing pleurodesis. They further demonstrated that pleural
large amounts of pleural fluid. These studies demon fibrosis can expand into the lung parenchyma from
strate that the species used has a significant influence the visceral layer, but not into the muscle from the
on the results for pleurodesis. parietal layer ( 1 72) . This same group ( 1 73) also
showed that carbon particles administered to the pleu
Rat Model. Pleurodesis has also been investi ral cavity caused severe pleural fibrosis in the presence
gated in the rat. Sugarmann et al. ( 1 66) studied the of bleomycin, whereas bleomycin alone had no fibro
effectiveness of pleural abrasion with or without genic effect. The pleural response was associated with
CHAPT E R 4 I A N I MAL M O D E LS I N PLE U RAL I NVESTIGAT I O N 47
progressive fibrosis in subpleural regions, similar to factor/vascular permeability factor by human lung cancer
human interstitial pulmonary fibrosis. Matrix accu cells. Am ] Pathol. 2000; 1 57 : 1 893- 1 903.
15. Nagamachi Y, Tani M, Shimizu K, et al. Orthotopic growth
mulation within this area evolved through "mesothe and metastases of human non-small cell lung carcinoma cells
lial-fibroblastoid transformation" where mesothelial injected into the pleural cavity of nude mice. Cancer Lett.
cells acquire myofibroblast characteristics. In con 1 998; 1 27:203-209.
trast, carbon did not exaggerate bleomycin-induced 16. Ohta Y, Kimura K, Tamura M, et al. Biological characteristics
of carcinomatosa pleuritis in orthotopic model systems using
pulmonary fibrosis after combined intratracheal
immune deficient rats. Int] Oncol. 200 1 ; 1 8 :499-5 0 5 .
administration ( 1 73) . These studies suggest that the 17. Ebright M l , Zager J S , Malhotra S, e t al. Replication-competent
mesothelium might be important in the production herpes virus NV! 020 as direct treatment of pleural cancer in a
of interstitial pulmonary fibrosis. rat model. J Thorac Cardiovasc Surg. 2002; 1 24: 1 23-1 29 .
18. Prosst RL, Winkler S, Boehm E , e t a l . Thoracoscopic fluores
cence diagnosis (TFD) of pleural malignancies: experimental
studies. Thorax. 2002;57: 1 005-1 009.
REFERENCES 1 9. Hatton MW, Southward SM, Ross BL, et al. Angiostatin II is
the predominant glycoform in pleural effusates of rabbit VX-2
I . Lee YCG. Animal models for pleural diseases. In: Light RW, lung tumors. J Lab Clin Med. 2002; 1 39:3 1 6-323 .
Lee YCG, eds. Textbook ofPleural Diseases. London, England: 20. Hatton MW, Southward SM, Ross BL, et al. Relationships
Arnold Publishers; 2003 : 1 49- 1 66. among tumor burden, tumor size, and the changing concen
2 . Stathopoulos GT, Kalomenidis I . Animal models of malignant trations of fibrin degradation products and fibrinolytic facrors
pleural effusion. Curr Opin Pulm Med. 2009; 1 5 :343-352. in the pleural effusions of rabbits with VX2 lung tumors. J Lab
3 . Stathopoulos GT, Zhu Z, Everhart MB, et al. Nuclear Clin Med. 2006; 1 47:27-3 5 .
Factor {kappa} B affects tumor progression in a mouse model 21. Wang B , Xiong Q, Shi Q, e t al. Genetic disruption o f host
of malignant pleural effusion. Am J Respir Cell Mo! Biol. nitric oxide synthase II gene impairs melanoma induced
2006;34: 1 42- 1 5 0 . angiogenesis and suppresses pleural effusion. Int J Cancer.
4. Z h u Z, Stathopoulos GT, Dikensoy 0 , et al. The efficacy o f 200 1 ;9 1 :607-6 1 1 .
S U ! 1 248 in inhibition of tumor growth and pleural fluid for 22. Yasutake N , Matsuzaki T, Kimura K, et al. Th e role o f tumor
mation in a mouse malignant pleural effusion model. Proc Am necrosis factor (TNF) alpha in the antitumor effect of intra
Thor Soc. 2005;2:A536. pleural injection of Lactobacillus casei strain Shirota in mice.
5 . Fang F, Chen P, Wu X, et al. Therapeutic effects of recom Med Microbiol Immunol (Berl). 1 999; 1 8 8 : 9- 1 4 .
binant human endostatin adenovirus in a mouse model 23. Wagner JC. Experimental production of mesothelial tumours
of malignant pleural effusion. J Cancer Res Clin Oncol. of the pleura by implantation of dusts in laboratory animals.
2009; 1 3 5 : 1 1 49-1 1 57. Nature. 1 962; 1 96: 1 80- 1 8 1 .
6. Stathopoulos GT, Sherrill TP, Karabela SP, et al. Host 24. Smith WE, Miller L , Churg J , e t al. Mesotheliomas i n ham
derived !nterleukin-5 promotes adenocarcinoma-induced sters following intrapleural injection of asbestos. J Mt Sinai
malignant pleural effusion. Am J Respir Grit Care Med. Hosp. 1 965;32: 1 -8 .
20 1 0; 1 82 : 1 273-1 28 1 . 25. Carthew P, Hill RJ, Edwards RE, e t al. Intrapleural admin
7 . Moschos C, Psallidas I , Cottin T, et al. A sulindac analogue istration of fibres induces mesothelioma in rats in the same
is effective against malignant pleural effusion in mice. Lung relative order of hazard as occurs in man after exposure. Hum
Cancer. 20 1 1 ;73 : 1 7 1 - 1 7 5 . Exp Toxicol. 1 9 92; 1 1 : 530-534.
8 . Moschos C, Psallidas I, Kollintza A , e t al. Th e angiopoietin/ 26. Davis JM. Structural variations between pleural and perito
Tie2 axis mediates malignant pleural effusion formation. neal mesotheliomas produced in rats by the injection of cro
Neoplasia. 2009; 1 1 :298-304. cidolite asbestos. Ann Anat Pathol (Paris). 1 976;2 l : 1 99-2 1 0 .
9 . Stathopoulos GT, Moschos C, Loutrari H, et al. Zoledronic 27. Jaurand M C , Fleury J, Monchaux G, e t al. Pleural carci
acid is effective against experimental malignant pleural effu nogenic potency of mineral fibers (asbestos, attapulgite)
sion. Am ] Respir Grit Care Med. 2008 ; 1 76 : 5 0-59. and their cytotoxicity on cultured cells. J Natl Cancer Inst.
10. Cui R, Takahashi F, Ohashi R, et al. Osteopontin is involved 1 9 87;79:797-804.
in the formation of malignant pleural effusion in lung cancer. 28. Monchaux G, Bignon J, Jaurand MC, et al. Mesotheliomas in
Lung Cancer. 2009;63;368-374. rats following inoculation with acid-leached chrysotile asbes
1 1 . Zhang JX, Xie CM, Zhu ZW, et al. Potential role of AQP l tos and other mineral fibres. Carcinogenesis. 1 9 8 1 ;2:229-236.
and VEGF in the development of malignant pleural effusion 29. Van der Meeren A, Fleury J, Nebut M , et al. Mesothelioma
in mice. Med Oncol. 2 0 1 2;29 :656-662. in rats following intrapleural injection of chrysotile and
12. Yano S, Nokihara H, Hanibuchi M, et al. Model of malig phosphorylated chrysotile (chtysophosphate) . Int J Cancer.
nant pleural effusion of human lung adenocarcinoma in SCIO 1 992;50 :937-942.
mice. Oncol Res. 1 997;9: 573-579. 30. Wagner JC, Berry G, Timbrell V. Mesotheliomata in rats
13. Yano S, Herbst RS, Shinohara H, et al. Treatment for malig after inoculation with asbestos and other materials. BrJ Cancer.
nant pleural effusion of human lung adenocarcinoma by inhi 1 973;28: 1 73-1 8 5 .
bition of vascular endothelial growth factor receptor tyrosine 31. Wagner J C , Hill RJ, Berry G, e t al. Treatments affecting
kinase phosphorylation. Clin Cancer Res. 2000;6:957-965 . the rate of asbestos-induced mesotheliomas. Br J Cancer.
1 4 . Yano S, Shinohara H, Herbst RS, e t al. Production o f experi 1 9 80;4 l : 9 1 8-922.
mental malignant pleural effusions is dependent on invasion 32. Wagner JC, Skidmore JW, Hill RJ, et al. Erionite exposure and
of the pleura and expression of vascular endothelial growth mesotheliomas in rats. Br J Cancer. 1 9 8 5 ; 5 1 :727-730.
48 P L E U RAL D I S EASES
33. Baris YI, Saracci R, Simonato L, et al. Malignant mesothe 5 1 . Boylan AM, Ruegg C, Kim KJ, et al. Evidence of a role
lioma and radiological chest abnormalities in two villages in for mesothelial cell-derived interleukin 8 in the pathogen
central Turkey. An epidemiological and environmental inves esis of asbesros-induced pleurisy in rabbits. J Clin Invest.
tigation. Lancet. 1 9 8 1 ; I :984-9 87. 1 992;89: 1 257- 1 267.
34. Fleury-Feith J, Nebut M, Saint-Etienne L, et al. Occurrence 52. Sahn SA, Antony VB. Pathogenesis of pleural plaques.
and morphology of tumors induced in nude mice transplanted Relationship of early cellular response and pathology. Am Rev
with chrysotile-transformed rat pleural mesothelial cells. Biol Respir Dis. 1 9 84; 1 30 : 8 84-887.
Cell. 1 9 89;65:45-50. 5 3 . Colice GL, Curtis A, Deslauriers J, et al. Medical and surgi
35. Berman OW, Crump KS, Chatfield EJ, et al. The sizes, shapes, cal treatment of parapneumonic effusions: an evidence-based
and mineralogy of asbestos structures that induce lung tumors guideline. Chest. 2000; 1 1 8 : 1 1 5 8-1 1 7 1 .
or mesothelioma in AF/HAN rats following inhalation. Risk 5 4 . Light RW. Animal models o f pleural investigation. In:
Anal. 1 995; 1 5 : 1 8 1 - 1 9 5 . Bouras D, ed. Pleural Disease. New York, NY: Marcel Dekker
3 6 . Davis J M , Jones AD . Comparisons o f the pathogenicity of Inc; 2004: 1 009-1 034.
long and short fibres of chrysotile asbestos in rats. Br J Exp 55. Mavroudis C, Ganze! BL, Katzmark S, et al. Effect of hemo
Pathol. 1 988;69:7 1 7-737. thorax on experimental empyema thoracis in the guinea pig.
37. Davis JM, Addison J, Bolton RE, et al. The pathogenicity of J Thorac Cardiovasc Surg. 1 9 85;89 :42-49.
long versus short fibre samples of amosite asbestos adminis 56. Graham EA, Bell RO . Open pneumothorax: its relations ro
tered to rats by inhalation and intraperitoneal injection. Br J the treatment of empyema. Amf Med Sci. 1 9 1 8 ; 1 56:839-87 1 .
Exp Pathol. 1 986;67:4 1 5-430. 57. Sahn SA, Potts DE. Turpentine pleurisy in rabbits: a model
3 8 . Davis JM, Addison J, Bolton RE, et al. Inhalation and of pleural fluid acidosis and low pleural fluid glucose. Am Rev
injection studies in rats using dust samples from chrysotile Respir Dis. 1 978; 1 1 8 : 893-90 1 .
asbestos prepared by a wet dispersion process. Br J Exp Pathol. 5 8 . Sahn SA, Taryle DA, Good JT Jr. Experimental empyema:
1 986;67: 1 1 3- 1 29. time course and pathogenesis of pleural fluid acidosis and low
39. Johnson NF, Edwards RE, Munday DE, et al. Pluripotential pleural fluid glucose. Am Rev Respir Dis. 1 979; 1 2 0 : 3 5 5-36 1 .
nature of mesotheliomata induced by inhalation of erionite in 5 9 . Shohet I , Yellin A, Meyerovitch J , et al. Pharmacokinetics
rats. Br J Exp Pathol. 1 984;65 :377-3 8 8 . and therapeutic efficacy of gentamicin in an experimental
40. Wagner J C , Berry G, Skidmore J W, et al. The effects o f the pleural empyema rabbit model. Antimicrob Agents Chemother.
inhalation of asbestos in rats. Br J Cancer. 1 974;29 :252-269 . 1 987;3 1 :982-9 8 5 .
4 1 . Wagner J C , Berry G, Skidmore J W, e t al. The comparative 6 0 . Liapakis I E , Light RW, Pitiakoudis MS, e t al. Penetration o f
effects of three chrysotiles by injection and inhalation in rats. clarithromycin i n experimental pleural empyema model fl u id.
!ARC Sci Pub/. 1 9 80;30:363-372. Respiration. 2005;72:296-300.
42. Webster I, Goldstein B, Coetzee FS, et al. Malignant mesothe 6 1 . Liapakis IE, Kottakis I, Tzatzarakis MN, et al. Penetration
lioma induced in baboons by inhalation of amosite asbestos. of newer quinolones in the empyema fluid. Eur Respir J
Am ] Ind Med. 1 993;24:659-666. 2004;24:466-470 .
43. Pylev LN. Pretumorous lesions and lung and pleural tumours 62. Mavroudis C, Ganze! BL, Cox SK, et al. Experimental aerobic
induced by asbesros in rats, Syrian golden hamsters and Macaca anaerobic thoracic empyema in the guinea pig. Ann Thorac
mulatta (rhesus) monkeys. !ARC Sci Pub/. 1 9 80;30:343-3 5 5 . Surg. 1 9 87;43:29 5-297.
4 4 . Adachi S, Kawamura K , Kimura K , e t al. Tumor incidence was 63. Sasse SA, Causing LA, Mulligan ME, et al. Serial pleural fluid
not related to the thickness of visceral pleural in female Syrian analysis in a new experimental model of empyema. Chest.
hamsters intratracheally administered amphibole asbestos or 1 996; 1 09 : 1 043- 1 048.
manmade fibers. Environ Res. 1 992; 5 8 : 5 5-65 . 64. Sasse S, Nguyen T, Teixeira LR, et al. The utility of daily
4 5 . Humphrey EW, Ewing SL, Wrigley J V, e t al. Th e produc therapeutic thoracentesis for the treatment of early empyema.
tion of malignant tumors of the lung and pleura in dogs Chest. 1 999; 1 1 6 : 1 703- 1 708.
from intratracheal asbestos instillation and cigarette smoking. 6 5 . Sasse S, Nguyen TK, Mulligan M, et al. The effects of
Cancer. 1 9 8 1 ;47: 1 994-1 999. early chest tube placement on empyema resolution. Chest.
46. Adachi S, Kawamura K, Takemoro K. A trial on the quanti 1 997; 1 1 1 : 1 679-1 683.
tative risk assessment of man-made mineral fibers by the rat 66. Teixeira LR, Sasse SA, Villarino MA, et al. Antibiotic levels in
intraperitoneal administration assay using the JFM standard empyemic pleural fluid. Chest. 2000; 1 1 7: 1 734- 1 739.
fibrous samples. Ind Health. 200 1 ;39: 1 68-1 74. 67. Sasse SA, Jadus MR, Kukes GD. Pleural fluid transform
47. Miller BG, Searl A, Davis JM, et al. Influence of fibre ing growth factor-/3 1 correlates with pleural fibrosis in
length, dissolution and biopersistence on the production of experimental empyema. Am J Respir Crit Care Med. 2003;
mesothelioma in the rat perironeal cavity. Ann Occup Hyg. 1 68 :700-70 5 .
1 999;43: 1 5 5- 1 66. 68. Kunz CR, Jadus MR, Kukes G, e t al. Intrapleural injection of
48. Davis MR, Manning LS, Whitaker 0, et al. Establishment transforming growth factor /3 antibody inhibits pleural fibrosis
of a murine model of malignant mesothelioma. Int J Cancer. in empyema. Chest. 2004; 1 26: 1 636-1 644.
1 992; 52: 8 8 1 -886. 69. Zhu Z, Hawthorne ML, Guo Y, et al. Tissue plasmino
49. Holiat SM, Smith WE, Hubert DD, et al. Chemotherapeutic gen activator combined with human recombinant DNase
trials with hamster mesothelioma I 0-24: responses to azaciti is effective therapy for empyema in a rabbit model. Chest.
dine, aziridinylbenzoquinone, cisplatin, and PCNU. Cancer 2006; 1 2 9 : 1 577- 1 5 8 3 .
Treat Rep. 1 9 8 1 ;6 5 : 1 1 1 3-1 1 1 5 . 70. Oikensoy 0 , Zhu Z, N a MJ, et al. Intrapleural heparin o r
5 0 . Shore BL, Daughaday CC, Spilberg I . Benign asbestos pleu heparin combined with human recombinant DNase is not
risy in the rabbit. A model for the study of pathogenesis. Am effective in the treatment of empyema in a rabbit model.
Rev Respir Dis. 1 983; 1 28:48 1-4 8 5 . Respirology. 2006; 1 1 :75 5-760.
CHAPT E R 4 I A N I MAL M O D E LS I N PLE U RAL I NVESTIGAT I O N 49
7 1 . N a MJ, Liao H, Moschos C, e t al. Tissue plasminogen acti 90. Hanada S, Sugawara SH, Sertie JA. Miconazole as inflamma
vator increases the volume of pleural fluid by inflammatory tory agent. II: Time course of pleurisy and drug interference.
responses in rabbits with empyema. Chest. 2005; 1 26 : 3 1 8S. Gen Pharmacol. 1 998;30:79 1-794.
72. Schopf LF, Fraga JC, Amantea SL, et al. Induction of pleural 9 1 . Vinegar R, Truax JF, Selph JL, et al. Pathway of onset, devel
empyema in rats by thoracentesis with intrapleural pressure opment, and decay of carrageenan pleurisy in the rat. Fed
monitoring. Pediatr Surg Int. 2004;20 : 5 1 5-5 1 9 . Proc. 1 982;4 l :942-946.
73. Luo L, Zhou Q, Chen XJ, et al. Effects of the TREM-1 path 92. Mielens ZE, Connolly K, Stecher VJ. Effect of disease
way modulation during empyema in rats. Chin Med J (Engl}. modifying antirheumatic drugs and nonsteroidal antiinflam
20 1 O ; 1 23 : 1 5 6 1-1 5 6 5 . matory drugs upon cellular and fibronectin responses in a
7 4 . Wilkosz S, Edwards LA , Bielsa S, et al. Characterization of a pleurisy model. ] Rheumatol. 1 9 8 5 ; 1 2 : 1 083- 1 087.
new mouse model of empyema and the mechanisms of pleural 93. Cuzzocrea S, Mazzon E, Calabro G, et al. Inducible nitric
invasion by Streptococcus pneumonia. Am J Respir Cell Mol oxide synthase-knockout mice exhibit resistance to pleurisy
Biol. 20 1 1 ;46: 1 80- 1 87. and lung injury caused by carrageenan. Am] Respir Crit Care
7 5 . Allen JC, Apicella MA. Experimental pleural effusion as a Med. 2000; 1 62 : 1 8 5 9- 1 866.
manifestation of delayed hypersensitivity to tuberculin PPD. 94. Pinheiro RM, Calixto JB. Effect of the selective COX-2
] Immunol. 1 968; 1 0 1 :48 1 -487. inhibitors, celecoxib and rofecoxib in rat acute models of
76. Leibowitz S, Kennedy L, Lessof MH. The tuberculin reaction inflammation. Injlamm Res. 2002;5 l :603-6 1 0 .
in the pleural cavity and its suppression by antilymphocyte 9 5 . D'Acquisto F, Ianaro A , Ialenti A , e t al. Activation o f nuclear
serum. Br] Exp Pathol. 1 973;54: 1 52- 1 62. transcription factor kappa B in rat carrageenin-induced pleu
77. Widstrom 0, Nilsson BS. Pleurisy induced by intrapleu risy. Eur] Pharmacol. 1 999;369:233-236.
ral BCG in immunized guinea pigs. Eur J Respir Dis. 96. Frode TS, Souza GE, Calixto JB. The modulatory role played
l 982;63 :425-434. by TNF-alpha and IL- 1 beta in the inflammatory responses
78. Widstrom 0, Nilsson BS. Low in vitro response to PPD and induced by carrageenan in the mouse model of pleurisy.
PHA in lymphocytes from BCG-induced pleurisy in guinea Cytokine. 200 1 ; 1 3 : 1 62- 1 6 8 .
pigs. EurJ Respir Dis. 1 9 82;63:43 5-444. 9 7 . Sautebin L, Ialenti A , Ianaro A , e t al. Relationship berween
79. Paterson RC. The pleural reaction to inoculation with tuber nitric oxide and prostaglandins in carrageenin pleurisy.
cule bacilli in vaccinated and normal guinea pigs. Am Rev Biochem Pharmacol. 1 99 8 ; 5 5 : 1 1 3-1 1 7 .
Tuberc. 1 9 1 7 ; 1 :353-37 1 . 9 8 . Frode T S , Souza G E , Calixto JB. Th e effects of IL-6 and
80. Antony VB, Sahn SA, Antony AC, e t al. Bacillus Calmette IL- 1 0 and their specific antibodies in the acute inflamma
Guerin stimulated neutrophils release chemotaxins for tory responses induced by carrageenan in the mouse model
monocyces in rabbit pleural space in vitro. ] Clin Invest. of pleurisy. Cytokine. 2002; 1 7 : 149- 1 56.
1 9 8 5;76: 1 5 1 4- 1 5 2 1 . 99. Cuzzocrea S, Sautebin L, De Sarro G, et al. Role of IL-6 in the
8 1 . Ly LH, Jeevan A, McMurray ON. Neutralization of TNF pleurisy and lung injury caused by carrageenan. J Immunol.
alpha alters infl a mmation in guinea pig tuberculous pleuritis. 1 999; 1 63:5 094-5 1 04.
Microbes Infect. 2009; 1 1 :680-688. 1 00 . Yuhki KI, Ueno A, Naraba H, et al. Prostaglandin recep
82. Du H, Xie C, He Q, et al. Increased expression of Aquaporin- 1 tors, EP2, EP3 and IP, mediate exudate formation in car
on the pleura of rats with a Tuberculous pleural effusion. Lung. rageenin induced mouse pleurisy. J Pharmacol Exp Ther.
2007; 1 8 5 :325-336. 2004;3 1 1 : 1 2 1 8- 1 224.
83. Moore AR. Pleural models of inflammation: immune and 1 0 1 . Cuzzocrea S, Rossi A, Serraino I, et al. 5 Lipoxygenase knock
nonimmune. Methods Mol Biol. 2003;22 5 : 1 23- 1 2 8 . out mice exhibit a resistance to pleurisy and lung injury
8 4 . Imai Y, Hayashi M, Ohishi S . Key role o f complement caused by carrageenan. J Leukoc Biol. 2003;73 :739-746.
activation and platelet activating factor in exudate forma 1 02 . Kikuchi M, Tsuzurahara K, Naito K. Involvement of leu
tion in zymosan induced rat pleurisy. Jpn J Pharmacol. kotriene B4 in zymosan induced rat pleurisy: inhibition of
1 9 9 1 ;57:225-232. leukocyte infiltration by the 5 lipoxygenase inhibitor T 0757.
85. Utsunomiya I, Ito M, Ohishi S . Generation of inflammatory Biol Pharm Bull. 1 99 5 ; 1 8 : 1 302-1 034.
cycokines in zymosan-induced pleurisy in rats: TNF induces 1 03 . Sampaio AL, Rae GA, Henriques MG. Participation of
IL-6 and cytokine-induced neutrophil chemoattractant (cine) endogenous endothelins in delayed eosinophil and neu
in vivo. Cytokine. 1 998; 1 0 :9 5 6-963. trophil recruitment in mouse pleurisy. Injlamm Res.
86. Silva AR, Larangeira AP, Pacheco P, et al. Bradykinin down 2000;49 : 1 70- 1 76.
regulates LPS-induced eosinophil accumulation in the pleural 1 04 . Yuhki Kl, Ushikubi F, Naraba H, et al. Prostaglandin I2 plays
cavity of mice through type 2-kinin receptor activation: a role a key role in Zymosan-induced mouse pleurisy. J Pharmacol
for prostaglandins. Br J Pharmacol. 1 999; 1 27: 569-575. Exp Ther. 2008;325 :60 1 -609.
87. Edamitsu S, Macsukawa A, Ohkawara S, et al. Role of TNF 105. Castro-Faria-Neto HC, Penido CM, Larangeira AP, et al.
alpha, IL 1, and IL 1 ra in the mediation ofleukocyce infiltration A role for lymphocytes and cytokines on the eosino
and increased vascular permeability in rabbits with LPS induced phil migration induced by LPS . Mem Inst Oswaldo Cruz.
pleurisy. Clin Immunol Immunopathol. 1 995;75:68-74. 1 997;92(suppl 2) : 1 97-200.
8 8 . Pasquale CP, Lima MC, Bandeira-Melo C, et al. Systemic 1 06. Fukumoto T, Matsukawa A, Yoshimura T, et al. IL 8 is an
and local dexamethasone treatments prevent allergic eosino essential mediator of the increased delayed phase vascular
philia in rats via distinct mechanisms. Eur ] Pharmacol. permeability in LPS induced rabbit pleurisy. ] Leukoc Biol.
1 999;368:67-74. 1 998;63:584-5 90.
89. Aihara S, Murakami N, Tomita T, et al. Inhibitory action of 1 07. Broaddus VC, Boylan AM, Hoeffel JM, et al. Neutralization
indomethacin on neutrophil infiltration in monosodium urate of IL 8 inhibits neutrophil influx in a rabbit model of endo
induced pleurisy in rats. Jpn ] Pharmacol. 1 99 5 ;68:27 1-277. toxin induced pleurisy. ] Immunol. 1 994; 1 52 :2960-2967.
50 P L E U RAL D I S EASES
108. Silva PMR, Martins MA, Castro-Faria-Neta HC, et al. Gen 1 26. Zhu Z, Donnelly E, Dikensoy 0, et al. Efficacy of ultra
eration of an eosinophilotactic activity in the pleural cavity of sound in the diagnosis of pleurodesis in rabbits. Chest.
platelet-activating facror-injected rats. J Pharmacol Exp Ther. 2005; 1 28:934-939.
1 9 9 1 ;257: I 039- 1 044. 1 27. Komissarov AA, Mazar AP, Koenig K, et al. Regulation of
1 09. Bozza PT, Castro-Faria-Nero HC, Maccins MA, et al. Phar intrapleural fibrinolysis by urokinase-{alpha}-macroglobulin
macological modulation of lipopolysaccharide-induced complexes in tetracycline-induced pleural injury in rabbits.
pleural eosinophilia in the rat; a role for a newly generated Am J Physiol Lung Cell Mo/ Physiol. 2009;297:L568-L577.
protein. EurJ Pharmacol. 1 993;248 : 4 1 -47. 128. Light RW, Wang N-S, Sassoon CSH, et al. Talc slurry
1 1 0 . Ferreira HH, Medeiros MV, Lima CS, et al. Inhibition of is an effective pleural sclerosanr in rabbits. Chest.
eosinophil chemotaxis by chronic blockade of nitric oxide 1 99 5 ; 1 07: 1 702- 1 706.
biosynthesis. EurJ Pharmacol. 1 996;3 l 0 :20 1 -207. 129. Xie C, Teixeira LR, Wang N-S, et al. Serial observations after
1 1 1 . Bozza PT, Castro-Faria-Nero HC, Penido C, et al. IL-5 high dose talc slurry in the rabbit model for pleurodesis.
accounts for the mouse pleural eosinophil accumulation Lung. 1 998; 1 76:299-307.
triggered by antigen but not by LPS . Immunopharmacology. 1 30. Vargas FS, Teixeira LR, Silva LMMF, et al. Comparison of
1 994;27: 1 3 1 - 1 3 8 . silver nitrate and tetracycline as pleural sclerosing agents in
1 1 2. Lima M C , Martins MA, Perez SA, e t al. Effect of azelastine rabbits. Chest. 1 99 5 ; 1 0 8 : 1 080- 1 0 83.
on platelet-activating factor and antigen-induced pleurisy in 1 3 1 . Xie C, McGovern J, Wu W, et al. Comparisons of pleurode
rats. EurJ Pharmacol. 1 99 1 ; 1 97:20 1 -207. sis induced by talc with or without thymol iodide in rabbits.
1 1 3 . Sampaio AL, Rae GA, Henriques MM. Role of endothelins Chest. 1 998; 1 1 3:795-79 9.
on lymphocyte accumulation in allergic pleurisy. ] Leukoc 1 32. Xie C, Teixeira LR, McGovern JP, et al. Systemic coccico
Biol. 2000;67: 1 8 9- 1 9 5 . steroids decrease the effectiveness of talc pleurodesis. Am
1 1 4. Bandeira Melo C , Calheiros AS, Silva PM, e t al. Suppressive J Respir Crit Care Med. 1 998; 1 57 : 1 44 1 - 1 444.
effect of distinct bradykinin B2 receptor antagonist on aller 1 33. Dresler CM, Olak J, Herndon JE II, et al. Phase III inter
gen evoked exudation and leukocyte infilcration in sensitized group study of talc poudrage vs talc slurry sclerosis for malig
rats. Br] Pharmacol. 1 999; 1 27:3 1 5-320. nant pleural effusion. Chest. 2005 ; 1 27:909-9 1 5 .
1 1 5 . Bandeira Melo C, Bozza PT, Diaz BL, et al. Cutting 1 34. Ferrer J , Montes JF, Villarino MA, e t al. Influence o f par
edge: lipoxin (DC) A4 and aspirin triggered 1 5 epi LXA4 ticle size on extrapleural talc dissemination after talc slurry
block allergen induced eosinophil trafficking. ] lmmunol. pleurodesis. Chest. 2002; 1 22: 1 0 1 8- 1 027.
2000; 1 64:2267-227 1 . 1 3 5 . Montes JF, Ferrer J, Villarino MA, et al. Influence of talc dose
1 1 6. Pinho V, Oliveira SH, Souza DG, et al. The role of CCL22 on extrapleural talc dissemination after talc pleurodesis. Am
(MDC) for the recruitment of eosinophils during allergic ] Respir Crit Care Med. 2003; 1 68 :348-3 5 5 .
pleurisy in mice. J Leukoc Biol. 2003;73 :356-362. 1 36. Kennedy L, Harley RA , Sahn SA, e t al. Talc slurry
1 1 7. Klein A, Talvani A, Cara DC, et al. Stem cell factor plays a pleurodesis. Pleural fluid and hiscologic analysis. Chest.
major role in the recruitment of eosinophils in allergic pleu 1 99 5 ; 1 07: 1 707- 1 7 1 2 .
risy in mice via the production of leukotriene B4. J Immunol. 1 37. Marchi E , Vargas F S , Acencio M M , e t al. Talc and silver
2000; 1 64:427 1 -4276. nitrate induce systemic inflammatory effects during the
1 1 8 . Klein A, Talvani A, Silva PM, et al. Stem cell factor-induced acute phase of experimental pleurodesis in rabbits. Chest.
leukotriene B4 production cooperates with eotaxin co medi 2004; 1 2 5 :2268-2277.
ate the recruitment of eosinophils during allergic pleurisy in 1 3 8 . Montes JF, Garcia-Valero J, Ferrer J . Evidence of innervation
mice. ] Immunol. 200 1 ; 1 67:524-5 3 1 . in talc-induced pleural adhesions. Chest. 2006; 1 30 :702-709.
1 1 9. Bozza PT, Castro Faria Nero HC, Penido C , e t al. I L 5 1 39. Vargas FS, Teixeira LR, Vaz MAC, et al. Silver nitrate is supe
accounts for the mouse pleural eosinophil accumulation rior to talc slurry in producing pleurodesis in rabbits. Chest.
triggered by antigen but not by LPS . Immunopharmacology. 2000; 1 1 8 : 8 08-8 1 3 .
1 994;27: 1 3 1 - 1 36. 1 40. Vargas F S , Teixeira LR, Anconangelo L , e t al. Experimental
120. Sahn SA, Good JT. The effect of common sclerosing agents on pleurodesis in rabbits induced by silver nitrate or talc: I -year
the rabbit pleural space. Am Rev Respir Dis. 1 98 1 ; 1 24:65-67. follow-up. Chest. 200 1 ; 1 1 9 : 1 5 1 6- 1 520.
1 2 1 . Light RW, Wang NS, Sassoon CS, et al. Comparison of the 1 4 1 . Marchi E, Vargas FS, Acencio MM, et al. Low doses of silver
effectiveness of tetracycline and minocycline as pleural scle nitrate induce pleurodesis with a limited systemic response.
rosing agents in rabbits. Chest. 1 994; 1 06: 577-582. Respirology. 2009; 1 4 : 8 8 5-889.
1 22. Wu W, Teixeira LR, Light RW. Doxycycline pleurodesis in 1 42. Tremblay A, Stather D, Kelly M . Effect of repeated admin
rabbits. Comparison of resulcs with and without chest tube. istration of low-dose silver nitrate for pleurodesis in a rabbit
Chest. 1 998; 1 1 4: 5 63-56 8 . model. Respirology. 20 1 1 ; 1 6 : 1 070- 1 075 .
1 23 . Cheng D - S , Rogers J, Wheeler A , e t al. Th e effects ofintrapleu 143. Guo Y, Tang K, Bilaceroglu S, et al. lodopovidone is as effective
ral polyclonal anti-tumor necrosis factor alpha (TNFa) fab as doxycycline in producing pleurodesis in rabbits. Respirology.
fragments on pleurodesis in rabbits. Lung. 2000; 1 78 : 1 9-30. 20 1 0; 1 5 : 1 1 9-125.
1 24. Bilaceroglu S, Guo Y, Hawthorne ML, et al. Oral forms 1 44. Vargas FS, Wang N-S, Lee HM, et al. Effectiveness of bleo
of tetracycline and doxycycline are effective in producing mycin in comparison co tetracycline as pleural sclerosing
pleurodesis. Chest. 200 5 ; 1 2 8 :3750-3756. agent in rabbits. Chest. 1 993; 1 04 : 1 582- 1 584.
1 2 5 . Ors Kaya S, Bir F, Atalay H, et al. Effect of diclofenac on 145. Marchi E, Vargas FS, Teixeira LR, et al. Comparison of nitro
experimental pleurodesis induced by tetracycline in rabbits. gen mustard, cytarabine and dacarbazine as pleural sclerosing
J lnvestig Med. 2005;53 :267-270. agents in rabbits. Eur Respir J 1 997; I 0:598-602.
CHAPT E R 4 I A N I MAL M O D E LS I N PLE U RAL I NVESTIGAT I O N 51
1 46. Light RW, Wang N-S, Despars ]A, et al. Comparison of 1 60 . Jerram RM, Fossum Tw, Berridge BR, et al. The efficacy of
mitoxantrone and tetracycline as pleural sclerosing agents in mechanical abrasion and talc slurry as methods of pleurode
rabbits. Lung. 1 996; 1 74:373-38 1 . sis in normal dogs. Vet Surg. 1 999;28:322-332.
1 47. Vargas FS, Teixeira LR, Antonangelo L, et al. Acute and 1 6 1 . Cole HG, Russack V, Chiu Y, et al. A comparison of tho
chronic pleural changes after the incrapleural instillation of racoscopic talc insuffiation, slurry, and mechanical abrasion
mitoxantrone in rabbits. Lung. 1 998; 1 76:227-236. pleurodesis. Chest. 1 997; 1 1 1 :442-448.
148. Light RW, Vargas FS. Pleural sclerosis for the treat 1 62 . Whitlow CB, Craig R, Brady K, et al. Thoracoscopic
ment of pneumothorax and pleural effusion. Lung. pleurodesis with minocycline vs talc in the porcine model.
1 997; 1 75 :2 1 3-223. Surg Endosc. 1 996; 1 0 : 1 057- 1 059.
149. Miller EJ, Kajikawa 0, Pueblitz S . Chemokine involve 1 63 . Cohen RG, Shely WW, Thompson SE, et al. Talc pleurode
ment in tetracycline-induced pleuritis. Eur Respir J sis: talc slurry versus choracoscopic talc insuffiation in a por
1 999; 1 4 : 1 387-1 393. cine model. Ann Thorac Surg. 1 996;62: 1 000- 1 002.
1 5 0. Teixeira LR, Wu W, Cheng D-S, et al. The effect of corti 1 64 . Lardinois D, Voge P, Yang L, et al. Non steroidal anti
costeroids on pleurodesis induced by doxycycline in rabbits. inflammatory drugs decrease the quality of pleurodesis
Chest. 2002; 1 2 1 :2 1 6-2 1 9 . after mechanical pleural abrasion. Eur J Cardiothorac Surg.
1 5 1 . Light RW, Cheng D-S, Lee YC, e t al. A single intrapleural 2004;2 5 : 865-87 1 .
injection of transforming growth factor-/32 produces excel 1 6 5 . Lee YCG, Lane KB, Parker RB, et al. Transforming growth
lent pleurodesis in rabbits. Am J Respir Crit Care Med. factor/32 (TGF-/32) produces effective pleurodesis in sheep with
2000; 1 62:98-1 04. no systemic complications. Thorax. 2000; 5 5 : 1 0 5 8- 1 062.
1 52. Lee YCG, Teixeira LR, Devin CJ, et al. Transforming growth 1 66. Sugarmann WM, Widmann WD, Mysh D , et al. Mesh
factor-/32 induces pleurodesis significantly faster than talc. insertion as an aid for pleurodesis. J Cardiovasc Surg.
Am ] Respir Crit Care Med. 200 1 ; 1 63 : 640-644. 1 996;37: 1 73- 1 7 5 .
1 53 . Lee YCG, Devin CJ, Teixeira LR, et al. Transforming growth 1 67. Cetin B, Kockaya EA , Atalay C, e t al. Th e efficacy o f polido
factor /32 induced pleurodesis is not inhibited by corticoste canol in pleurodesis in rats. Surg Today. 2003;33 :688-692.
roids. Thorax. 200 1 ;5 6 :643-648 . 1 68 . Cetin B, Acalay C, Arzu Kockaya E, et al. The efficacy of
1 54. Kalomenidis I, G u o Y, Lane K B , et al. Transforming growth fibrin tissue adhesives in pleurodesis in rats. Exp Lung Res.
factor-/33 induces pleurodesis in rabbits and collagen produc 2005;3 1 :7 1 3-7 1 8 .
tion ofhuman mesothelial cells. Chest. 200 5 ; 1 27 : 1 335-1 340. 1 69 . Werebe EC, Pazecci R , Campos JR, e t al. Systemic distribu
1 5 5 . Idell S, Mazar A, Cines D, et al. Single-chain urokinase alone tion of talc after intrapleural administration in rats. Chest.
or complexed to its receptor in tetracycline-induced pleuritis 1 999 ; 1 1 5 : 1 90- 1 93 .
in rabbits. Am J Respir Crit Care Med. 2002; 1 66:920-926. 1 70 . Fraticelli A , Robaglia-Schlupp A , Riera H, e t al. Distribution
1 56 . Marchi E, Vargas FS, Acencio MM, et al. Evidence that of calibrated talc after intrapleural administration: an experi
mesothelial cells regulate the acute inflammatory response in mental study in rats. Chest. 2002; 1 22 : 1 737- 1 74 1 .
talc pleurodesis. Eur Respir J 2006;28:929-932. 1 7 1 . Kalomenidis I , Lane K, Blackwell TS, et al. Mice
1 57. Guo YB, Kalomenidis I, Hawthorne M, et al. Pleurodesis is are resistant to the induction of a pleurodesis. Chest.
inhibited by anti vascular endothelial growth factor antibody. 2003; 1 24:2407-2408.
Chest. 2005; 1 28 : 1 790- 1 797. 1 72 . Decologne N, Kolb M, Margeccs PJ, et al. TGF-beca l
1 5 8 . Teixeira LR, Vargas FS, Acencio MM, et al. Blockage of vas induces progressive pleural scarring and subpleural fibrosis.
cular endothelial growth factor (VEGF) reduces experimen J Immunol. 2007; 1 79: 6043-605 1 .
tal pleurodesis. Lung Cancer. 2 0 1 1 ;74:392-395. 1 73 . Decologne N , Wettstein G , Kolb M , et al. Bleomycin
1 5 9. Brescicker MA, Oba], LoCicero J III, et al. Optimal pleurode induces pleural and subpleural fibrosis in the presence of car
sis: a comparison study. Ann Thorac Surg. 1 993; 5 5 : 364-366. bon particles. Eur Respir J 2 0 1 0;3 5 : 1 76- 1 8 5 .
Cytok i nes and th e Pleura
Cytokines are soluble peptides secreted by cells that macrophages (3) . IL- 1/3 induces mesothelial cells to
affect the behavior of either the same or nearby cells release plasminogen activator inhibitor type 1 (PAI-
through nonenzymatic means. Often they are glyco 1 ) and transforming growth factor beta (TGF-/3) (4) .
peptides and typically exert their effects at very low The highest levels of pleural fluid IL- 1 (5) and
concentrations in the picomolar to nanomolar range. IL- 1/3 (6) are seen with empyema, which is the pleu
Within this broad class are a number of subclasses, ral disease with the most inflammation. Pleural fluid
including polypeptide growth factors, interleukins IL- 1 and tissue necrosis factor alpha (TNF-a) are sig
(ILs) , interferons, and colony-stimulating factors. nificantly correlated ( 1 ) . The mean levels of IL- 1 are
In recent years, much research has been devoted significantly higher in tuberculous than in malignant
toward understanding the role of these peptides. effusions ( 1 ,7) , but there is sufficient overlap that
It is not surprising, therefore, that there have been IL- 1 levels are not useful diagnostically. In one paper,
numerous reports assessing the diagnostic utility of the pleural fluid levels of IL- 1 were higher than the
the levels of different cytokines in the pleural fluid. simultaneous serum levels, suggesting local produc
In general, with the exception of interferon-gamma, tion ( 1 ) . In another paper (3) the pleural fluid levels
the diagnostic usefulness of cytokine levels in pleu of both IL- 1 a and IL- 1/3 were lower than the simul
ral fluid remains to be demonstrated. Nevertheless, taneously obtained serum levels in patients with both
I have elected to include a chapter on cytokines and malignant and benign pleural effusions. The pleural
the pleura because their study has provided clues about fluid levels of IL- 1/3 are higher in exudates than in
the pathogenesis and resolution of pleural injury. transudates.
Pleural fluid levels of IL- 1 appear to be associated
with loculated pleural effusions and the development
I NTERLE U K I N 1
of pleural fibrosis. Patients with tuberculous pleuritis
IL- 1 has an essential role in T-cell activation and is who develop chronic pleural thickening have signifi
considered a proinflammatory cytokine ( 1 ) . It is a cantly higher IL- 1 levels than those who do not ( 1 ) .
strong immune adjuvant and contributes to the stim Patients with loculated pleural effusions have much
ulation of nonspecific host responses besides promot higher pleural fluid levels of IL- 1/3 than patients with
ing wound healing (2) . It enhances blood flow and nonloculated pleural effusions irrespective of whether
the induction of chemoattractants, which bring to the the effusions are due to malignancy, pneumonia, or
injury sites the key inflammatory cells. The adminis tuberculosis (4) . The levels of IL- 1/3 correlate posi
tration of high doses of IL- 1 to animals produces a tively with the levels ofTGF-/3 and PAI- 1 in the pleu
clinical picture of systemic inflammation that mimics ral fluid, but negatively with the levels of tissue type
septic shock (2) . IL- 1 is composed of two separate plasminogen activator (tPA) (4) .
cytokines, IL- l a and IL- 1{3 (3) . IL- l a is an immu There is a naturally occurring interleukin- I recep
noregulatory cytokine with an essential role in T-cell tor antagonist (IL- I RA) . Marie et al. (8) measured
activation (3) . IL- 1{3 is the predominant form ofIL- 1 the IL- 1 RA levels in the pleural fluid and serum of 24
released upon stimulation o f human monocytes and patients. They found that the mean plasma level of
52
CHAPT E R 5 I CYTO K I N E S A N D T H E PLE U RA 53
IL- l RA was significantly higher than the mean pleu receptor (IL-2R) along with the release o f a shorter
ral fluid level and patients with infections did not soluble form of IL-2R. The median levels of the
have higher levels than those without infections (8) . soluble IL-2R are higher in exudates than in tran
sudates and are higher in tuberculous effusions than
I NTERLE U K I N 2 in malignant pleural effusions, and parapneumonic
pleural effusions, but again there is substantial over
IL-2 plays a crucial role in the mediation of the lap ( 1 5 , 1 7) . The highest pleural fluid IL-2R levels are
immune response. The interaction of IL-2 with seen with rheumatoid pleuritis ( 1 8) .
the IL-2 receptor stimulates a cytokine cascade
that includes various interleukins, interferons, and
TNF-a (2) . IL-2 also induces and maintains the I NTERLE U KI N 3
proliferation of T lymphocytes following mitogen or IL-3 induces the proliferation of eosinophils in vitro
antigen activation and it also induces production of and also prolongs their survival. In patients with eosin
cytotoxic lymphocytes, natural killer (NK) cells, and ophilic pleural effusions, IL-3 appears to help promote
lymphokine-activated killer cells (2) . the eosinophil proliferation and also prolongs the sur
The ability of IL-2 to induce the production of vival of eosinophils. It appears, however, that IL-3 is less
these lymphocytes has led to the evaluation of its important than IL-5 in promoting these two activities
antitumor effects. Intrapleural IL-2 has been used to in patients with eosinophilic pleural effusions. Block
treat malignant pleural effusions because of its abil ing antibodies to IL-5 neutralize more of these activi
ity to induce the production of the various lympho ties than do blocking antibodies to IL-3 ( 1 9) . IL-3 is
cytes (9) . When IL-2 is administered intrapleurally, not detectable by enzyme-linked immunosorbent assay
there are increases in the pleural fluid levels of IL-6, (ELISA) in eosinophilic pleural effusions (20,2 1 ) .
but not ofTNF-a or IL- 1 (9) . When IL-2 is injected
intrapleurally in patients with positive pleural fluid
cytology, the cytology becomes negative within I NTERLE U KI N 4
approximately 1 week ( 1 0) . When mesothelial cells Human immunity has two major components
are incubated with IL-2, there is a significant increase cellular and humoral. The T-helper type 1 (Th l )
in their proliferative response and cytolytic activ pathway favors cellular immunity, whereas the 1h2
ity against autologous tumors ( 1 1 ) . In one study, pathway favors humoral immunity (22) . Early deter
the intrapleural administration of low-dose IL-2, as mination toward Th l and 1h2 cells in the immune
an initial treatment, resulted in an objective clinical response is dependent on the balance between IL- 1 2,
response in 72 of 1 00 patients (72%) with a median which favors the Th 1 response, and IL-4, which favors
duration of 5 months ( 1 2) . the 1h2 response. In a murine model of delayed
There i s one other situation i n which IL-2 is rel hypersensitivity, the subcutaneous administration of
evant to the pleural space. When high-dose intra recombinant murine IL-4 significantly blocked cell
venous IL-2 is used to treat metastatic disease, a trafficking into the pleural space (23) .
capillary leak syndrome develops in some patients In one study on 2 1 patients with malignant pleural
(see Chapter 22) (2) . This syndrome is characterized effusions, IL-4 levels in the pleural fluid were below
by an increase in vascular permeability which fre minimal detectable concentrations (22) . In another
quently results in pleural effusions ( 1 3) . In mice, this paper, the IL-4 pleural fluid levels were measurable
capillary leak syndrome can be prevented by the oral in at least some patients, but there was no significant
administration of a nitric oxide inhibitor ( 1 4) . difference in the levels in patients with tuberculosis
As with IL- 1 , the pleural fluid levels o f IL-2 are or malignancy (24) . However, in a subsequent study,
higher with exudates than with transudates and with IL-4 levels were detectable by cytometric bead array
tuberculous pleuritis than with malignant pleural in almost all the pleural fluids that were tested ( 1 6) .
effusion, but there is much overlap (7, 1 5) . IL-2 levels
tend to be lower in the pleural fluid than in the simul
I NTERLE U KI N 5
taneously obtained serum (3) . The pleural fluid levels
of IL-2 are significantly correlated with the pleural IL-5 is a T-helper 2 cytokine, which is important in the
fluid levels of IL-4, IL-5, IL- 1 0, and TNF-a ( 1 6) . trafficking of eosinophils. IL-5 induces the prolifera
One o f the first events with T-cell activation is the tion of eosinophils in vitro and prolongs their survival.
synthesis and surface expression of an interleukin 2 Pleural fluid IL-5 levels are elevated in patients with
54 P L E U RAL D I S EASES
posttraumatic eosinophilic pleural effusions (20) . The mJection of carrageenan in the mouse, the exuda
pleural fluid from such individuals acts as a stimulus tion and total and differential leukocyte migration in
for eosinophil colony formation, and this stimula both the early and late response are reduced in a dose
tory capability is largely blocked by specific antibodies dependen t and significant manner (30) . The intrapleu
toward IL-5. In like manner, eosinophilic pleural fluid ral injection of IL-6 antibodies 30 minutes before the
enhances the survival of eosinophils, and this capability intrapleural injection of carrageenan decreases both the
is largely blocked by specific antibodies toward IL-5. total and differential leukocyte influx, but significantly
When air is introduced into the pleural space of increases the exudation (30) .
mice, there is a brisk eosinophilic response with a The levels of IL-6 are much higher in the pleural
1 00-fold increase in pleural eosinophils by 1 2 hours, fluid than they are in the serum (3,29,3 1 ) and the
which peaks at 48 hours (25). When IL-5 knockout mean IL-6 levels are much higher in exudates than in
mice are injected with air, the number of eosinophils transudates (6,29,32) . Tuberculous effusions contain
in the pleural lavage increases, but only approximately a significantly higher level of IL-6 than do malignant
1 0 % as much as they did in the wild-type mice (25) . pleural effusions (32) or parapneumonic effusions
In 40 patients with pleural fluid eosinophilia, (6) . The pleural fluid from patients with mesothe
including 30 with more than 1 0% eosinophils, there lioma has a significantly higher mean IL-6 level than
was a significant relationship between the number of does the pleural fluid from patients with adenocarci
eosinophils in the pleural fluid and the pleural fluid noma (33) . It has been postulated that the thrombo
IL-5 level (r = 0.55) (2 1 ) . In a subsequent paper, the cytosis seen in patients with mesothelioma is due to
pleural fluid and serum IL-5 levels were measured the intrapleural production of large amounts of IL-6
in 38 patients with pleural effusions occurring after (33) . The intrapleural administration of IL-2 results
coronary artery bypass graft surgery including 1 3 with in increased pleural fluid levels of IL-6 in malignant
eosinophilic pleural effusions (26) . In this study, the pleural effusions (9) . One study reported that the lev
pleural fluid IL-5 levels significantly correlated with els of soluble IL-6 receptor in the pleural fluid were
the pleural fluid eosinophil counts and the serum lower than those in the serum and were comparable
IL-5 levels significantly correlated with the number of in different diagnostic categories (34) .
blood eosinophils (26) . The pleural fluid IL-5 levels
were significantly higher than the serum IL-5 levels
I NTERLE U KI N 7
(26) . In patients with paragonimiasis, the pleural fluid
IL-5 levels are markedly elevated and correlate with IL-7 was originally discovered as a pre-B-cell growth
the number of eosinophils in the pleural fluid (27) . factor (35) . Soon thereafter, it was found to be a criti
Therefore, it appears that IL-5 is one of the primary cal cytokine for normal T and B lymphopoiesis and
factors responsible for eosinophilic pleural effusions. a mobilizer of pluripotent stem cells and myeloid
It appears that pleural fluid IL-5 is important progenitors. It has also been found to enhance T-cell
in the pathogenesis of malignant pleural effusions functioning and induce cytokine expression in mono
at least in mice. When Lewis Lung Cancer cells or cytes (35). In cell culture, the proliferative response
adenocarcinoma cells are injected directly into the of lymphocytes from malignant pleural effusions is
pleural space of mice, host derived IL-5 promotes the increased significantly more with IL-7 plus IL-2 than
formation of malignant pleural fluid (28) . Knockout with IL- 1 2 plus IL-2. Chen et al (35) concluded that
mice for IL-5 produce much less pleural fluid and IL-7 in the presence of IL-2 could restore the immu
have less tumor growth than wild type mice (28) . nosuppressed cytolytic activity of the lymphocytes of
malignant pleural effusion against autologous tumor.
I NTERLE U K I N 6
To my knowledge, there have been no studies in which
the pleural fluid levels of IL-7 have been measured.
IL-6, also called B-cell stimulatoryJactor-2 or hepatocyte
stimulating growth factor, is a multifunctional cytokine
I NTERLE U KI N 8
produced by several different cell types such as mono
cytes, fibroblasts, and endothelial cells (29) . IL-6 has IL-8 is a powerful neutrophil chemotaxin that con
a pivotal role in many regulatory functions including tributes to the influx of neutrophils into the pleural
maturation of B-cells to antibody-producing cells and space (36-38) . IL-8 is a downstream cytokine to IL- 1
induction of the synthesis of acute phase proteins. and TNF-a. Cultured mesothelial cells produce IL-8
When IL-6 is injected 5 minutes before the intrapleural in response to IL- 1 , TNF-a, or endotoxin (3 5). In
CHAPT E R 5 I CYTO K I N E S AND T H E PLE U RA 55
contrast, antibodies to IL- 1 or TNF-a inhibit IL-8 model o f carrageenan pleurisy, the intrapleural injec
release from mesothelial cells (2) . Mesothelial cells tion of IL- 1 0 5 minutes before the injection of car
produce IL-8 in basal conditions and the produc rageenan led to a significant inhibition of the early
tion is increased if the mesothelial cells are stimu phase (4 hours) but had no significant effect on the
lated with inflammatory stimuli, asbestos fibers, or late phase (48 hours) of the response to carrageenan
infective agents (2) . The pleural fluid IL-8 levels are (30) . In the same model, the administration of anti
higher than the serum IL-8 levels in exudative pleu IL- 1 0 antibody caused a graded and marked increase
ral effusions suggesting that in vivo IL-8 is produced of both total and differential leukocyte influx and also
in the pleural space (3) . IL-8 can also stimulate the increased fluid leakage in the early phase, but had no
growth of certain tumors. In mesothelioma cell lines, effect on the late phase (30) .
IL-8 causes a dose-dependent increase in proliferating Chen et al. (22) measured the pleural fluid and
activity. Lastly IL-8 has angiogenic properties (3) . serum IL- 1 0 in 2 1 patients with a malignant pleural
Pleural fluid IL-8 levels are higher in exudates than effusion. They reported that IL- 1 0 was detectable in
in transudates (6) . Pleural fluid IL-8 levels are most 1 9 of the 2 1 pleural fluids (90%) and that the lev
elevated in patients with empyema (37-40) . There els of IL- 1 0 were higher in the pleural fluid than in
are also relatively high levels of IL-8 in the pleural the serum (22) . However, a second study reported
fluid from patients with cancer or tuberculosis, but that the IL- 1 0 levels were comparable in the pleural
the levels of IL-8 in the pleural fluid from patients fluid and in the serum (3 1 ) . Chen et al (22) found
with congestive heart failure are low (4 1 ,42) . There is no correlation between IL- 1 0 levels and lymphocyte
a significant correlation between the number of neu subpopulations. Aoe et al. ( 1 6) measured the pleural
trophils in empyema fluid and the level of IL-8 in the fluid levels of IL- 1 0 in 93 pleural fluids and found
fluid (36,37) . Neutrophil chemotactic activity is cor detectable levels in 92. The levels were similar in
related with IL-8 activity, and most of the neutrophil patients with malignant effusions, tuberculous pleu
chemotactic activity in pleural fluid is neutralized ritis, and other miscellaneous effusions ( 1 6) . In this
with anti-IL-8 antibodies (36) . study, the pleural fluid IL- 1 0 levels were significantly
IL-8 may also induce lymphocyte chemotaxis for correlated with the levels of IL-2 and IL-4 ( 1 6) .
the pleural space (4 1 ) . Pace et al. (4 1 ) demonstrated
that in patients with malignant and tuberculous pleu
I NTERLE U KI N 1 2
ral effusions, the lymphocyte count was more closely
correlated with the IL-8 level than was the neutro IL- 1 2 is a heterodimeric cytokine composed of two
phil count. Moreover, these workers reported that subunits of molecular masses of 40 kd (p40) and 35 kd
the pleural fluid was chemotactic for lymphocytes (p35) (43) . IL- 1 2 is capable of enhancing cell-mediated
and that the chemotactic activity could be eliminated and cytotoxic immune responses to intracellular patho
with antibodies to IL-8 (4 1 ) . gens and tumors. It is produced primarily by antigen
presenting cells and is considered crucial in promoting
Thl responses and subsequent cell-mediated immu
I NTERLE U KI N 1 0
nity (2) . Knockout mice with deficient IL- 1 2 genes
IL- 1 0 is the most important antiinflammatory cyto have defective cell-mediated immunity, fail to develop
kine found within the human immune response (42) . granulomatous reactions, and are prone to develop
An antiinflammatory cytokine, by definition, is one tuberculosis (2) . When pleural cells are incubated with
that can inhibit the synthesis of IL- 1 , TNF-a, or heat-killed Mycobacterium tuberculosis, IL- 1 2 is detect
other major proinflammatory cytokines (42) . IL- 1 0 able in the supernatants (43) . The addition of anti
is a potent inhibitor o f Th l cell cytokines, includ IL- 1 2 antibodies suppressed proliferative responses of
ing IL-2 and interferon-gamma (42) . In humans, pleural fluid cells to M. tuberculosis by 36% (43) .
the main sources of IL- 1 0 are the lymphocytes and The mean pleural fluid level of IL- 1 2 was approxi
monocytes, but macrophages, mast cells, and eosin mately 1 0 times higher (585 pg/mL) than that in the
ophils also synthesize IL- 1 0 (2) . In a mouse model simultaneously obtained serum when an ELISA was
of hypersensitivity pleuritis, the administration of used that detects both the IL- 1 2 and p40 (43). When
recombinant murine IL- 1 0 before challenge signifi an ELISA was used that detects only heterodimeric
cantly blocked cell trafficking to the pleural cavity IL- 1 2, the mean IL- 1 2 concentration in the pleu
(23) . IL-4 was more potent than IL- 1 0 in block ral fluid in patients with tuberculous pleuritis was
ing this trafficking in this model (23) . In the mouse 165 :±: 28 pg/mL with undetectable levels in serum
56 P L E U RAL D I S EASES
of the same patients or in pleural fluid of patients with endotoxin produce MCP- 1 (39) . Antony et al. (37)
malignancy (43) . Although pleural fluid levels of IL- have shown that the pleural fluid levels ofMCP- 1 are
1 2p40 are higher in tuberculous than in malignant higher in patients with malignant pleural effusions
effusions (44) , they are less efficient at identifying and tuberculous pleural effusions than they are in
tuberculous pleural effusions than are pleural fluid lev patients with parapneumonic effusions or congestive
els of adenosine deaminase (45). In another study on heart failure. There is a correlation between the num
2 1 patients with malignant pleural effusion, pleural ber of monocytes and the pleural fluid MCP- 1 levels
fluid IL- 1 2 was below the minimal detectable concen in patients with malignant pleural effusions. Specific
tration for all serum and pleural fluid samples (22) . It neutralizing antibodies to MCP- 1 eliminate approxi
is interesting to note that rhe intralesional injection mately 70% of the monocyte chemotactic activity in
of IL- 1 2 into mesotheliomas in the mouse model of pleural fluid (37) .
mesothelioma leads to tumor regression (46) .
T U M O R N E C R O S I S FA CTO R a
in mice by the intrapleural injection of Lewis Lung pleural fluid VEGF are correlated significantly with
cancer cells, mice that are given SUI 1 248, a VEGF both the levels of TGF-/3 1 and TGF-/3 2 (63) . In one
inhibitor, have less pleural fluid and less tumor bur study of 83 patients, the levels of VEGF and MMP-9
den (70) . It is unclear as to whether the reduction in were highly correlated (r = 0.88) (76) . The levels of
the fluid volume is due to lessened permeability or VEGF and LDH in the pleural fluid were significantly
a reduction in the size of the tumor (70) . The addi correlated in one study (66) but not in another (75).
tion of VEGF to mesothelioma cell cultures increases It has also been shown that both VEGF mRNA and
their proliferation (7 1 ) . endostatin mRNA are higher in malignant effusions
It appears that VEGF i s important in the pro than in benign effusions and it has been suggested
duction of a pleurodesis. The intrapleural injection that their measurement might assist in the diagnosis
of TGF-/3 in rabbits results in an excellent pleurode of malignant effusions (77) .
sis (72) . However, when anti-VEGF antibodies are
administered before the TGF-/3 is given, the degree of
T U M O R N E C R O S I S FACTO R - a
pleurodesis is markedly diminished (72) . When the
pleurodesis scores are correlated with the angiogenesis Tumor necrosis factor-a (TNF-a) plays a predomi
score in all the rabbits that received either TGF-/3 alone nant role in inflammatory processes. It increases
or TGF-/3 plus anti-VEGF antibodies, there is a close neutrophil margination and activates neutrophils,
correlation between the angiogenesis score in the vis monocytes, macrophages, and eosinophils. It often
ceral pleura and the pleurodesis scores (72) . elicits acute-phase reactions characterized by fever
In a second study, anti-VEGF antibodies were and anorexia (2) . Mesothelial cells produce TNF-ll'
administered intravenously 30 minutes before ani in response to a variety of stimuli. TNF-a is pres
mals received talc or silver nitrate intrapleurally (73) . ent in pleural effusions and levels are significantly
Animals pretreated with anti-VEGF antibody showed higher in exudative than in transudative effusions.
significant reductions in pleural fluid volumes after Pleural fluid TNF-a levels are higher in malignant
talc or silver nitrate injection. IL-8 levels, vascular than in benign effusions but there is considerable
permeability and macroscopic pleural adhesion scores overlap (2) . Anti-TNF-a antibodies inhibit talc
were also reduced in the groups that received the anti pleurodesis (78) .
VEGF antibody (73). These studies demonstrate the
importance of angiogenesis, as promoted by VEGF,
BA S I C F I B R O B LAST G R OWTH FACT O R
in the production of a pleurodesis.
Most VEGF in pleural effusions is believed to Mesothelial cells synthesize and release considerable
originate from local production (2) . VEGF is pro amounts of basic fibroblastic growth factor (b-FGF)
duced by most cell types, but mesothelial cells are (2) . Most b-FGF is intracellular and the remainder is
thought to represent the principal source of pleu associated with the extracellular matrix components
ral fluid VEGF, although infiltrating inflammatory on the mesothelial cell surface (2) . b-FGF stimulates
cells and malignant cells also contribute (2) . VEGF mesothelial cell proliferation in vitro and in vivo (2) .
production can be stimulated by various cytokines Like VEGF, b-FGF is a potent angiogenic factor and
among which TGF-/3 is the most consistent. In vitro is implicated in cancer growth and metastasis (2) .
many other cytokines including IL- 1 , IL-6, TNF-a, Ruiz et al. (79) found that the b-FGF levels were
platelet derived growth factor (PDGF) and platelet below detectable levels (3 pg/mL) in most uncompli
activating factor (PAF) can also stimulate VEGF syn cated parapneumonic, tuberculous, malignant, and
thesis (2) . transudative pleural effusions. In contrast, b-FGF was
There are detectable levels of VEGF in all pleural detectable in most empyemas or complicated parap
effusions (66,70,74) . In patients with exudative pleu neumonic effusions (79) . In an earlier study, Strizzi et
ral effusions, the pleural fluid levels are several times al. (80) reported that the mean pleural fluid b-FGF
higher than the serum values suggesting local produc level in malignant effusions (8.5 ± 6. 1 pg/mL) was
tion (74) . The levels of VEGF are higher in exudates significantly lower than the mean level in nonmalig
than in transudates, but there is significant overlap nant effusions (23 .0 ± 1 9.8 pg/mL) . The explanation
(66,74) . The VEGF levels tend to be higher in malig for the contrasting results in the above two studies is
nant effusions than in tuberculous or parapneumonic not clear. Economidou et al. (8 1 ) reported that the
effusions (75), but there is too much overlap for the levels of b-FGF were comparable in patients with
differences to be useful diagnostically. The levels of transudates and exudates.
CHAPT E R 5 I CYTO K I N E S A N D T H E PLE U RA 59
Levels of MMP- 1 , MMP-2, TIMP- 1 and TIMP-2 in had no active procoagulant. Fibrinolytic activity was
tuberculous pleural fluid are higher than in the simul absent in all the exudates, whereas five of the eight
taneously obtained serum (92) . transudates had fibrinolytic activity (93) . In view of
the above-mentioned observations measurement of the
F I B R I N O G E N E S I S A N D F I B R I N O LY S I S
procoagulant activity in pleural fluids does not appear
I N T H E P L E U RA L S PA C E
to be useful either diagnostically or prognostically.
Since the publication of this study more than
Pleural fibrin deposition is a characteristic of many 20 years ago, there have been several more publica
diseases of the pleura, and fibrin membranes are tions regarding fibrogenesis and fibrinolysis in the
responsible for the loculations that make the drainage pleural space. Lin et al. (94) demonstrated that the
of complicated parapneumonic effusions and empy pleural fluid levels of PAI- 1 and the PAI- 1 /tPA ratio
ema difficult. Fibrin membranes can also produce a were approximately four times higher in patients
loculated pleural effusion in other situations such as with parapneumonic effusions or empyema than in
tuberculosis or malignancy, and the progress of fibrin patients with tuberculous pleuritis. Their tuberculosis
deposition followed by collagen deposition can lead patients with residual pleural thickening had signifi
to chronic pleural thickening. cantly higher PAI- 1 /tPA levels (94) . Chung et al. (95)
When the pleura is inflamed, the amount of fibrin compared the PAI- 1 and the tPA levels in the pleural
that is laid down is the result of the balance between fluid from patients with loculated and nonloculated
fibrinogenesis and fibrinolysis. Thrombin acts on fibrin pleural effusions due to malignancy, tuberculosis, and
ogen to produce fibrin. Fibrogenesis occurs when the pneumonia. As expected, the patients with loculated
factors that favor fibrogenesis such as TNF-a, TGF-{3, pleural effusion had higher levels of pleural fluid
and plasminogen activation inhibitor- I (PAI- 1) are PAI- 1 than did the patients with nonloculated pleural
dominant. Fibrinolysis occurs when more fibrin is being effusions. However, the pleural fluid levels of tPA did
broken down than is being created. Plasminogen breaks not differ significantly between the two groups (95) .
down fibrin and is activated by tPA (55). Human meso In addition, the PAI- 1 /tPA ratio correlated positively
thelial cells express tPA, but no detectable fibrinolytic with the levels ofTNF-a, IL- 1/3 and TGF-{3 1 (95).
activity is found in a fibrin plate assay. The explanation The performance of serial thoracenteses leads
for the lack of fibrinolytic activity appears to be the pro to loculation of the pleural fluid in some patients.
duction of the PAI- 1 . In mesothelial cell cultures, PAI-1 Chung et al. (95) performed thoracentesis daily for
increases in response to both TNF-a and TGF-/3 (55). 3 days in 26 patients with malignant pleural effu
In pleural fluids from humans, patients with tuber sions. They noted, through ultrasound, that on day
culosis have higher levels of PAI- 1 than do patients with 6 after the initial thoracentesis, 1 1 of the patients
pleural malignancy, and the levels in the pleural fluid had fibrous strands in their pleural fluid. (95). When
are approximately four times higher than those in the the PAI- 1 was measured in the pleural fluid, it had
blood ( 1 ) . In contrast, the pleural fluid levels of tPA are increased significantly in the patients who developed
approximately three times higher in patients with malig the fibrous strands but not in those who did not (95).
nancy than in patients with tuberculosis ( 1 ) . This fac All the above facts suggest that the balance
tor possibly explains why there is so much more fibrin between fibrinogenesis and fibrinolysis, as reflected
deposition in patients with tuberculosis than in patients by the PAI- 1 /tPA ratio, is very important in deter
with pleural malignancy. Patients with pleural TB who mining whether a patient will get a loculated pleural
develop residual pleural thickening have significantly effusion or develop residual pleural thickening. Pos
greater pleural fluid levels ofTNF-a and PAI- 1 ( 1 ) . sibly, therapeutic interventions aimed at decreasing
One might hypothesize that the development o f the this ratio will decrease the amount of pleural locula
loculations would depend on the balance between the tion and residual pleural thickening.
procoagulant and the fibrinolytic activity in the pleu
ral space. Idell et al. (93) measured the procoagulant
and fibrinolytic activity in 36 pleural fluids, includ
REFERENCES
ing 2 1 due to malignancy, 3 due to empyema, 8 due
to congestive heart failure, and 4 due to pneumonia. I . Hua CC, Chang LC, Chen YC, et al. Proinflammatory cyto
They found that procoagulant activity was present in kines and fibrinolytic enzymes in tuberculous and malignant
pleural effusions. Chest. 1 999; 1 1 6 : 1 292- 1 296.
some of the exudates, but its presence did not necessar 2 . Lee YCG. Cytokines in pleural diseases. In: Light RW, Lee
ily correlate with loculation and its level did not serve YCG, eds. Textbook of Pleural Diseases. London, England:
to differentiate the different exudates. The transudates Arnold Publishers; 2003 ;63-8 9 .
CHAPT E R 5 I CYTO K l N E S AND THE PLE U RA 61
3. Alexandrakis MG, Coulocheri SA, Bouros D, et al. Evaluation 22. Chen YM , Yang WK, Whang-Peng J , et al. Elevation of
of inflammatory cycokines in malignant and benign pleutal interleukin- I 0 levels in malignant pleural effusion. Chest.
effusions. Oncol Rep. 2000;7: 1 327- 1 332. 1 996; 1 1 0:433-436.
4. Chung CL, Chen CH, Sheu JR, et al. Proinflammatory cyco 23. Fine JS, Rojas-Triana A, Jackson JV, et al. Impairment of leu
kines, transforming growth faccor-beta l , and fibrinolytic kocyce trafficking in a murine pleuritis model by IL-4 and
enzymes in loculated and free-flowing pleural exudates. Chest. IL- 1 0 . Inflammation. 2003;27: 1 6 1 - 1 74 .
2005; 1 2 8 :690-697. 2 4 . Okamoto M, Hasegawa Y, Hara T, et a l . T helper type l /T
5. Silva-Mejias C, Gamboa-Antinolo F, Lopez-Cortes LF, et helper type 2 balance in malignant pleural effusions compared
al. lnterleukin- 1 beta in pleural fluids of different etiolo co tuberculous pleural effusions. Chest. 2005 ; 1 28 :4030-403 5 .
gies. Its role as inflammatory mediator in empyema. Chest. 2 5 . Kalomenidis I, Guo Y, Peebles RS, e t al. Pneumothorax
1 995; 1 08:942-945. associated pleural eosinophilia in mice is interleukin-5 but not
6. Akarsu S, Kure AN, Dogan Y, et al. The differential diagnos interleukin- 1 3 dependent. Chest. 2005 ; 1 28 :2978-2983.
tic values of cycokine levels in pleural effusions. Mediators 26. Kalomenidis I, Stathopoulos GT, Barnette R, et al. Eotaxin-3
Injlamm. 2005;200 5 ( 1 ) :2-8 . and interleukin-5 pleural fluid levels are associated with pleu
7. Shimokata K, Saka H, Murate T, et al. Cytokine content in ral fluid eosinophilia in post-coronary artery bypass grafting
pleural effusion. Chest. 1 9 9 1 ;99: 1 1 03-1 1 07. pleural effusions. Chest. 200 5 ; 1 27:2094-2 1 00.
8 . Marie C, Losser MR, Fitting C, et al. Cycokines and soluble 27. Taniguchi H , Mukae H, Matsumoto N, et al. Elevated
cycokine receptors in pleural effusions from septic and nonsep IL-5 levels in pleural fluid of patients with paragonimiasis
tic patients. Am] Respir Crit Care Med. 1 997; 1 5 6: 1 5 1 5- 1 522. westermani. Clin Exp Immunol. 2000; 1 2 3 : 94-9 8 .
9 . Yanagawa H, Sone S, Munekata M, et al. IL-6 in malignant 28. Stathopoulos GT, Sherrill TP, Karabela SP, et a l . Host
pleural effusions and its augmentation by incrapleural instilla derived interleukin-5 promotes adenocarcinoma-induced
tion of IL-2. Clin Exp Immunol. 1 992;88 :207-2 1 2 . malignant pleural effusion. Am J Respir Crit Care Med.
1 0 . Yanagawa H, Sone S, Nii A , e t al. Lymphokine-activated killer 20 1 0 ; 1 82: 1 273- 1 28 1 .
induction and its regulation by macrophages in malignant 29. Alexandrakis MG, Coulocheri SA, Bouros D , e t al. Evaluation
pleural effusions. Jpn J Cancer Res. 1 989;80: 1 220- 1 227. of ferritin, interleukin-6, interleukin-8 and tumor necrosis
1 1 . Chen YM , Hsieh YL, Tsai CM, et al. lnterleukin-2 stimula factor alpha in the differentiation of exudates and transudates
tion activates mesothelial cellular functioning against aucolo in pleural effusions. Anticancer Res. 1 999; 1 9 :3607-3 6 1 2 .
gous tumor cells. ] Chin Med Assoc. 2004;67:323-330. 30. Frode T S , Souza G E , Calixto JB. The effects of IL-6 and
1 2 . Lissoni P, Mandala M, Curigliano G, et al. Progress report IL- 1 0 and their specific antibodies in the acute inflammatory
on the palliative therapy of 1 00 patients with neoplastic responses induced by carrageenan in the mouse model of pleu
effusions by intracavitary low-dose interleukin-2. Oncology. risy. Cytokine. 2002; 1 7: 1 49- 1 5 6.
200 1 ;60:308-3 1 2 . 3 1 . Shirakabe A, Hara N, Yokoyama S, et al. Cycokine levels in
1 3 . Baluna R , Vitetta ES. Vascular leak syndrome: a side effect of pleural effusions of patients under intensive care. J Nippon
immunotherapy. Immunopharmacology. 1 997;37: 1 1 7- 1 32. Med Sch. 2008;75:262-268.
1 4 . Orucevic A, Lala PK. NG nitro L arginine methyl ester, an 32. Yokoyama A, Maruyama M, Ito M, et al. Interleukin 6 activity
inhibitor of nitric oxide synthesis, ameliorates interleukin 2 in pleural effusion. Chest. 1 992; 1 02 : 1 0 5 5 - 1 0 5 9 .
induced capillary leak syndrome in healthy mice. J Immuno 33. Nakano T, Chahinian A P, Shinj o M, et a l . Interleukin 6 and
ther Emphasis Tumor Immunol. 1 995; 1 8 : 2 1 0-220. its relationship co clinical parameters in patients with malig
1 5 . Harita S, Nogami N, Kikuchi T, et al. Preliminary evaluation nant pleural mesothelioma. Br] Cancer. 1 998 ;77:907-9 1 2 .
of soluble IL-2 receptor and type III procollagen N-terminal 3 4 . Yokoyama A , Kohno N, Fuj ino S, e t al. Soluble inter
aminopeptide in pleural fluid for differentiating tuberculous, leukin-6 receptor levels in pleural effusions. Respir Med.
carcinomatous and parapneumonic pleural effusions. Respirol I 996;90:329-332.
ogy. 2002;7:3 1 1 -3 1 5 . 35. Chen YM, Tsai CM, Whang-Peng J, et al. lnterleukin-7 and
1 6 . Aoe K , Hiraki A , Murakami T, e t al. Relative abundance interleukin- 1 2 have different effects in rescue of depressed cel
and patterns of correlation among six cycokines in pleu lular immunity: comparison of malignant and tuberculous
ral fluid measured by cycometric bead array. Int J Mol Med. pleural effusions. J Interferon Cytokine Res. 200 I ; 2 1 :249-256.
2003; 1 2 : 1 93- 1 9 8 . 36. Broaddus VC, Hebert CA, Vitangcol RV, et al. lnterleu
1 7 . Chang S C , H s u YT, Chen YC, e t al. Usefulness o f soluble inter kin-8 is a major neutrophil chemotactic factor in pleu
leukin 2 receptor in differentiating tuberculous and carcinoma ral liquid of patients with empyema. Am Rev Respir Dis.
tous pleural effusions. Arch Intern Med. 1 994; 1 54: 1 097- 1 1 0 1 . 1 992; 1 46: 825-830.
1 8 . Pettersson T, Soderblom T, Nyberg P, e t al. Pleural fluid solu 37. Antony VB, Godbey SW, Kunkel SL, et al. Recruitment of
ble interleukin 2 receptor in rheumatoid arthritis and systemic inflammatory cells co the pleural space. Chemotactic cyco
lupus erythematosus. ] Rheumatol. 1 994;2 1 : 1 820- 1 824. kines, IL-8, and monocyce chemotactic peptide- I in human
19. Nakamura Y, Ozaki T, Kamei T, et al. Factors chat stimulate pleural fluids. ] Immunol. 1 993; 1 5 1 :72 1 6-7223 .
the proliferation and survival of eosinophils in eosinophilic 3 8 . Miller EJ, Idell S. lnterleukin-8 : a n important neutrophil che
pleural effusion: relationship to granulocyte-/macrophage col motaxin in some cases of exudative pleural effusions. Exp Lung
ony-stimulating factor, interleukin-5, and interleukin-3 . Am J Res. 1 993; 1 9 : 5 89-60 1 .
Respir Cell Mol Biol. I 993;8:605-6 1 1 . 39. Antony VB, Hott JW, Kunkel SL, et al. Pleural mesochelial
20. Schandene L, Namias B , Crusiaux A, et al. IL-5 i n post cell expression of C-C (monocyte chemotactic peptide) and
traumatic eosinophilic pleural effusion. Clin Exp Immunol. C-X-C (interleukin 8) chemokines. Am J Respir Cell Mol Biol.
1 993;93: 1 1 5- 1 1 9 . 1 99 5 ; 1 2 : 5 8 1-588.
2 1 . Mohamed KH, Abdelhamid Al, Lee YC, e t al. Pleural fluid 40. Segura RM, Alegre J, Varela E, et al. lnterleukin-8 and mark
levels of interleukin-5 and eosinophils are closely correlated. ers of neutrophil degranulation in pleural effusions. Am J
Chest. 2002; 1 22: 576-5 80. Respir Cell Mol Biol 1 997; 1 57: 1 565-1 572.
62 P L E U RAL D I S EASES
4 1 . Pace E, Gjomarkaj M, Melis M, et al. !nterleukin-8 induces 6 1 . Sasse SA, Jadus MR, Kukes GD. Pleural fluid transforming
lymphocyte chemotaxis into the pleural space. Role of pleural growth factor-{3 1 correlates with pleural fibrosis in experimen
macrophages. Am J Respir Crit Care Med 1 999; 1 59: 1 5 92-1 599. tal empyema. Am J Respir Crit Care Med. 2003; 1 68 :700-705 .
42. Opal SM, DePalo VA. Anti-inflammatory cytokines. Chest 62. Kunz CR, Jadus MR, Kukes G D , e t al. lntrapleural injection
2000; 1 1 7: 1 1 62- 1 1 72. of transforming growth factor f3 antibody inhibits pleural
43. Zhang M, Gately MK, Wang E, et al. Interleukin 12 at the site fibrosis in empyema. Chest. 2004; 1 26: 1 636-1 644.
of disease in tuberculosis. ] Clin Invest. 1 994;93 : 1 733- 1 739. 63. Cheng D-S, Lee YC, Rogers JT, et al. Vascular endo
44. Yang CS, Lee JS, Lee HM, et al. Differential cyrokine levels thelial growth factor level correlates with transforming
and immunoreactivities against Mycobacterium tuberculosis growth facror-{3 isoform levels in pleural effusions. Chest.
antigens between tuberculous and malignant effusions. Respir 2000; 1 1 8 : 1 747- 1 753.
Med. 2008; 1 02:280-286. 64. Maeda J, Ueki N, Ohkawa T, et al. Transforming growth fac
45. Valdes L, San Jose E, Alvarez Dobano JM, et al. Diagnostic ror-beta I (TGF-bera ! )-and beta 2-like activities in malig
value of lnterleukin 12 p40 in tuberculous pleural effusions. nant pleural effusions caused by malignant mesothelioma or
Eur Respir J 2009;33 : 8 1 6-820. primary lung cancer. Clin Exp Immunl. 1 994;98 :3 1 9-322.
46. Caminschi I, Venetsanakos E, Leong CC, et al. !nterleukin- 1 2 65. Grove CS, Lee YC. Vascular endothelial growth facror: the key
induces an effective antitumor response i n malignant meso mediator in pleural effusion formation. Curr Opin Pulm Med.
chelioma. A m ] Respir Cell Mo! Biol. 1 998; 1 9 :738-746. 2002;8:294-30 1 .
47. Rossi Ml, de Oliveira Barreca E, Pires AL, et al. Long-term 66. Cheng C-S, Rodriguez RM, Perkett EA, et al. Vascular endothe
exacerbation by interleukin 13 of IgE-mediaced eosinophilia lial growth factor in pleural fluid. Chest. 1 999; 1 1 5 : 760-765 .
in rats. Int Immunopharmacol. 200 5 ; 5 : 1 353-1 364. 6 7 . Light RW, Hamm H . Malignant pleural effusion: would the
48. Qin XJ, Shi HZ, Huang ZX, et al. lnterleukin- 1 6 in real cause please stand up? Eur Respir J 1 997; 1 0 : 1 70 1 - 1 702.
tuberculous and malignant pleural effusions. Eur Respir J 68. Zebrowski BK, Yano S, Liu W, et al. Vascular endothelial
2005;25 :605-6 1 1 . growth factor levels and induction of permeability in malig
49. Cheng D-S, Rogers J , Wheeler A, et al. The effects ofintrapleu nant pleural effusions. Clin Cancer Res. l 999;5:3364-3368.
ral polyclonal anti-rumor necrosis facror alpha (TNF-a) fab 69. Yano S, Shinohara H, Herbst RS, et al. Production of experi
fragments on pleurodesis in rabbits. Lung. 2000; 178: 1 9-30. mental malignant pleural effusions is dependent on invasion
50. Odeh M , Sabo E, Srugo I, et al. Correlation between levels of of the pleura and expression of vascular endothelial growth
tumour necrosis facror-alpha and levels of pH, glucose, and factor/vascular permeability factor by human lung cancer
lactate dehydrogenase in parapneumonic effusions. J Infect. cells. Am ] Pathol. 2000; 1 57: 1 893- 1 903.
2005;50: 1 1 4- 1 1 9. 70. Zhu Z, Stathopoulos GT, Dikensoy 0 , et al. The efficacy of
5 1 . Gursel G, Gokcora N, Elbeg S, et al. Tumor necrosis fac SUI 1 248 in inhibition of rumor growth and pleural fluid
ror-alpha (TNF-alpha) in pleural fluids. Tuberc Lung Dis. formation in a mouse malignant pleural effusion model. Proc
1 99 5 ;76: 370-37 1 . Am Thorac Soc. 2005 ;2:A536.
52. Perkett EA. Role o f growth facrors i n lung repair and diseases. 7 1 . Strizzi L, Catalano A, Vianale G, et al. Vascular endothelial
Curr Opin Pediatr. 1 99 5 ;7:242-249. growth factor is an autocrine growth factor in human malig
5 3 . Nasreen N, Mohammed KA, Mubarak KK, et al. Pleural nant mesothelioma. J Pathol. 200 1 ; 1 93 :468-47 5 .
mesothelial cell transformation inro myofibroblasts and hap 7 2 . Guo YB, Kalomenidis ], Hawthorne M, e t a l . Pleurodesis is
rotactic migration in response ro TGF- {beta} I in vitro. Am J inhibited by anti vascular endothelial growth factor antibody.
Physiol Lung Cell Mo! Physiol. 2009;297:L46 1 -L470. Chest. 2005; 1 2 8 : 1 790- 1 797.
54. Lee YC, Lane KB, Zola 0, et al. Transforming growth factor 73. Ribeiro SC, Vargas FS, Antonangelo L, et al. Monoclonal
induces collagen synthesis without inducing IL-8 production anti-vascular endothelial growth factor antibody reduces fluid
in mesothelial cells. Eur Respir J 2003;22 : 1 97-202. volume in an experimental model of inflammatory pleural
55. Idell S, Zwieb C, Kumar A, et al. Pathways of fibrin turnover effusion. Respirology. 2009 ; 1 4 : 1 1 8 8-1 1 93 .
of human pleural mesothelial cells in vitro. Am J Respir Cell 7 4 . Sack U, Hoffmann M, Zhao XJ, e t al. Vascular endothelial
Mo! Biol. 1 992;7:4 1 4-426. growth factor in pleural effusions of different origin. Eur
56. Light RW, Cheng D-S, Lee YC, et al. A single inrrapleural Respir J 2005;25:600-604.
injection of transforming growth factor-2 produces excel 75. Momi H , Matsuyama W, Inoue K, et al. Vascular endothelial
lent pleurodesis in rabbits. Am J Respir Crit Care Med. growth factor and proinflammatoty cytokines in pleural effu
2000 ; 1 62:98-1 04. sions. Respir Med. 2002;96 : 8 1 7-822.
57. Lee YC, Lane KB, Parker RE, et al. Transforming growth fac 76. Jin HY, Lee KS , Jin SM, et al. Vascular endothelial growth
tor beta-2 (TGF-/3 , ) produces effective pleurodesis in sheep factor correlates with matrix metalloproteinase 9 in the pleural
with no systemic complications. Thorax. 2000; 5 5 : I 058-1 062. effusion. Respir Med. 2004;98 : 1 1 5- 1 22 .
58. Lee YCG , Teixeira LR, Devin CJ, et al. Transforming growth 7 7 . Chen Y, Liang B, Zhao YJ, e t a l . Transcription expression
factor-beta(2) induces pleurodesis significantly faster than and clinical significance of vascular endothelial growth factor
talc. Am J Respir Crit Care Med. 200 I ; 1 63:640-644. mRNA and endostatin mRNA in pleural effusions of patients
59. Kalomenidis I , Guo Y, Lane KB, et al. Transforming growth fac with lung cancer. Diagn Cytopathol. 20 1 2;40:287-29 1 .
tor-beta3 induces pleurodesis in rabbits and collagen produc 78. Cheng D-S, Rogers J , Wheeler A, et al. The effects ofinrrapleu
tion of human mesothelial cells. Chest. 200 5 ; 1 27 : 1 335-1 340. ral polyclonal anti-rumor necrosis facror alpha (TNF-a) Fab
60. Lee YCG, Malkerneker D, Thompson PJ, et al. Transforming fragments on pleurodesis in rabbits. Lung. 2000; 1 7 8 : 1 9-30.
growth factor-beta2 induces vascular endothelial growth fac 79. Ruiz E, Aleman C, Alegre J, et al. Angiogenic factors and
tor elaboration from pleural mesothelial cells in vivo and in angiogenesis inhibitors in exudative pleural effusions. Lung.
vitro. Am ] Respir Crit Care Med. 2002; 1 6 5 : 88-94. 200 5 ; 1 83 : 1 8 5 - 1 9 5 .
CHAPT E R 5 / CYTO K l N E S AND THE PLE U RA 63
80. Strizzi L, Vianale G, Catalano A, et al. Basic fibroblast 88. Di Carlo A, Terracciano D, Mariano A, et al. Matrix metal
growth facror in mesothelioma pleural effusions: correla loproteinase-2 and matrix metalloproteinase-9 type IV colla
tion with patient survival and angiogenesis. Int J Oncol. genases in serum of patients with pleural effusions. IntJ Oncol.
200 1 ; 1 8 : 1 093-1 098. 2005;26: 1 363- 1 368.
8 1 . Economidou F, Antoniou KM , Tzanakis N, et al. Angiogenic 89. Park KJ, Hwang SC, Sheen SS, et al. Expression of matrix
molecule Tie-2 and VEGF in the pathogenesis of pleural effu metalloproteinase-9 in pleural effusions of tuberculosis and
sions. Respir Med. 200 8 ; 1 02:774-779 . lung cancer. Respiration. 2005;72: 1 66- 1 7 5 .
82. Kalomenidis I , Kollintza A , Sigala I, e t al. Angiopoietin-2 9 0 . D'Agostino P, Camemi AR, Caruso R , e t al. Matrix metal
levels are elevated in exudative pleural effusions. Chest. loproteinases production in malignant pleural effusions after
2006; 1 2 9 : 1 259-1 266. talc pleurodesis. Clin Exp Immunol. 2003; 1 34: 1 3 8-142.
83. Sumi M, Kagohashi K, Satoh H, et al. Endostatin levels in 9 1 . Iglesias D, Alegre J, Aleman C, et al. Metalloproteinases and
exudative pleural effusions. Lung. 2003; 1 8 1 :329-334. tissue inhibitors of metalloproteinases in exudative pleural
84. Fang F, Chen P, Wu X, et al. Therapeutic effects of recom effusions. Eur Respir]. 2005;25 : 1 04- 1 09.
binant human endostatin adenovirus in a mouse model 92. Hoheisel G, Sack U, Hui DS, et al. Occurrence of matrix metal
of malignant pleural effusion. J Cancer Res Clin Oncol. loproteinases and tissue inhibitors of metalloproteinases in
2009; 1 3 5 : 1 1 49-1 1 57. tuberculous pleuritis. Tuberculosis (Edinb). 200 1 ; 8 1 :203-209.
8 5 . Kimura I, Salramoto Y, Shibasaki M, et al. Release of endothe 93. Idell S, Girard W, Koenig KB, et al. Abnormalities of path
lins and platelet-activating factor by a rat pleural mesothelial ways of fibrin turnover in the human pleural space. Am Rev
cell line. Eur Respir]. 2000; 1 5 : 1 70- 1 76. Respir Dis. 1 9 9 1 ; 1 44 : 1 87- 1 94 .
86. Smit HJ, van den Heuvel MM, Barbierato SB, et al. Analysis 9 4 . L i n F C , Chen YC, Chen FJ, et a l . Cyrokines and fibrinolytic
of pleural fluid in idiopathic spontaneous pneumothorax; cor enzymes in tuberculous and parapneumonic effusions. Clin
relation of eosinophil percentage with the duration of air in Immunol. 200 5 ; 1 28 :690-697.
the pleural space. Respir Med. 1 999;93 :262-267. 95. Chung CL, Chen YC, Chang SC. Effect of repeated thoracen
87. Oda M, Satouchi K, Ikeda I, et al. The presence of platelet teses on fluid characteristics, cyrokines, and fibrinolytic activ
activating factor associated with eosinophil and/or neutro ity in malignant pleural effusion. Chest. 2003; 1 2 3 : 1 1 88-1 1 9 5 .
phil accumulations in the pleural fluids. Am Rev Respir Dis.
1 990; 1 4 1 : 1 469- 1 473.
Rad i ographi c E xaminations
64
C H APTE R 6 / RA D I O G RAPH I C EXA M I N AT I O N S 65
chest radiograph does not indicate that pleural fluid of the lung base and the typical meniscus shape of the
is present. When more fluid accumulates, it spills fluid, as demonstrated in Figure 6. 1 .
over into the posterior costophrenic angle and oblit
erates that sinus as viewed in the lateral projection
Subpulmonic or Intrapulmonary Effusions
(Fig. 6. I B) . The normally sharp posterior costo
phrenic angle is obliterated by a shallow, homoge At times, for unknown reasons, substantial amounts
neous shadow whose upper surface is meniscus of pleural fluid (more than 1 ,000 mL) can be present
shaped. The pleural line up the posterior thoracic wall and may remain in an infrapulmonary location with
is also widened. Anytime the posterior costophrenic out spilling into the costophrenic sulci or extending
angle is obliterated or the posterior part of one or up the chest wall. Such pleural fluid accumulations are
both diaphragms is obscured, the presence of pleural called subpulmonic or inftapulmonary pleural effesions
fluid is suggested and further diagnostic efforts should (Fig. 6.3). Although the posterior costophrenic angle
be made. Moreover, if both posterior costophrenic is usually blunted, at times it is perfectly clear ( 1 ) .
angles are clear and sharp, the presence of clinically Th e following radiologic characteristics are com
significant amounts of free pleural fluid can be nearly mon to most cases of subpulmonic effusions ( 1 ) , and
excluded. the presence of one or more of these characteristics
Increasing amounts of fluid blunt the lateral costo should serve as an indication for decubitus exami
phrenic angle of the posteroanterior (PA) radiograph. nations to rule out the possibility of a subpulmonic
Collins et al. (3) injected fluid into the pleural spaces pleural effusion: (a) apparent elevation of one or both
of upright cadavers. They demonstrated that at least diaphragms; (b) in the posteroanterior projection
1 75 mL pleural fluid had to be injected before the with subpulmonic effusions, the apex of the appar
lateral costophrenic angle was blunted and, in some ent diaphragm is more lateral than usual, near the
cases, more than 500 mL pleural fluid could be pres junction of the middle third and the lateral third of
ent without blunting the lateral costophrenic angle. the diaphragm, rather than at the center of the dia
As more fluid accumulates, the entire outline of the phragm; (c) the apparent diaphragm slopes much
diaphragm on the affected side is lost, and the fluid more sharply toward the lateral costophrenic angle
extends upward around the anterior, lateral, and pos (Fig. 6.4); (d) if the subpulmonic effusion is on the left
terior thoracic walls. This fluid produces opacification side, the lower border of the lung is separated farther
66 P L E U RAL D I S EASES
Diaphragmatic Inversion
At times, the weight of the fluid may cause the dia
phragm to become inverted such that its normally con
vex superior border becomes concave. This inversion
occurs much more commonly with effusions on the left,
but it can occur with effusions on the right (5). Radio
logically, the gastric air bubble is pushed inferiorly, and
the superior border of the diaphragm is concave upward
rather than convex. When viewed under fluoroscopy,
such inverted diaphragms move paradoxically with
respiration, rising on inspiration and descending on
expiration (6). At times, patients with large left pleural
effusions suddenly become dyspneic coincidentally with
the development of inversion of the left diaphragm. In
such instances, therapeutic thoracentesis is indicated (see
Chapter 28) . The removal of some of the pleural fluid
F
restores the normal configuration to the diaphragm and
rapidly relieves the patient's symptoms (6) .
Supine Position
Until this time, I have only discussed the radiologic
characteristics of pleural effusions with the patient
in the upright position. Many chest radiographs,
however, particularly those in acutely ill patients, are
obtained with the patient in the supine position.
When the patient is in this position, pleural fluid
FIGURE 6.2 • D i a g r a m matic exp l a n at i o n for gravitates to the posterior parts of the thoracic cavity.
the m e n iscus s h a p e of p l e u ra l f l u i d . The d i sta n ce Because the pleural fluid is spread over a large area,
between the l u n g a n d the c hest wa l l is the sa m e
considerable quantities must be present before any
a ro u n d the e n t i re l u n g . The d e pth o f the fl u i d
w h e n viewed en face A t>: t o C C ' is n ot suffi cient
radiographic changes are seen. In one study of 1 1 7
to i ncrease the ra d i o d e n s ity. M ore latera l ly at D D ' pleural effusions documented by CT scan, the por
to F F ' , h oweve r, the x-ray bea m passes t h rou g h table anteroposterior radiograph demonstrated only
m o re a n d m o re p l e u ra l f l u i d , s o that a n i ncrease 66% of the effusions including 57% of small, 7 1 % of
in d e n s ity i s rad i o l o g i ca l ly evident. moderate, and 91 % of large-sized effusions (7) . How
ever, in a recent study (8) of 6 1 patients with para
pneumonic effusions, the portable chest radiograph
from the gastric air bubble than usual; normally, the identified 76% of the effusions correctly while the
top of the left diaphragm on the posteroanterior view later radiograph identified 87% and the PA identi
is less than 2 cm above the stomach air bubble (4) ; fied 8 1 % correctly compared with the CT scan. Most
a separation greater than 2 cm suggests a subpul of the patients in whom the pleural effusions were
monic effusion, but, of course, it can also be due to missed had lower lobe infiltrates (8) .
subdiaphragmatic fluid accumulation; if no gastric air The presence of free pleural fluid elicits several signs
bubble is present, the ingestion of a carbonated bev on the supine radiograph. These signs include blunt
erage by the patient will allow evaluation of this sign; ing of the costophrenic angle, increased homogeneous
(e) in the lateral projection, the major fissure often density superimposed over the lung, loss of the hemi
bows anteriorly where it meets the convex upper diaphragm silhouette, apical capping, elevation of
margin of the fluid; a small amount of fluid is usu the hemidiaphragm, decreased visibility of lower lobe
ally apparent in the lower end of the major fissure vasculature, and accentuation of the minor fissure
C H APTE R 6 / RA D I O G RAPH I C EXA M I N AT I O N S 67
(9, 1 0) . Some patients with a small-to-moderate-sized density, blunted costophrenic angle, and loss
pleural effusion have none of these signs. In one of a diaphragm silhouette were the most accurate
study (9) , none of these radiologic signs were present signs in diagnosing a pleural effusion with an accu
in 9 of 1 6 patients with small effusions (defined as racy of approximately 80% ( 1 1 ) . The increased
measuring < 1 . 5 cm on the decubitus radiograph) homogeneous density in most cases with a pleural
and in 3 of 1 3 patients with moderate effusions effusion was limited to the lower one- or two-thirds
(defined as measuring 1 . 5-4.5 cm on the decubitus of the lung field or was more pronounced there ( 1 1 ) .
radiograph) . In a second study, the supine chest radio Th e earliest sign i s blunting o f the costophrenic
graph suggested the presence of pleural Buid in 29 of angle (9) . Subsequently, increased density of the hemi
30 patients (97%) who had more than 300 mL pleu thorax, loss of the hemidiaphragm, and decreased
ral Buid ( 1 1 ) . In this study, increased homogeneous visibility of the lower lobe vasculature occur. Apical
capping does occur with pleural effusion, but it does
not appear to be related to the size of the pleural
effusion (9) . Elevation of the hemidiaphragm and
accentuation of the minor fissure are insensitive signs
in that they occur in a minority of patients and are
not related to the size of the effusion (9) .
Three characteristics serve to differentiate the
increased density due to pleural Buid from that due to
a parenchymal infiltrate. First, if the density is caused
by pleural Buid, the vascular structures of the lung
will be readily visible through the density in a prop
erly exposed film. Any intrapulmonary process that
produces a similar density, however, obliterates the
vascular structure by the "silhouette effect." Second,
if the density is due to pleural Buid, it is usually com
pletely homogeneous. In contrast, infiltrates caused
by intrapulmonary processes are usually less homoge
neous. Third, air bronchograms are present only if the
increased density is due to a parenchymal infiltrate.
FIGURE 6.4 • S u b p u l m o n i c p l e u ra l effu s i o n .
N ote t h a t the r i g h t l atera l costo p h re n i c a n g l e
i s c l e a r, b u t the a pex o f t h e r i g h t d i a ph ra g m
Atypical Effusion
i s m o re l atera l t h a n u s u a l , a n d the a p p a rent The typical arrangement of Buid in the pleural space
d i a p h ra g m s l o pes s h a rp ly towa rd the l atera l
depends on an underlying lung free of disease and
costo p h r e n i c a n g le .
thereby having uniform elastic recoil. If the lung
68 P L E U RAL D I S EASES
underlying the effusion is diseased, the elastic recoil of the lateral view. Fluid encapsulated in a fissure has
the diseased portion is frequently different from that a profile similar to a biconvex lens. Its margins are
of the remainder of the lung, and fluid accumulates sharply defined and blend imperceptibly into interlo
most where the elastic recoil is greatest. Therefore, an bar fissures (Fig. 6.6) . In some situations, the loculated
atypical collection of pleural fluid is an indication of effusion may simulate a mass on the posteroanterior
underlying parenchymal as well as pleural disease. For radiograph. This situation is most frequently seen in
example, if disease in a lower lobe increases its elas patients with congestive heart failure, and because
tic recoil, fluid will collect posteromedially. Accord the fluid absorbs spontaneously when the conges
ingly, in the posteroanterior projection, the opacity is tive heart failure is treated, these fluid collections
higher on the mediastinal than on the axillary border, have been termed vanishing tumors or pseudotumors.
in contrast to the typical appearance in which the The most common location of these "tumors" is in
opacity is higher at the axillary border. Moreover, the the right horizontal fissure ( 1 3) . These pseudotumors
upper surface curves downward and laterally toward are usually less than 4 cm in diameter, but pseudo
the lateral costophrenic sulcus and thereby simulates tumors as large as 1 0 cm in their largest diameter
atelectasis and consolidation of the middle and lower have been reported ( 1 4) . The distinctive configura
lobes. In the lateral projection, the upper border of tion of the loculated interlobar effusion should estab
the density roughly parallels the major fissure, begin lish the diagnosis. The disappearance of the apparent
ning high in the thorax posteriorly and running mass as the effusion resolves definitely establishes the
downward and anteriorly to the anterior costophrenic diagnosis.
sulcus. For the interested reader, Fleischner ( 1 2) has At times, it is difficult to differentiate encapsu
detailed the radiographic appearance of atypical pleu lated fluid in the lower half of a major fissure from
ral fluid accumulation in disease affecting all the indi atelectasis or combined atelectasis and consolidation
vidual lobes. of the right middle lobe. The following three points
help make the distinction ( 1 ) . First, if the minor fis
Loculated Effusion sure is visible as a separate shadow, the diagnosis of
encapsulated fluid is certain. Second, encapsulated
Pleural fluid may become encapsulated by adhesions fluid does not usually obscure the border of the right
anywhere between the parietal and the visceral pleura side of the heart; in contrast, middle lobe atelectasis
or in the interlobar fissures. Because the encapsula almost invariably does. Third, in the lateral projec
tion is caused by adhesions between contiguous pleu tion, loculated effusions usually have a convex border
ral surfaces, it occurs most frequently in association on one or both sides. When the right middle lobe is
with conditions that cause intense pleural inflamma diseased, the borders of the shadow are either straight
tion, such as empyema, hemothorax, or tuberculous or slightly concave.
pleuritis. Loculations occurring between the lung and
the chest wall produce a characteristic radiographic
picture. When viewed in profile (Fig. 6.5), the locu RAD I O LO G I C D O C U M E N TAT I O N
lation is 0-shaped, with the base of the D against
the chest wall and the smooth convexity protruding Most o f the changes discussed i n the previous sec
inward toward the lung because of the compressibil tions are suggestive rather than diagnostic of the pres
ity of the lung parenchyma. If the loculation is in the ence of pleural fluid. For example, blunting of the
lower part of the thoracic cavity, its lower border may posterior or lateral costophrenic angles can be due to
not be visible. Loculation may be differentiated from pleural effusion, but it can also be caused by pleural
parenchymal infiltrates by the absence of air bron thickening or hyperinflation of the lung. Pleural effu
chograms. A definitive diagnosis of loculated pleu sion can obliterate one or both diaphragms on the
ral effusion is best established by ultrasound (see the lateral radiograph, but so can atelectasis or parenchy
discussion on ultrasound) . Because multiple locules mal infiltrates. Therefore, when the posteroanterior
are common, the demonstration of one locule should or the lateral chest radiograph suggests a pleural effu
serve as an indication to search for additional locules. sion, further radiographic studies are usually needed
to document the presence of pleural fluid. Currently,
ultrasound is used most frequently to document the
Loculation in the Fissures
presence of a pleural effusion, but lateral decubitus
The plane of the lung fissures is such that fluid encap radiographs and computed tomography (CT) scans
sulated in the fissure is usually seen in profile in are useful in this situation.
C H APT E R 6 / RA D I O G RAPH I C EXA M I N AT I O N S 69
FIGURE 6 . 5 • Loc u l ated p l e u ra l effu s i o n . A: Poste roa nte r i o r ra d io g r a p h d e m o n strati n g a D-shaped d e n s ity,
with base of the D a g a i nst the r i g h t l atera l ch est wa l l . B: R i g ht l atera l d e c u b itus ra d io g ra p h d e m o n strat i n g
the a bsence o f free p l e u ra l f l u i d i n the s a m e patient. C: C o m p uted tomography (CT) sca n d e m o n strat i n g
p a re n chym a l i nvolve m e n t adjacent to l o c u l ated p l e u ra l effu s i o n . T h i s patient h a s a n a n a e r o b i c i nfect i o n o f
the l u n g a n d p l e u r a l space.
A B
FIGURE 6.6 • Posteroa nte r i o r (A) a n d l atera l (B) ra d i o g r a p h s of a patient with cong estive h e a rt fa i l u re .
A: Two m a ss- l i ke l e s i o n s a re visi b l e i n the l owe r right l u ng fi e l d . B: The b i convex confi g u ration o f l ocu l ated
fl u i d in both the major and the m i n o r fissu res is evident. With treat m e nt of the patie nt's h e a rt fa i l u re, the
l u n g f i e l d s c l e a red, and t h e a p p a rent m a sses d i sappeared. (Courtesy of Dr Harry Sassoon.)
70 P L E U RAL D I S EASES
A B
Pleural fluid on ultrasound appears echo free the visceral and parietal pleura. In one series of 320
(anechoic) , or hypoechoic (reduced echogenicity patients with pleural effusions, 1 72 (54%) were
relative to the liver) (20) . For simplicity, pleural fluid anechoic, 50 ( 1 6%) were complex nonseptated, 76
collections with ultrasound can be characterized as (24%) were complex septated, and 22 (7%) were
echo free (anechoic) , complex septated if there are homogeneously echogenic (22) . Interestingly, all
fibrin strands or septa floating inside the hypoechoic the patients who had complex-nonseptated, com
pleural effusions, complex nonseptated if heteroge plex-septated, or homogeneously echogenic results
neous echogenic material is inside the hypoechoic had exudative pleural effusions. Patients who had
pleural effusion, and homogeneously echogenic if anechoic effusions could have either transudative or
homogeneously echogenic spaces are present between exudative pleural effusions (22) . However, in a more
72 P L E U RAL D I S EASES
recent study 70 of 1 27 patients (55%) with transu in the identification of pleural fluid (28) . With the
dative effusions had a complex nonseptated pattern color Doppler ultrasound, pleural fluid is identified
(24) . because it provides a color signal. In one report of
Transthoracic chest ultrasonography can be per 5 1 patients with minimal pleural effusions, color
formed with any modern ultrasound unit. A 3 . 5 to Doppler ultrasound correctly demonstrated color in
5 . 0 MHz with a small footprint allows visualization 33 of 35 (94%) patients with pleural fluid but was
of the deeper structures, and the sector scan field negative in the 1 6 patients without pleural fluid. In
allows a wider field of view through a small acoustic contrast, real-time gray-scale ultrasound identified
window. Once an abnormality has been identified, a fluid in all 35 patients with fluid but also in 5 of
high-resolution 7 .5 to 1 0 MHz linear probe can be 16 patients without fluid (28) . Another study recently
used to provide detailed depiction of any chest wall, confirmed that color Doppler ultrasound was supe
pleural, or peripheral lung abnormality ( 1 9,20) . The rior to gray-scale ultrasound in demonstrating the
high-frequency transducer improves resolution in presence of pleural effusion (29) .
the near field so that internal echoes in a solid lesion are Ultrasound can demonstrate pleural fluid in
easier to image, which helps the examiner distinguish normal individuals. Kocijancic et al. (30) performed
solid lesions from cystic lesions. In addition, near ultrasonography on a group of 1 06 healthy volun
field reverberation artifacts, which might otherwise teers in the lateral decubitus position and then lean
obscure fluid close to the skin, are reduced. Real-time ing on the elbow. They reported that a layer of pleural
scanning is preferred to conventional static scanning fluid greater than 2 mm in thickness was found in 28
because it allows one to assess the changing configu of the 1 06 volunteers (26%) . The effusion was bilat
ration of pleural fluid with respiration, is easier to use eral in 1 7 volunteers and unilateral in 1 1 (30) . When
in the intercostal spaces, and requires less time to scan the ultrasonography was repeated 2 to 4 months later,
large areas (2 1 ) . The best distinguishing characteristic 2 1 of the original volunteers with fluid still had it,
of a pleural fluid collection on ultrasound is that it whereas an additional 1 1 patients had fluid which
changes its shape with respiration (25). was not present initially (30) .
The posterior chest is best imaged with the patient
sitting upright, whereas the anterior and lateral chest Identifying the Site for Thoracentesis or
are best assessed in the lateral decubitus position. If Thoracoscopy
the patient's arm is raised above his head, the distance
Ultrasonic techniques are useful in identifying the
between the rib spaces is increased and this facili
appropriate site for thoracentesis (2 1 ,22,3 1 ) and
tates scanning the patient in the erect or recumbent
ultrasound guidance is being used more and more
position ( 1 9) .
frequently to identify the site for thoracentesis. The
appropriate site can be identified both in patients
Identifying the Presence of Pleural Fluid
with loculated pleural effusions and in those with
Ultrasound is accurate in identifying the presence small amounts of pleural fluid. In addition to identi
of pleural fluid. In one recent study of 60 patients fying the site for aspiration, the appropriate depth for
with congestive heart failure, CT scans demonstrated aspiration can also be ascertained, thereby increasing
pleural effusions in 52 of the 60 (87%) (26) . Ultra the safety of the procedure. It is important to perform
sound demonstrated pleural fluid in more than 90% the thoracentesis at the time the fluid is identified by
of the patients in whom the fluid was identified by ultrasound. When the skin is only marked at the time
CT scans (26) . In another study, lateral decubitus of the ultrasonic examination and the patient is sent
radiographs were able to identify the presence of back to the ward, the patient is frequently in a dif
pleural fluid in 92% of 52 patients who had small ferent position when the thoracentesis is attempted.
amounts of pleural fluid demonstrated by ultrasound In such instances, the relationship between the skin
(27) . Moreover, there was a close correlation between and the pleural fluid is altered, and the thoracentesis
the thickness of the fluid on the ultrasound examina attempt may occur at the wrong location (3 1 ) . The
tion and the decubitus radiograph (27) . In a recent safety of this method was documented in one retro
study, pulmonologists were able to correctly identify spective review of94 l thoracenteses performed during
the presence or absence of pleural fluid in 95 1 of a 3-year period at a single institution. Of these, only
9 5 5 patients (99.6%) (23) . 24 patients (2. 5 %) developed a pneumothorax and
It has been suggested that color Doppler ultra only 8 (0.8%) required a chest tube (32) . It is prob
sound is superior to real-time gray-scale ultrasound ably safer to perform thoracentesis with ultrasound
C H APT E R 6 / RA D I O G RAPH I C EXA M I N AT I O N S 73
guidance. Barnes et al. (33) , in a retrospective study, patients could be selected for thoracentesis based on
reported that the incidence of pneumothorax and their ultrasonography findings (39) .
chest tube requirement was 4.9% and 0.7% in 305
ultrasonically guided procedures and 1 0.3% and Semiquantitating the Amount of
4 . 1 % in 1 45 thoracenteses done without ultrasound. Pleural Fluid
When ultrasound is used to guide thoracentesis, the
It is possible to estimate, with some degree of reli
lung gliding sign should be assessed pre- and post
ability, the amount of pleural fluid with either ultra
procedure to ascertain whether the procedure resulted
sonography or lateral decubitus chest radiographs.
in a pneumothorax (20) .
Eiben berger et al. (40) studied 5 1 patients who under
In a similar manner, the site for medical thora
went lateral decubitus chest radiography and ultraso
coscopy can be evaluated with ultrasound. Feller
nography while supine. The thickness of the fluid on
Kopman et al. evaluated the use of transthoracic
the lateral decubitus radiograph and on ultrasonogra
ultrasound to locate a safe entry site for trocar place
phy was measured just cranial to the base of the lung.
ment during medical thoracoscopy in 20 patients
Subsequent to these studies, the patients underwent
without induction of a preprocedure pneumotho
therapeutic thoracentesis with removal of all of the
rax (34) . The ultrasound identified entry sites in all
pleural fluid. The relationship between the measure
20 patients including three patients with adhesions
ments and the amount of fluid withdrawn is shown
(34) . These authors felt that the use of ultrasound
in Figure 6.8. It can be seen in the same figure that
may replace the practice of pneumothorax induction
the amount of fluid is more closely correlated with the
before medical thoracoscopy. Ultrasound guidance
ultrasonic measurement than with the lateral decubitus
also facilitates the insertion of chest tubes ( 1 9) and
measurement. Also note that on the lateral decubitus
indwelling pleural catheters such as the Pleurx cathe
radio graph, a fluid thickness of 30 mm corresponds to
ter (3 5). Lastly, ultrasound guidance facilitates needle
a volume of 1 ,000 mL, whereas on the ultrasound
biopsy of the pleura with a standard Abrams needle or
measurement, a fluid thickness of 40 mm corresponds
an automated cutting needle device ( 1 9) . One study
to a volume of 1 ,000 mL. Roch et al. (4 1 ) evaluated
compared the use of ultrasound guided Tru-cut nee
the utility of ultrasound in detecting pleural effusions
dle biopsy and unaided Abrams needle biopsy in the
with volumes greater than 500 mL in patients in the
diagnosis of pleural malignancy and tuberculosis and
ICU. They concluded that if the distance between
reported that the ultrasonically guided procedure had
the lung and the parietal pleura was more than 5 cm at
higher sensitivity and specificity (36) .
the base of the lung, there was a sensitivity of 83% and
a specificity of 90% that the patient had a pleural effu
Ultrasound in the Intensive Care U nit
sion with more than 500 mL (4 1 ) . In a recent study,
The intensive care unit (ICU) is one of the locations Balik et al. (42) found that the amount of pleural
where ultrasound has its greatest utility (37,38) . fluid in 8 1 patients on mechanical ventilation could
Because many chest radiographs are taken with the be predicted by the following formula: volume pleural
patient in the supine position, pleural effusions fluid (mL) = 20 X the separation of the visceral and
and pneumothorax are often missed. Ultrasound can parietal pleura in mm. The mean prediction error was
identify which of the patients in the ICU have pleu 1 58 ± 1 60 mL in this study (42) .
ral effusions or pneumothorax. Identification of the The amount of pleural fluid can also be estimated
presence of significant amounts of pleural fluid in from the CT scan (43) . The volumes estimated from
patients in the ICU is important for at least two rea the CT scan correlate well with the volumes esti
sons. If the patient is on a ventilator, removal of the mated by ultrasound (43) . The amount of pleural
fluid may facilitate weaning. If the patient is febrile, fluid can also be roughly semiquantitated from the
it is important to ascertain whether the patient has posteroanterior and lateral chest radiographs (44) .
infected pleural fluid. In a prospective study, Tu et al. Blackmore et al. (44) have demonstrated that when
(39) performed a 1 -year study of febrile patients more than 50 mL pleural fluid is present, it becomes
with physical, radiographic and ultrasonographic visible on the lateral radiograph as a meniscus pos
evidence of pleural effusion. They performed tho teriorly. When more than 200 mL fluid is present, a
racentesis with ultrasound guidance on 94 patients meniscus can be identified in the lateral costophrenic
and 1 5 ( 1 6%) of them had empyema (39) . All the angle of the posteroanterior radiograph. When more
patients with empyema had a hyperechoic pattern than 500 mL is present, the meniscus obscures the
or a complex-septated pattern. They suggested that entire hemidiaphragm.
74 P L E U RAL D I S EASES
1 00
I I
1 00
0
90 E 90
.s 80
80
"E
E' ( Q)
u
.s 70 E 70
� R
"E .;;....
Q) 60 "' 60
"' 0
::0
E 0 R
/ Q) 0 _,.,.L---" ""
� 50
- 0
E 50
0
n
0 .... " v-
�
"' 0
v
Q) 40
I Au n "- 40
0
n
E 0 v- � 0
v
IA_..... , ;:,�
0 0 0
a: 30
"' 30
( 0 00
0 c
o n �v::.f9no -� no 0 A
v
_J 0
20 - "' 20 ....
0 0
c .§
10
�� A s 10
0 0 c
0
0 0 I I
0 250 500 750 1 000 1 250 1 500 1 750 2000 2250 0 250 500 750 1 000 1 250 1 500 1 750 2000 2250
FIGURE 6.8 • Corre l a t i o n of a ctu a l vo l u m e of p l e u ra l effu s i o n with (A) t h i c k n ess of fl u i d on the l atera l
d e c u b itus ra d i o g ra p h (LDR) a n d with (B) t h i ck n ess of fl u i d on s o n o g ra p hy. Va l u es a re from 5 1 patients before
t h o racentesis. The s o n o g ra p h i c measu re m e nts were m o re closely corre l ated with the vo l u me of the fl u i d
( r 0 . 80) t h a n were the latera l decu b itus m e a s u r e m e nts ( r 0 . 58) . (From Eibenberger KL, Dock WI, Ammann ME, et
= =
al. Quantification of pleural effusions: sonography versus radiography. Radiology. 1 994; 1 9 1 :68 1 -684, with permission.)
Pleural abnormalities can be more readily detected was present in 36% of transudates (54) . An added
and distinguished from lung parenchymal and extra bonus with CT is the clear demonstration of bone
pleural disease by CT than by standard radiographs pathology such as metastases or tuberculosis.
because these anatomic compartments are distinct Findings on CT can help distinguish benign and
on the cross-sectional image with CT (52) . Pleural malignant effusions. In one study of 1 46 patients
collections or masses tend to conform to the pleu with pleural disease, the following findings (with
ral space. As with chest radiographs, the angle of the their sensitivities and specificities) were suggestive of
lesion with the chest wall may help identify whether malignancy: pleural nodularity (37%/97%), pleural
the lesion is pleural or parenchymal. If the angle of rind (22%/97%) , mediastinal pleural involvement
the abnormality with the chest wall is acute, then the (3 1 %/85%) , and pleural thickening greater than
lesion probably has a parenchymal origin, whereas if 1 cm (35%/87%) (55). Traill et al. (56) reported that
the angle is obtuse, the lesion probably has a pleu 27 of 32 patients with malignant pleural effusion had
ral origin. Sometimes, however, the CT findings are pleural nodularity or irregularity or pleural thickness
as ambiguous as the radiographs, particularly when greater than 1 cm whereas none of 8 patients with
there is atelectasis or pneumonia or when a pleural benign disease met these criteria.
collection forms acute angles with the chest wall. Chest CT is not indicated in all patients with sus
Free-flowing pleural fluid produces a sickle pected pleural disease. The density coefficients from
shaped opacity in the most dependent part of the CT are not specific enough to distinguish among
thorax posteriorly (2 1 ) . Loculated fluid collections parenchymal lesions, solid pleural masses, or pleural
are seen as lenticular opacities of fixed position. collections of serous fluid, blood, or pus ( 1 9) . More
When free fluid lies in the posterior costophrenic over, the attenuation coefficients from the CT scan
recess adjacent to the diaphragm, it may be difficult do not differ significantly between transudates and
to differentiate from ascites. Several CT features have exudates (54) . Ultrasonic examinations are preferred
been described that aid in the differentiation of pleu to CT when the primary question is whether pleural
ral fluid from ascites. These are the displaced crus fluid is present.
sign, the interface sign, the diaphragm sign, and the CT examinations of the chest have also provided
bare-area sign (2 1 ) . With the displaced crus sign, the additional information concerning the effects of a
displacement of the diaphragmatic crus away from pleural effusion on the underlying lung. Paling and
the spine by the fluid indicates that the fluid is in Griffin (57) reviewed the chest CT, obtained in
the pleural space. In contrast, ascites lies lateral and the supine position, of 46 cases with a moderate or
anterior to the crus. With the interface sign, a sharp large pleural effusion. The volume of the underlying
interface can be identified between the fluid and the lung, particularly the lower lobe, was reduced in all
liver or spleen, and this interface indicates that asci patients, and there was atelectasis of the underlying
tes is present. This line is much less distinct if pleural lung in 44 of the 46 (96%) patients. In 1 9 patients,
fluid is present. With the diaphragm sign, fluid that there was segmental collapse of the lower lobe. In
is outside the diaphragm is pleural fluid, whereas that seven patients, the atelectatic segment was so large as
which is inside the diaphragm is ascites. With the to produce an appearance initially suggestive of com
bare-area sign, restriction of ascites by the coronary plete collapse of the lower lobe. In 25 of 46 patients,
ligaments from the bare area of the liver indicates the lower lobe collapse involved all except the superior
that the patient has ascites. segment, which tended to remain aerated. Recogni
CT is effective in demonstrating abnormalities in tion of a major degree of volume loss in the lower lobe
the lung parenchyma that are obscured on the con depended on identification of the bronchial anatomy
ventional chest radiograph by the pleural disease. serving the airless lung and on the loss of an iden
Chest CT is particularly useful in distinguishing tifiable inferior pulmonary vein, which was buried
empyema with air-fluid levels from lung abscess, as within the collapsed lung.
discussed subsequently. In patients with pleural effu CT examinations of the chest have also been
sions, CT can also identify pleural thickening, which used to evaluate the major and minor fissures. In
suggests that the patient has an exudative effusion. In one report, 1 00 CT scans of patients with normal
one study, 36 of 59 exudative effusions (6 1 %) had lungs were reviewed to determine the normal charac
associated pleural thickening, whereas only 1 of 27 teristics of the major and minor fissures. Each major
transudates (4%) had associated pleural thickening fissure was imaged most often as a lucent band,
(53) . However, in another study, pleural thickening less often as a line, and least often as a dense band.
76 P L E U RAL D I S EASES
In contrast, the minor fissure was imaged as a lucent mesothelioma, the CT and MRI provide comparable
area, which was usually triangular with its apex at the information in most instances. The MRI is superior to
hilar region (58) . CT in revealing solitary foci of chest wall invasion and
endothoracic fascia involvement, and also in show
ing diaphragmatic muscle invasion (62) . However,
M AG N ETIC R E S O N A N C E I MAG I N G
these findings do not affect surgical treatment (62) . It
Magnetic resonance imaging (MRI) has generated should be noted that many normal individuals have
considerable interest as a safe and sensitive tech a small amount of fluid on MRI. Nguyen et al. (63)
nique for imaging human pathologic conditions. The reviewed 200 normal women who had an MRI for
technique basically consists of inducing transitions breast cancer screening and reported that 17 4 patients
between energy states by causing certain atoms to had a pleural effusion of which 1 24 were bilateral.
absorb and transfer energy. This is accomplished Most of the effusions were less than 7 mm in maximal
by directing a radio frequency pulse at a substance thickness (63) .
placed within a large magnetic field. Measures of In summary, MRI of the chest is currently less
the time required for the material to return to a satisfactory than ultrasound or CT in identifying the
baseline energy state (relaxation time) can be trans presence of a pleural effusion. At present, there are
lated by a complex computer algorithm to a visual no definite clinical indications for MRI of the chest
image. There are two time constants associated with in the management of patients with pleural disease.
relaxation, called T l and T2. The T l relaxation time
characterizes a time constant with which the nuclei
P O S I T R O N E M I S S I O N TO M O G RA P H
align in a given magnetic field. In contrast, T2 reflects
( P E T) S CA N S
the time constant for loss of phase coherence of
excited spins (59,60) . Co-registration o f PET and CT using combined
With MRI, the lungs are seen as regions of black scanners has expanded the use of PET imaging
signal intensity similar to the black appearance of the throughout the body, although its relatively high
lungs on CT. When evaluating the soft tissues, how cost, limited availability, and length of examination
ever, several differences are noted. The subcutaneous time limits it use (64) . Malignant cells concentrate
fat on MRI has bright signal intensity, compared 1 8-fluorodeoxyglucose ( 1 8FDG) more avidly than
with the low signal intensity of CT. On MRI, the normal tissue. However, any area with intense
vascular structures including the aorta and main, left, inflammation is likely to be PET positive. PET scans
and right pulmonary arteries are seen as regions of appear to be useful in the evaluation of pleural effu
signal void (black) that is distinct from the surround sions and pleural thickening or nodules. Orki et al.
ing mediastinal fat. With noncontrast CT, the vessels (65) performed PET scans combined with CT in
and masses have a similar attenuation. The bony 60 patients with pleural effusion and 23 patients with
structures on MRI may be seen as regions of bright pleural thickening before thoracoscopy. The PET
signal intensity because of the fat within the marrow scans were positive in all 44 cases with malignancy
or low intensity for cortical bone (60) . (including 25 cases with malignant mesothelioma)
Pleural effusions can be identified with MRI. A and in two of the 39 benign cases both of whom had
pleural fluid collection is visualized as an area of abnor tuberculosis (65) .
mally low signal intensity on T l -weighted images that
increases in brightness on Tl-weighted images. This
A I R- F L U I D L E V E LS I N T H E
characteristic is consistent with the relatively long T l
P L E U RA L S PA C E
and T 2 values o f pleural fluid. With MRI, different
types of pleural fluid collections such as transudative When both air and fluid are present in the pleural
fluid, chylothorax, hemothorax, or pus may appear space, an air-fluid level is apparent on radiographs
somewhat different, but their characteristics are not obtained in the erect position (Fig. 6.9) . An air
sufficiently distinct as to be diagnostic. A diagnos fluid level is manifested as an absolutely straight line
tic thoracentesis is certainly more definitive and less parallel to the bottom of the radiograph. Of course, if
expensive. the radiograph is obtained with the patient in the
In general, MRI has a limited role in the inves supine position, no air-fluid level will be present
tigation of pleural disease because of poor spatial unless it is a cross-table lateral view. The presence of
resolution and motion artifact (6 1 ) . For patients with an air-fluid level in the pleural space indicates that
C H APT E R 6 / RA D I O G RAPH I C EXA M I N AT I O N S 77
air has gained entry into the pleural space. The differ of the proximal (chest wall-parietal pleura) and the
ential diagnosis includes bronchopleural fistula from distal (visceral pleura-lung) interface occurs when
pulmonary infections, spontaneous pneumothorax the process is in the pleural space. If the process
with pleural effusion or hemothorax, trauma (iatro is within the lung parenchyma, the proximal and
genic or noniatrogenic) , the presence of gas-forming distal interfaces (anterior and posterior walls of the
organisms in the pleural space, and rupture of the caviry) move symmetrically (66) . Other characteristics
esophagus into the pleural space. Air-Buid levels in of pyopneumothorax, but not of a lung abscess, are
the pleural space must be distinguished from air loss of the gliding sign above the air-Buid level and
Buid levels in dilated loops of bowel entering the tho the curtain sign which is movement of the air-Buid
racic caviry through a diaphragmatic hernia. Contrast level synchronized with respiration (67) . In a study
media studies of the gastrointestinal tract are diagnos of 1 6 patients with lung abscess and 1 9 patients with
tic in doubtful cases. pyopneumothorax, Lin et al. (67) reported that the
It is frequently difficult to distinguish a loculated absence of the gliding sign and the curtain sign were
pyopneumothorax with a bronchopleural fistula from the findings which best separated the two entities (67) .
a peripheral lung abscess. This differentiation is impor Recently, it has been shown that color Doppler effec
tant because pyopneumothorax should be treated on an tively distinguishes lung abscess for pyopneumothorax
emergency basis with pleural drainage (see Chapter 1 2), (68) . Chen et al. (68) performed a color Doppler study
whereas a lung abscess usually responds to antibiotics of 34 patients with lung abscess and 30 patient with
and postural drainage alone. One indication that the empyema and demonstrated that if the identification
air-Buid is in the pleural space is whether the length of color Doppler vessel signals in peri-cavitary con
of the air-Buid level varies markedly between the PA solidation was used as an indicator of peripheral lung
and lateral radiographs. A lung abscess is round and abscess, the sensitiviry, specificiry, positive predictive
the length of the air-Buid level is about the same on value, and negative predictive value were 94%, 1 00%,
both the PA and lateral radiographs. 1 00%, and 94%, respectively.
Both ultrasound and CT are useful in distinguishing Empyema with a bronchopleural fistula can also
between these two conditions. With ultrasonic exami be distinguished from a lung abscess by chest CT
nation during hyperventilation, asymmetric motion (69) . With CT scanning, a pyopneumothorax is
78 P L E U RAL D I S EASES
characterized by unequal fluid levels on positional the chest is fluid-filled bullae or lung cysts in which
scanning that closely approximate the chest wall. The the CT findings may resemble those of empyema. On
space characteristically has a smooth, regular mar CT scan, the fluid-filled cavities have many character
gin that is sharply defined without side pockets. The istics of loculated pleural fluid collections including
appearance of the cavity often changes with variations lenticular shape; air-fluid levels of different length on
in the patient's position. In contrast, a lung abscess orthogonal views; uniform, smooth inner walls; and
is typically round with an irregular, thick wall and mass effect on the adjacent lung. Two features are use
has an air-fluid level of equal length in all positions. ful in differentiating fluid-filled cysts from empyema:
When the patient's position is changed, the shapes of (a) cysts tend to be located in the upper lobes and
the cavity and of the mass do not change. Frequently, the air-fluid levels are limited by fissures, and (b) one
multiple side pockets are adjacent to the main cav notes the absence of preexisting or coexisting large
ity. An additional distinguishing feature is that the pleural effusion with fluid-filled cysts (7 1 ) .
larger empyemas displace the adjacent lung and lung
abscesses do not (69) .
MASSIVE EFFUSION
CT is not infallible in distinguishing empyemas
from lung abscesses. Bressler et al. (70) reviewed the When an entire hemithorax is opacified, one should
CT scans from 7 1 patients in whom the question was first examine the position of the mediastinum because
whether the individual had a lung abscess or empy its position is influenced by the pleural pressures
ema. In 5 of the 7 1 patients, the foregoing morpho (Fig. 6. 1 0) . If the pleural pressure is lower on the
logic criteria were insufficient to make the distinction. side of the effusion, the mediastinum will be shifted
The intravenous administration of a bolus of contrast toward the side of the effusion (Fig. 6 . 1 OA) . Alter
medium in conjunction with CT was diagnostically nately, if the pleural pressure is higher on the side of
useful. The demonstration of vessels within a lesion the effusion, the mediastinum will be shifted toward
unequivocally identifies the lesion as parenchymal the contralateral side (Fig. 6. l OB). Of course, if the
rather than pleural. Moreover, after administration mediastinum is invaded by tumor or other infiltrative
of sufficient amounts of contrast material, pulmo processes, it will be fixed, and no shift will be evident
nary parenchyma is enhanced, whereas most pleural on the PA radiograph.
lesions show minimal or no enhancement (70) . When the patient's mediastinum is shifted toward
Another condition that must be differentiated the side of the effusion, the lung underlying the effu
from empyema in a patient with air-fluid levels in sion is usually diseased. In such a case, overexpansion
A B
FIGURE 6 . 1 0 • A: M a ssive p l e u ra l effu s i o n with m a rked sh ift of the trachea a n d m e d i a sti n u m towa rd
the side of the effu s i o n . This patient had a bronchoge n i c carc i n o m a o bstruct i n g the l eft m a i n bro n c h u s .
B: M a ssive p l e u r a l effu s i o n w i t h m a rked sh ift of the trachea a n d m e d i a sti n u m a w a y f r o m the s i d e o f
the effus i o n .
C H APT E R 6 / RA D I O G RAPH I C EXA M I N AT I O N S 79
Asbestos-Induced Thickening
Pleural thickening can also result from asbestos expo
sure (see Chapter 27) . In contrast to other types of
pleural thickening, the parietal pleura rather than
the visceral pleura is thickened following asbestos
exposure. The pleural thickening can either be local
ized, in which case the thickenings are called pleural
plaques, or generalized (77) . An average of 30 years
elapse between the first exposure to asbestos and the
appearance of pleural plaques (77) . The pleural thick
ening or plaques associated with asbestos exposure are
usually bilateral, more prominent in the lower half
of the thorax, and follow the rib contours (78) . The
pleural thickening due to asbestos exposure eventu
ally becomes calcified. The calcification ranges from
small linear or circular shadows, which are usually
situated over the diaphragmatic domes, to complete
encirclement of the lower portion of the lungs. CT FIGURE 6 . 1 1 • Poste roa nte r i o r ra d io g r a p h of
of the chest is more sensitive than other radiologic a pat i e nt with a p n e u mot h o rax on the l eft s i d e .
N ote the o bvi o u s p l e u r a l l i n e (arrows) se p a rat i n g
procedures in identifying both pleural thickening and
the l u n g f r o m the a i r i n the p l e u ra l space. The
pleural calcification due to asbestos exposure (79) . d e nsity of the ra d i o g r a p h i ns i d e and outs i d e the
In obese patients, subcostal fat may mimic pleu p l e u ra l line is s i m i l a r. Also n ote that a b l e b (upper
ral thickening. Typically, it appears as a symmetric, arrow) is p resent a l o n g the su rfa ce of the a p i c a l
smooth, soft tissue density that parallels the chest wall p l e u ra l l i n e . T h i s b l e b w a s p r o b a b l y respo n s i b l e
and is of greatest thickness over the apices. In prob f o r the p n e u m ot h o rax. (Courtesy of D r. Harry Sassoon.)
lem cases, subcostal fat can be distinguished from
either diffuse thickening or localized plaques with
CT. On CT, subcostal fat can be identified as low lung parenchyma from the remainder of the thoracic
density tissue internal to the ribs and external to the cavity, which is clear and devoid of lung markings.
parietal pleura (2 1 ) . Although one might suppose that the partially col
lapsed lung would have increased density radiologi
cally, it does not for the following reasons. First, blood
P N E U M OT H O RAX
flow through the partially collapsed lung, which con
The radiographic signs of a pneumothorax are influ tributes substantially to the density radiologically,
enced by two factors ( 1 ) . First, air in the pleural space decreases proportionately to the degree of collapse.
accumulates in the highest part of the thoracic cav Second, the thorax is a cylinder, and with a pneumo
ity because air is less dense than the lung. Second, thorax, the presence of air both anterior and posterior
the lobes of the lung maintain their traditional shape to the partially collapsed lung decreases the overall
at all stages of collapse. Note that these are the same density of the lung. The radiologic density of the lung
factors that influence pleural fluid accumulation. does not increase until the lung loses approximately
The only difference is that with pneumothorax, air 90% of its volume. Complete atelectasis of a lung due
rises to the apex of the hemithorax and causes early to pneumothorax is characterized ipsilaterally by an
collapse of the upper lobes of the lung, whereas with enlarged hemithorax, a depressed diaphragm, a shift
pleural effusion, the pleural fluid falls to the bottom of the mediastinum to the contralateral side, and a
of the hemithorax and collapses the lower lobes. fist-sized mass of increased density at the lower part of
the hilum representing the collapsed lung (Fig. 6. 1 2) .
The diagnosis o f pneumothorax is usually easily
Radiologic Signs
established by demonstrating the visceral pleural line
A definitive radiologic diagnosis of pneumothorax on the posteroanterior radiograph. With small pneu
can be made only when a visceral pleural line can be mothoraces, however, the visceral pleural line may not
identified (Fig. 6. 1 1 ) . The visceral pleural line is evi be visible on the routine radiographs. In such cases,
dent as a faint but sharply defined line separating the one of the following two procedures can establish
C H APT E R 6 / RA D I O G RAPH I C EXA M I N AT I O N S 81
FIGURE 6. 1 3 • Anteroposte r i o r ra d i o g r a p h of
a patient with a p n e u m ot h o rax in the r i g h t a pex
a n d a s k i n fo l d in the l eft a pex. On the s i d e with
the p n e u m ot h o rax, n ote the sharp b l ack-wh ite
i nterface m ed i a l to the a p p a rent p l e u ra l l i n e . I n
FIGURE 6 . 1 2 • Poste roa nte r i o r rad i o g r a p h of
contra st, note that o n t h e s i d e with t h e s k i n fo l d
a pat i e nt with a p n e u mot h o rax a n d co m p l ete
(l eft s i d e), t h e re i s g ra d u a l fa d i n g from the wh ite
atelectas i s of the right l u n g .
p l e u ra l l i ne a s one m oves m e d i a l ly.
the diagnosis: (a) radiographs can be obtained in In one series, the pneumothorax could not be iden
the upright position in full expiration; the rationale tified on the lateral projection in 1 3 of 1 22 patients
is that, although the volume of gas in the pleural ( 1 1 %) (8 1 ) . When the pneumothorax is identifi
space is constant, with full expiration, the lung vol able, the displaced pleural line is usually anteriorly
ume is reduced, and therefore, the percentage of the or posteriorly located rather than at the lung apex.
hemithorax occupied by air increases, making iden In 1 0% of the patients, an air-fluid level was the only
tification of the visceral pleural line much easier; and recognizable finding of a pneumothorax on the lateral
(b) radiographs can be obtained in the lateral decu projection (8 1 ) .
bitus position, with the side of the suspected pneu Skin folds may superficially mimic a pleural line
mothorax superior; the free air in the pleural space and possibly lead to a misdiagnosis of pneumothorax.
rises, increasing the distance between the lung and A skin fold results in an abnormal edge with a sharp
the chest wall; additionally, fewer conflicting shadows black-white interface laterally, with gradual fading of
are seen over the lateral chest wall than at the apex. It the density from white to black medially (Fig. 6. 1 3) .
appears that the decubitus position is the most sen In contrast, there is n o such fading medial to the line
sitive for detecting a pneumothorax. Carr et al. (SO) with pneumothorax. In addition, lung markings are
obtained conventional chest radiographs and CT seen peripheral to the edge of the skin fold, in con
scans on cadavers in which varying amounts of intra trast to absence of lung markings peripheral to the
pleural air had been introduced. They found that the line of a pneumothorax (82) .
lateral decubitus film was most sensitive (88%) for It is much more difficult to establish the diagnosis
the diagnosis of pneumothorax, followed by the erect of pneumothorax on a supine radiograph. In a review
(59%) and supine (37%) views. These researchers of 88 critically ill patients with 1 1 2 cases of pneumo
reported that the pneumothorax was always detected thorax, 30% of the cases were not initially detected
in the lateral decubitus position when there was by the radiologist, and half these patients progressed
more than 40 mL of intrapleural air. In addition, they to tension pneumothorax (82) . The most common
found that a CT scan was no more sensitive than the location for collections of air on the supine film is
decubitus views (80) . the anteromedial location because this area is the least
Pneumothoraces are more difficult to recognize on dependent pleural recess. The three other locations
lateral projections than on posteroanterior projections. in which air collects on the supine radiograph are
82 P L E U RAL D I S EASES
the increased pleural pressure prevents the shift of the 2 0 . Koenig SJ, Narasimhan M, Mayo P H . Thoracic ulcraso
mediastinum toward the involved side on inspiration, nography for the pulmonary specialise. Chest. 20 1 1 ; 1 40:
1 332- 1 34 1 .
as occurs with a nontension pneumothorax, and the 2 1 . McLoud TC, Flower CD. Imaging the pleura: sonography,
movement of the ipsilateral diaphragm is restricted CT, and MR imaging. A]R Am ] Roentgenol. 1 9 9 1 ; 1 56:
( 1 ) . In general, however, it is better to insert a needle 1 1 45-1 1 53 .
into the pleural space to ascertain the presence of a 2 2 . Yang P C , Luh KT, Chang D B , e t al. Value o f sonography
in determining the nature of pleural effusion: analysis of
tension pneumothorax than to waste time with radio
320 cases. A]R A m j Roentgenol. 1 992; 1 5 9:29-33.
logic procedures (see Chapter 24) (88) . 23. Rahman NM, Singanayagam A, Davies HE, et al. Diagnostic
accuracy, safety and utilisation of respiratory physician
delivered thoracic ulcrasound. Thorax. 2 0 1 0;65 :449-453.
REFERENCES 24. Chen HJ, Tu CY, Ling SJ, et al. Sonographic appearances in
cransudative pleural effusions: not always an anechoic pattern.
I . Fraser RS, Muller NL, Colman N, et al. Diagnosis of Diseases Ultrasound Med Biol. 2008;34: 362-369.
of the Chest, 4th ed, Vol. I. Philadelphia, PA: WB Saunders; 2 5 . Marks WM, Filly RA, Callen PW Real-time evaluation of
1 999:563-5 94. pleural lesions: new observations regarding the probability of
2 . Davis S, Gardner F, Qvist G . The shape of a pleural effusion. obtaining free fluid. Radiology. 1 982; 1 42: 1 63- 1 64.
Br] Med. 1 963 ; 1 :436-437. 26. Kataoka H, Takada S . The role of thoracic ultrasonography
3 . Collins JD, Burwell D, Furmanski S, et al. Minimal detectable for evaluation of patients with decompensated chronic heart
pleural effusions. Radiology. 1 972; 1 0 5 : 5 1-53. failure. ] Am Coll Cardiol. 2000;3 5 : 1 638-1 646.
4. Vix VA. Roentgenographic recognition of pleural effusion. 27. Kocijancic I, Vidmar K, lvanovi-Herceg Z. Chest sonography
JAMA. 1 974;229:69 5-698 . versus lateral decubitus radiography in the diagnosis of small
5 . Wang JS, Tseng CH. Changes in pulmonary mechanics and pleural effusions. ] Clin Ultrasound. 2003;3 1 :69-74.
gas exchange after choracentesis on patients with inversion of 28. Wu R-G, Yuan A, Liaw Y-S, et al. Image comparison of real
a hemidiaphragm secondary to large pleural effusion. Chest. time gray-scale ultrasound and color Doppler ultrasound for
1 995; 1 07: 1 6 1 0- 1 6 1 4. use in diagnosis of minimal pleural effusion. Am J Respir Crit
6. Mulvey RB. The effect of pleural fluid on the diaphragm. Care Med. 1 994; 1 5 0 : 5 1 0-5 1 4 .
Radiology. 1 965;84: 1 080- 1 0 8 5 . 29. Kalokairinou-Motogna M, Maratou K , Paianid I, e t al. Appli
7. Kicazono MT, Lau CT, Parada AN, e t al. Differentiation cation of color Doppler ultrasound in the study of small pleural
of pleural effusions from parenchymal opacities: accuracy effusion. Med Ultrason. 20 1 0 ; 1 2: 1 2- 1 6 .
of bedside chest radiography. A]R Am J Roentgenol. 20 1 O; 30. Kocijancic K , Kocijancic I, Vidmar G. Sonography o f pleural
1 94:407-4 1 2. space in healthy individuals. J Clin Ultrasound. 2005;33:
8. Brixey AG, Luo Y, Skouras V, et al. The efficacy of chest radio 386-389.
graphs in detecting parapneumonic effusions. Respirology. 3 1 . Lomas DJ, Padley SG, Flower CD. The sonographic appear
20 1 1 ; 1 6: 1 000-1 004. ances of pleural fluid. Br J Radio/. 1 993;66 : 6 1 9-624 .
9 . Ruskin JA, Gurney JW, Thorsen MK, et al. Detection of pleu 32. Jones PW, Moyers JP, Rogers JT, et al. Ultrasound-guided
ral effusions on supine chest radiographs. A]RAm J Roentgenol. choracencesis. Is it a safer method? Chest. 2003; 1 23 :
1 9 87; 1 48:68 1 -683. 4 1 8-423.
10. Woodring JH. Recognition of pleural effusion on supine 33. Barnes Tw, Morgenthaler TI, Olson EJ, et al. Sonographi
radiographs: how much fluid is required? A]RAmj Roentgenol. cally guided choracentesis and race of pneumochorax. J Clin
1 984; 1 42 : 5 9-64. Ultrasound. 2005;33 :442-446.
1 1 . Emamian SA, Kaasbol MA, 0 lsen JF, et al. Accuracy of the 34. Hersh CP, Feller-Kopman D, Wahidi M, et al. Ultrasound guid
diagnosis of pleural effusion on supine chest x-ray. Eur Radio/. ance for medical choracoscopy: a novel approach. Respiration.
1 997;7:57-60. 2003;70:299-30 1 .
12. Fleischner FG. Atypical arrangement of free pleural effusion. 3 5 . Tremblay A, Michaud G . Single-center experience with 250
Radio/ Clin North Am. 1 963; 1 :347-362. tunnelled pleural catheter insertions for malignant pleural
13. Higgins JA, Juergens JL, Bruwer AJ, et al. Loculated interlo effusion. Chest. 2006; 1 29:362-368.
bar pleural effusion due to congestive heart failure. Arch Intern 36. Chang D-B, Yang P-C, Luh K-T, et al. Ultrasound-guided
Med. 1 9 5 5 ;96: 1 80- 1 87. pleural biopsy with Tm-Cuc needle. Chest. 1 99 1 ; 1 00 :
14. Haus BM, Stark P, Shofer SL, et al. Massive pulmonary pseu 1 328- 1 333.
dotumor. Chest. 2003; 1 24:758-760 . 37. Scefanidis K, Dimopoulos S, Nanas S . Basic principles and
1 5 . Moskowitz H, Platt RT, Schachar R , et a l . Roentgen visualiza current applications of lung ulcrasonography in the intensive
tion of minute pleural effusion. Radiology. 1 973; 1 09:33-3 5 . care unit. Respirology. 20 1 1 ; 1 6:249-256.
1 6 . Kocijancic I, Tercelj M , Vidmar K , e t al. The value of 3 8 . Xirouchaki N, Magkanas E, Vaporidi K, et al. Lung ultra
inspiratory-expiratory lateral decubitus views in the diagnosis sound in ill patients: comparison with bedside chest radiogra
of small pleural effusions. Clin Radio/. 1 999;54:595-597. phy. Intensive Care Med. 20 1 1 ;37: 1 488-1 493.
17. Mathis G . Thoraxsonography-part I: chest wall and pleura. 39. Tu CY, Hsu WH, Hsia TC, et al. Pleural effusions in
Ultrasound Med Biol. 1 997;23: 1 1 3 1 - 1 1 39. febrile medical ICU patients: chest ultrasound study. Chest.
18. Diacon AH, Theron J, Bolliger CT. Transchoracic ultrasound for 2004; 1 26: 1 274- 1 280.
the pulmonologist. Curr Opin Pulm Med. 2005; 1 1 :307-3 1 2 . 40. Eibenberger KL, Dock WI, Ammann ME, et al. Quantifi
1 9. Koh DM, Burke S, Davies N, et al. Transthoracic US of the cation of pleural effusions: sonography versus radiography.
chest: clinical uses and applications. Radiographies. 2002;22:E I . Radiology. 1 994; 1 9 1 :68 1-684.
84 P L E U RAL D I S EASES
4 1 . Roch A, Bojan M , Michelet P, et al. Usefulness of ultraso 63. Nguyen J, Nicholson BT, Patrie JT, et al. Incidental pleural
nography in predicting pleural effusions > 5 0 0 ml in patients effusions detected on screening breast MRI. AJR Am j Roent
receiving mechanical ventilation. Chest. 2005 ; 1 27:224-232. genol. 20 1 2 ; 1 99:Wl 42-W l 4 5 .
42. Balik M, Plasil P, Waldauf P, et al. Ultrasound estimation of 6 4 . Ayres J, Gleeson F. Imaging of rhe pleura. Semin Respir Crit
volume of pleural fluid in mechanically ventilated patients. Care Med. 2 0 1 0 ; 3 1 :674-688.
Intensive Care Med. 2006;24 : 1 -4. 6 5 . Orki A, Akin 0 , Tasci AE, et al. The role of positron emission
43. Remerand F, Dellamonica J, Mao Z, et al. Multiplane ultra tomography/computed tomography in rhe diagnosis of pleu
sound approach ro quantify pleural effusion at the bedside. ral diseases. 1horac Cardiovasc Surg. 2009 ;5 7 : 2 1 7-22 1 .
Intensive Care Med. 2 0 1 0;36:656-664. 66. Adams FV, Kolodny E . M-mode ultrasonic localization and
44. Blackmore CC, Black WC, Dallas RV, et al. Pleural fluid identification of fluid-containing pulmonary cysts. Chest.
volume estimation: a chest radiograph prediction rule. Acad 1 979;75 :330-333.
Radio!. 1 996;3 : 1 03- 1 09 . 67. Lin FC, Chou CW, Chang SC. Differentiating pyopneumo
4 5 . Zhu Z, Donnelly E, Dikensoy 0 , e t al. Efficacy of ultra thorax and peripheral lung abscess: chest ultrasonography. Am
sound in the diagnosis of pleurodesis in rabbits. Chest. j Med Sci. 2004;327:330-335.
2005 ; 1 28 : 934-939. 68. Chen HJ, Yu YH, Tu CY, et al. Ultrasound in peripheral
46. Dikensoy 0 , Zhu Z, Donnelly E, et al. Combination pulmonary air-fluid lesions: color Doppler Imaging as an
therapy with intrapleural doxycycline and talc in reduced aid in differentiating empyema and abscess. Chest. 2009;
doses is effective in producing pleurodesis in rabbits. Chest. 1 3 5 : 1 426- 1 432.
2005; 1 28:373 5-3742. 69. Pugatch RD, Spirn PW. Radiology of the pleura. Clin Chest
47. Leo F, Dellamonica J, Venissac N, et al. Can chest ultrasonog Med. 1 985;6: 1 7-32.
raphy assess pleurodesis after VATS for spontaneous pneumo 70. Bressler EL, Francis IR, Glazer GM, et al. Bolus contrast
rhorax? Eurj Cardiothorac Surg. 2005;28 :47-49. medium enhancement for distinguishing pleural from paren
48. Ma OJ, Mateer JR, Ogata M, et al. Prospective analysis of chymal lung disease: CT features. j Comput Assist Tomogr.
a rapid trauma ultrasound examination performed by emer 1 987; 1 1 :436-440.
gency physicians. j Trauma. 1 995;38 :879-8 8 5 . 7 1 . Zinn WL, Naidich DP, Whelan CA, et al. Fluid within pre
49. Sisley AC, Rozycki GS, Ballard R B , e t al. Rapid detection of existing pulmonary air-spaces: a potential pitfall in the CT
traumatic effusion using surgeon-performed ultrasonography. differentiation of pleural from parenchymal disease. j Comput
j Trauma. 1 998 ;44:29 1 -296. Assist Tomogr. 1 987; 1 1 :44 1 -448.
5 0 . Abboud PA, Kendall J . Emergency department ultrasound 72. Light RW, Jenkinson SG, Minh V, et al. Observations on pleu
for hemothorax after blunt traumatic injury. j Emerg Med. ral pressures as fluid is withdrawn during thoracentesis. Am
2003;25: 1 8 1 -1 84. Rev Respir Dis. 1 980; 1 2 1 :799-804.
5 1 . Rowan KR, Kirkpatrick AW, Liu D, et al. Traumatic pneu 73. Maher GG, Berger HW Massive pleural effusion: malignant
mothorax detection with rhoracic US: correlation with and non-malignant causes in 46 patients. Am Rev Respir Dis.
chest radiography and CT-initial experience. Radiology. 1 972; 1 05:45 8-460 .
2002;225 : 2 l 0-2 1 4 . 7 4 . Liberson M . Diagnostic significance of the mediastinal pro
52. Henschke C l , Yankelevitz OF, Davis SD. Pleural diseases: file in massive unilateral pleural effusions. Am Rev Respir Dis.
multimodality imaging and clinical management. Curr Prob 1 963;88 : 1 76- 1 80 .
Diagn Radio!. 1 99 1 ;20: 1 5 5-1 8 1 . 7 5 . Yu CJ, Yang P C , Wu HD, et a l . Ultrasound study i n unilateral
5 3 . Aquino SL, Webb WR, Gushiken BJ. Pleural exudates and hemithorax opacification. Image comparison with computed
transudates: diagnosis with contrast-enhanced CT. Radiology. tomography. Am Rev Respir Dis. 1 993; 1 47:430-434.
1 994; 1 92:803-808. 76. Renner RR, Markarian B, Pernice NJ, et al. The apical cap.
54. Abramowitz Y, Simanovsky N, Goldstein MS, et al. Pleural Radiology. 1 974; 1 1 0: 5 69-573.
effusion: characterization with CT attenuation values and CT 77. Hillerdal G. Non-malignant asbestos pleural disease. 1horax.
appearance. AJR Am j Roentgenol. 2009; 1 92:6 1 8-623. 1 9 8 1 ;36:669-675 .
5 5 . Yilmaz U, Polar G, Sahin N, et al. CT in differential diagnosis 7 8 . Fletcher DE, Edge J R . Th e early radiological changes in
of benign and malignant pleural disease. Monaldi Arch Chest pulmonary and pleural asbestosis. Clin Radio!. 1 970;2 l :
Dis. 2005;63 : 1 7-22. 3 5 5-365 .
56. Traill ZC, Davies RJ, Gleeson FV. Thoracic computed romog 7 9 . Katz D, Kreel L. Computed tomography i n pulmonary asbes
raphy in patients with suspected malignant pleural effusions. tosis. Clin Radio!. 1 979;30:207-2 1 3 .
Clin Radio!. 200 1 ; 56: 1 93- 1 96. 80. Carr JJ, Reed J C , Choplin RH, e t al. Plain and computed
57. Paling MR, Griffin GK. Lower lobe collapse due co pleural effu radiography for detecting experimentally induced pneumo
sion: a CT analysis. ] ComputAssist Tomogr. 1 985;9: 1 079-1 083. thorax in cadavers: implications for detection in patients.
58. Proco AV, Ball JB Jr. Computed tomography of the major and Radiology. 1 992; 1 83 : 1 93- 1 99.
minor fissures. AJR Am j Roentgenol. 1 983; 1 40:439-448 . 8 1 . Glazer HS, Anderson DJ, Wilson BS, et al. Pneumothorax:
5 9 . Gamsu G, Sostman D . Magnetic resonance imaging of the appearance on lateral chest radiographs. Radiology. 1 989;
thorax. Am Rev Respir Dis. 1 98 9 ; 1 39:254-274. 1 73 :707-7 1 1 .
60. Fisher MR. Magnetic resonance for evaluation of the thorax. 82. Buckner CB, Harmon BH, Pallin JS. Th e radiology o f abnor
Chest. 1 989;9 5 : 1 66- 1 73 . mal intrathoracic air. Curr ProblDiagn Radiol. 1 9 88; 1 7:37-7 1 .
6 1 . Evans AL, Gleeson FV. Radiology in pleural disease: state of 8 3 . Wernecke K . Sonographic features o f pleural disease. AJR Am
the art. Respirology. 2004;9 :300-3 1 2 . j Roentgenol. 1 997; 1 6 8 : 1 06 1 - 1 066.
6 2 . Heelan RT, Rusch VW, Begg C B , e t al. Staging o f malignant 84. Lichtenstein DA, Menu Y. A bedside ultrasound sign ruling
pleural mesothelioma: comparison of CT and MR imaging. out pneumothorax in the critically ill. Lung sliding. Chest.
AjR Am j Roentgenol. 1 999; 1 72 : 1 039-1 047. 1 9 9 5 ; 1 0 8 : 1 345-1348.
C H APT E R 6 / RA D I O G RAPH I C EXA M I N AT I O N S 85
8 5 . Lichtenstein D, Meziere G, Biderman P, et al. The comet-tail 87. Bourgouin P, Cousineau G, Lemire P, e t al. Computed tomog
artifact: an ultrasound sign ruling out pneumothorax. Intensive raphy used to exclude pneumothorax in bullous lung disease.
Care Med. 1 999;25:383-388. ] Can Assoc Radio!. 1 9 85 ;36:34 1 -342.
86. Murphy M , Nagdev A, Sisson C. Lack of lung sliding on ultra 88. Light RW Tension pneumothorax. Intensive Care Med.
sound does not always indicate a pneumothorax. Resuscitation. 1 994; 1 06: 1 1 62- 1 1 6 5 .
2008;77:270.
C l i n i cal M an i festati ons and U seful Tests
Normally, the pleural space contains only a few indicates that the parietal pleura is probably involved
milliliters of pleural fluid. If fluid in the pleural space and that the patient has an exudative pleural effusion.
is detected on a radiologic examination, it is abnor Ordinarily, the pain associated with pleural disease
mal. Many conditions can be associated with pleural is well localized and coincides with the affected area
fluid accumulation (Table 8 . 1 ) . When pleural fluid of the pleura because the parietal pleura is innervated
is detected, an effort should be made to determine mostly by the intercostal nerves. At times, however,
which of the many conditions listed in Table 8 . 1 is pleuritic pain is referred to the abdomen because
responsible. In this chapter, the clinical manifesta intercostal nerves are also distributed to the abdomen.
tions of pleural effusions are first discussed. Then, A notable exception to the localization of the pain
the various tests used in the differential diagnosis of occurs when the central portion of the diaphragmatic
pleural effusions are reviewed. In Chapter 8, recom pleura is involved. The nerve supply to this portion
mendations are given for a systematic approach to the of the parietal pleura is the phrenic nerve; therefore,
patient with an undiagnosed pleural effusion. inflammation of the central portion of the diaphragm
is referred to the tip of the ipsilateral shoulder. Pleu
C L I N I CA L M A N I F E STAT I O N S ritic pain felt simultaneously in the lower chest and
ipsilateral shoulder is pathognomonic of diaphrag
The presence o f moderate-to-large amounts o f pleural matic involvement.
fluid produces symptoms and characteristic changes A second symptom of pleural effusion is a dry,
on physical examination. nonproductive cough. The mechanism producing
the cough is not clear, although it may be related to
Symptoms
pleural inflammation. Alternately, lung compression
The symptoms of a patient with a pleural effusion are by the fluid may bring opposing bronchial walls into
mainly dictated by the underlying process causing contact, stimulating the cough reflex.
the effusion. Many patients have no symptoms refer The third symptom of pleural effusion is dyspnea.
able to the effusion. When symptoms are related to A pleural effusion acts as a space-occupying process in
the effusion, they arise either from inflammation of the thoracic cavity and therefore reduces all subdivi
the pleura, from compromise of pulmonary mechan sions of lung volumes. Small-to-moderate-sized pleural
ics, from interference with gas exchange, or, on rare effusions displace rather than compress the lung and
occasions, from decreased cardiac output. A pleural have little effect on pulmonary function (2) . Larger
effusion associated with pleuritic chest pain indicates pleural effusions obviously cause a significant reduction
inflammation of the pleura, specifically, the pari in lung volumes, but the improvement in pulmonary
etal pleura as the visceral pleura does not have pain function following therapeutic thoracentesis is much
fibers. Some patients with pleural effusions experi less than what one would anticipate. We obtained
ence a dull, aching chest pain rather than pleuritic spirometry before and 24 hours after thoracentesis in
chest pain. This symptom is very suggestive that the 26 patients from whom a mean of 1 ,740 mL pleural
patient has pleural malignancy ( 1 ) . The presence of fluid was withdrawn (3) . In these patients, the mean
either pleuritic chest pain or dull, aching chest pain vital capacity improved 4 1 0 ± 390 mL. Patients in this
86
C H APT E R 7 I C L I N ICAL M A N I F E STAT I O N S A N D U S E F U L TE STS 87
study, with higher pleural pressures after rhe removal lung from the chest wall, tactile fremitus is absent or
of 800 mL pleural fluid and patients with smaller attenuated because the fluid absorbs the vibrations
decreases in the pleural pressure after the removal of emanating from the lung. Tactile fremitus is much
800 mL pleural fluid, had greater improvements in rhe more reliable than percussion for identifying both
forced vital capacity (FVC) after thoracentesis. the upper border of the pleural fluid and the proper
Associated parenchymal disease probably explains site to attempt a thoracentesis. With a thin rim of
this small increase in pulmonary function following fluid, the percussion note may still be resonant, but
rherapeutic rhoracentesis. The degree of dyspnea is the tactile fremitus is diminished. Palpation may also
frequently out of proportion to rhe size of the pleu reveal that the cardiac point of maximum impulse is
ral effusion. Often, rhis feature is the result of com shifted to one side or the other. With large left pleural
promised diaphragmatic function due to the weight effusions, the cardiac point of maximum impulse may
of fluid on rhe diaphragm. At times, rhe diaphragm not be palpable. In patients with pleural effusions, the
becomes inverted and this usually results in dispropor position of the trachea should always be ascertained
tionate dyspnea (4) . Either pleuritic chest pain, with because it indicates the relationship between the
rhe resultant splinting, or concomitant parenchymal pleural pressures in the two hemithoraces.
disease can also be responsible for rhe disproportionate The percussion note over a pleural effusion is dull
dyspnea. When the pleural effusion is large, ventricular or flat. The dullness is maximum at the lung bases
filling may be impeded, leading to decreased cardiac where the thickness of the fluid is the greatest. As
output and dyspnea (5) . Arterial blood gases usually mentioned earlier, however, the percussion note may
remain at clinically acceptable levels whatever the size not be duller if only a thin rim of fluid is present.
of rhe effusion (6) because of the reflex reduction in Light percussion is better than heavy percussion for
perfusion to the lung underlying the effusion. identifying small amounts of pleural fluid. If the dull
ness to percussion shifts as the position of the patient
is changed, one can be almost certain that free pleural
Physical Examination
fluid is present (9) .
When a patient presents with chest symptoms, the Auscultation over the pleural fluid characteristi
physical examination is useful in suggesting whether cally reveals decreased or absent breath sounds. Near
pleural fluid is present (7) . In such an examination, the superior border of the fluid, however, breath
attention should be paid to the relative sizes of the sounds may be accentuated and may take on a bron
hemithoraces and the intercostal spaces. If the pleu chial characteristic. This phenomenon has been
ral pressure is increased on the side of the effusion, attributed to increased conductance of breath sounds
that hemithorax will be larger, and the usual concav through the partially atelectatic lung compressed by
ity of the intercostal spaces will be blunted or even the fluid ( 1 0) . This accentuation of breath sounds
convex. In contrast, if the pleural pressure on the side does not mean that an associated parenchymal infil
of the effusion is decreased, as with obstruction of trate is present. Auscultation may also reveal a pleural
a major bronchus or a trapped lung, the ipsilateral rub. Pleural rubs are characterized by coarse, creak
hemithorax will be smaller, and the normal concavity ing, leathery sounds most commonly heard during
of the intercostal spaces will be exaggerated. In addi the latter part of inspiration and the early part of
tion, with inspiratory efforts, the intercostal spaces expiration, producing a to-and-fro pattern of sound.
retract. Enlargement of the hemithorax with bulging Pleural rubs, caused by the rubbing together of the
of the intercostal spaces is an indication for thera roughened pleural surfaces during respiration, are
peutic thoracentesis to relieve the increased pleural often associated with local pain on breathing that
pressure. Signs of decreased pleural pressure are a subsides with breath-holding. Pleural rubs often
relative contraindication to therapeutic thoracente appear as pleural effusions diminish in size, either
sis because the decreased pleural pressure can lead spontaneously or as a result of treatment, because the
to reexpansion pulmonary edema (8) . Of course, in pleural fluid is no longer present between the rough
many patients with pleural effusions, the hemitho ened pleural surfaces.
races are equal in size and the intercostal spaces are It is important to realize that an elevated hemidia
normal. phragm can produce all the classic physical findings
Palpation of the chest in patients with pleural effu associated with a pleural effusion. Obviously, rhe chest
sions is useful in delineating the extent of the effusion. is not the only structure that should be examined when
In areas of the chest where pleural fluid separates the evaluating a patient with a pleural effusion; clues to
88 P L E U RAL D I S EASES
the origin of the effusion are often present elsewhere. or an exudate. If the effusion is a transudate, no fur
The effusion is probably due to congestive heart failure ther diagnostic pleural procedures are necessary, and
(CHF) if the patient has cardiomegaly, neck vein dis therapy is directed to the underlying CHF, cirrhosis,
tension, or peripheral edema. Signs of joint disease or or nephrosis. Alternately, if the effusion proves to be
subcutaneous nodules suggest that the pleural effusion an exudate, a more extensive diagnostic investigation
is due to rheumatoid disease or lupus erythematosus. is indicated to delineate the cause of the effusion. It
An enlarged, nontender nodular liver or the presence has been shown that pulmonary specialists are not very
of hypertrophic osteoarthropathy suggests metastatic accurate at doing this on the basis of clinical history,
disease, as do breast masses or the absence of a breast. physical examination, and radiographic findings ( 1 3) .
Abdominal tenderness suggests a subdiaphragmatic For many years, a pleural fluid protein level of
process, whereas tense ascites suggests cirrhosis and a 3.0 g/dL was used to separate transudates from exu
hepatothorax. Lymphadenopathy suggests lymphoma, dates, with exudative pleural effusions characterized
metastatic disease, or sarcoidosis. by a protein level above 3.0 g/dL ( 1 4, 1 5) . Use of
this one simple test led to the misclassification of
approximately 1 0 % of pleural effusions ( 1 4- 1 6) .
S E PA RAT I O N O F TRAN S U DAT I V E
Light e t al. subsequently demonstrated that with the
F R O M E X U DAT I V E E F F U S I O N S
use of simultaneously obtained serum and pleural
Th e accumulation o f clinically detectable quantities fluid protein and lactic acid dehydrogenase (LDH)
of pleural fluid is distinctly abnormal. A diagnostic values, 99% of pleural effusions could be correctly
thoracentesis (see Chapter 28) should be attempted classified as either transudates or exudates ( 1 6) . Exu
whenever the thickness of pleural fluid on ultrasound dative pleural effusions meet at least one of the fol
or the decubitus radiograph is greater than I 0 mm lowing criteria, whereas transudative pleural effusions
or whenever loculated pleural fluid is demonstrated meet none (Light's criteria) :
with ultrasound unless the etiology of the effusion is
known. A properly performed diagnostic thoracen 1 . Pleural fluid protein divided by serum protein
tesis takes less than I 0 minutes and should cause no greater than 0 . 5
more morbidity than a venipuncture. The informa 2. Pleural fluid LDH divided b y serum LDH
tion available from examination of the pleural fluid is greater than 0.6
invaluable in the management of the patient. 3. Pleural fluid LDH greater than two thirds of the
Pleural effusions have classically been divided into upper limit of normal serum LDH
transudates and exudates ( 1 1 ) . A transudative pleural
effusion develops when the systemic factors influenc Subsequent studies have demonstrated that Light's
ing the formation or absorption of pleural fluid are criteria classify virtually all exudates correctly but
altered so that pleural fluid accumulates. The pleural misclassify about 25% of transudates as exudates.
fluid is a transudate. The fluid may originate in the In recent years, other tests have been proposed for
lung, the pleura, or the peritoneal cavity ( 1 2) . The the separation of transudates from exudates. The tests
permeability of the capillaries to proteins is normal that have been proposed to indicate a pleural exudate
in the area where the fluid is formed. Examples of have included a pleural fluid cholesterol greater than
conditions producing transudative pleural effusions 60 mg/dL ( 1 7, 1 8), a pleural fluid cholesterol greater
are left ventricular failure producing increased pul than 45 mg/dL ( 1 9), a gradient of less than 1 .2 g/dL
monary interstitial fluid and a resulting pleural effu for the difference in the pleural fluid and serum albu
sion, ascites due to cirrhosis with movement of fluid min level (20), a pleural fluid-to-serum bilirubin ratio
through the diaphragm, and decreased serum oncotic above 0.6 (2 1 ) , a high pleural fluid viscosity (22), a high
pressure with hypoproteinemia. level of oxidative stress markers (23), soluble leukocyte
In contrast, an exudative pleural effusion devel selectin (24) , cytokines (25), uric acid (26), and a pleu
ops when the pleural surfaces or the capillaries in the ral fluid-to-serum cholinesterase ratio above 0.23 (27) .
location where the fluid originates are altered such Two subsequent reports (28,29) have compared
that fluid accumulates. The pleural fluid is an exu Light's criteria with some of the other proposed tests
date. The most common causes of exudative pleural and have concluded that Light's criteria best separate
effusions are pleural malignancy, parapneumonic exudates and transudates. In the study of Romero et al.
effusions, and pulmonary embolism. (28) of297 patients including 44 transudates and 253
The first question to ask in assessing a patient with a exudates, Light's criteria were superior to cholesterol
pleural effusion is whether that effusion is a transudate measurement in making the distinction. In this study
C H APT E R 7 I C L I N ICAL M A N I F E STAT I O N S A N D U S E F U L TE STS 89
with Light's criteria, 98% of the exudates and 77% effusions correctly than did the protein gradient. It is
of the transudates were correctly classified (28) . In a suggested that the protein gradient first be examined
subsequent study of 393 patients including 1 23 with because it is already available from Light's criteria. If
transudates and 270 with exudates from South Africa the protein gradient is not definitive, then one may
(29), Light's criteria were found to be superior to the use the albumin gradient or the NT-pro-BNP.
serum effusion albumin gradient, the effusion cho In the discussion in the preceding text, pleural
lesterol concentration, and the pleura fluid-to-serum effusions have been dichotomized into transudates
bilirubin ratio (29) . Again in this study, Light's cri or exudates on the basis of a single cutoff point. An
teria identified 98% of the exudates correctly, but alternative approach is to use likelihood ratios for
they were less accurate in identifying transudates, identifying whether a pleural fluid is a transudate or
misclassifying 1 9 of 1 1 2 ( 1 7%) (29) . Two additional an exudate (36,37) . The idea behind this approach is
studies (30, 3 1 ) have come to similar conclusions. that the higher a value, for example, the pleural fluid
It is unlikely that the pleural fluid cholesterol mea LDH, the more likely the effusion is to be an exudate
surement will provide additional information to the and the lower the value, the less likely the effusion is
ratio of the pleural fluid to the serum protein because to be an exudate. Heffner et al. have derived multi
the pleural fluid cholesterol level can be accurately level (36) and continuous (37) likelihood ratios for
predicted from the serum cholesterol and the ratio of the usual biochemical tests used to differentiate tran
the pleural fluid to the serum protein level (32) . sudates and exudates. When these likelihood ratios
The number of false positives and false negatives are used in conjunction with pretest probabilities
with any test depends upon the cutoff level chosen for using Bayes' theorem, posttest probabilities can be
the identification of an exudate. If a high cutoff level derived (37) . Difficulties in using this approach occur
is chosen, all transudates will be identified correctly, because the pretest probabilities vary significantly
whereas if a low level is chosen, all exudates will be from physician to physician, and most physicians
identified correctly. Light et al. originally developed do not understand the mathematics involved. This
Light's criteria with the goal to identify all exudates approach does emphasize that it is important to take
correctly, and the criteria are remarkably effective in into consideration the absolute value of the measure
achieving this goal. ments. Very high or very low values are almost always
An alternative approach is to select the cutoff level indicative of exudates and transudates, respectively,
that will correctly identify the highest percentage of whereas values near the cutoff levels can be associated
patients. Using this approach, Heffner et al. (33) ana with either transudates or exudates.
lyzed the data from eight studies with a total of 1 ,448 The following approach is recommended for deter
patients and concluded that the best cutoff levels mining whether a pleural effusion is a transudate or
for the different pleural fluid tests were as follows: an exudate. First assess the fluid with Light's criteria.
protein ratio 0.5, pleural fluid LDH 0.45 of the upper The higher the value for the protein ratio, the LDH
limits of the normal for serum, LDH ratio 0.45. ratio, and the absolute value of the LDH, the more
As demonstrated in the preceding text, Light's likely the fluid is an exudate. If the fluid meets the
criteria identify approximately 25% of transudative criteria for a transudative effusion, it is a transudate. If
effusions as exudates. This mislabeling occurs most the fluid meets the criteria for an exudative effusion by
commonly when patients with CHF are treated with only a small margin and the clinical picture is compat
diuretics before thoracentesis is performed (34) . ible with a transudative effusion, measure the protein
These mislabeled transudates barely meet the exuda gradient between the serum and pleural fluid. If this
tive criteria. The protein ratio is less than 0.65, the value is greater than 3 . 1 g/dL, then the fluid can be
LDH ratio is less than 1 .0 and the level of the LDH relabeled a transudate. An alternative approach is to
is less than the upper limit of normal. How can these measure the brain natriuretic peptide (BNP) level in
mislabeled transudates be identified? One possible the pleural fluid or the serum. If this is greater than
means is to examine the gradient between the serum 1 ,500 pg/ mL, the diagnosis of CHF is established (see
and the pleural fluid protein levels. If this gradient is the discussion on NT-pro-BNP later in this chapter) .
greater than 3 . 1 g/dL, one can presume that the fluid
is actually a transudate (34) . In a previous edition of
Specific Gravity
this book, it was recommended that an albumin gra
dient of 1 .2 g/dL (29) rather than the protein gradi The specific gravity of the pleural fluid as measured
ent of 3 . 1 g/dL be used. Bielsa et al. (3 5) reported with a hydrometer was used in the past to separate
that the albumin gradient identified more of these transudates from exudates (38) because it was a simple
90 P L E U RAL D I S EASES
and rapid method of estimating the protein content no single predominant cell type (4 1 ) . The pleural fluid
of the fluid. A specific gravity of 1 .0 1 5 corresponds glucose level is similar to the serum glucose level, and
to a protein content of 3 g/dL, and this value was the pleural fluid amylase level is low (42) . The pleural
used to separate transudates from exudates (38) . At fluid pH with transudative pleural effusions is higher
the present time, the specific gravity of pleural fluid than the simultaneously obtained blood pH (43),
is usually measured with a refractometer rather than probably because of active transport of bicarbonate
a hydrometer. Unfortunately, the scale on the com from the blood into the pleural space (44) .
mercially available refractometers is calibrated for the
specific gravity of urine rather than for pleural fluid.
Probrain Natriuretic Peptide ( B N P)
A reading of 1 .020 on the urine specific gravity scale
corresponds to a pleural fluid protein level of 3 g/dL. The levels ofNT-pro-BNP and BNP in the pleural fluid
However, there is a scale on the same refractometer for are useful in establishing that the etiology of the pleural
protein levels that is valid for pleural fluid. Because the effusion is CHE When the ventricles are subjected to
only reason to measure specific gravity is to estimate increased pressure or volume, these peptides are released
the protein level, and because the pleural fluid specific into the circulation (45). The biologically active BNP
gravity measurement is extraneous and confusing, it and the larger amino terminal part NT-pro-BNP
should no longer be ordered (39) . However, a rapid are released in equimolar amounts into the circulation
estimate of the pleural fluid protein content can be (45). The serum levels of BNP are used to help establish
obtained at the patient's bedside with the protein scale the diagnosis of CHE In clinical practice, levels above
on the refractometer (39) . 500 pg/mL are considered diagnostic of CHF whereas
levels below 1 00 pg/mL are thought to make the diag
nosis of CHF unlikely (46) .
Other Characteristics of Transudates
Poree! et al. (47) first demonstrated that the pleural
Most transudates are clear, straw colored, nonviscid, fluid levels of NT-pro-BNP are elevated in patients with
and odorless. It takes a pleural fluid RBC count of heart failure. They measured NT-pro-BNP levels in 1 1 7
more than 1 O,OOO/mm3 to give the pleural fluid a pleural fluid samples with the following diagnoses: CHF
pinkish tinge. Approximately 1 5% have RBC counts in 44 samples, malignancy in 25, tuberculous pleuritis
above this level. Therefore, the discovery of blood in 20, hepatic hydrothorax in 1 0, and miscellaneous in
tinged pleural fluid does not mean that the fluid 1 8. The mean NT-pro-BNP fluid level in the CHF
is not a transudate. Because RBCs contain a large patients (6,93 1 pg/mL) was significantly higher than
amount of LDH, one might suppose that the LDH the 5 5 1 pg/mL in the patients with hepatic hydrotho
level in a blood-tinged or bloody transudative pleural rax and the 292 pg/mL in the patients with exudative
effusion would be so elevated that it would meet the pleural effusions (45) . When a cutoff level of 1 ,500 pg/
criteria for an exudative pleural effusion. Such does mL was used, the sensitivity was 9 1 % and the specificity
not appear to be the case, however. The LDH isoen was 93% for the diagnosis of CHE We have compared
zyme present in RB Cs is LD H- 1 , and in one study of the pleural fluid NT-pro-BNP levels in 1 0 patients each
23 patients with bloody pleural effusions (pleural with effusions due to CHF, pulmonary embolism, post
fluid red cell counts greater than 1 OO,OOO/mm3) , coronary artery bypass surgery, and malignancy (48) .
the fraction o f LDH-1 i n the pleural fluid was only All the patients with CHF had NT-pro-BNP levels
slightly increased (40) . above 1 ,500 pg/mL, and none of the other patients had
The pleural fluid white blood cell (WBC) count of NT-pro-BNP levels this high (48).
most transudates is less than l ,OOO/mm3, but approxi It should be emphasized that the serum or pleural
mately 20% have WBC counts that exceed l ,OOO/mm3 fluid BNP and NT-pro-BNP cannot be used inter
(4 1 ) . Pleural fluid WBC counts above 1 0,000/mm3 are changeably in the diagnosis of pleural effusions due to
rare with transudative pleural effusions. The differen CHF (49). The BNP levels are only about 1 0 % of the
tial WBC count in transudative pleural effusions may NT-pro-BNP levels. There is not a close correlation
be dominated by polymorphonuclear leukocytes, small between the BNP levels and the NT-pro-BNP levels
lymphocytes, or other mononuclear cells. In a series ( r = 0.78) (50) . Moreover, the diagnostic usefulness
of 47 transudative effusions, 6 ( 1 3%) had more than of the NT-pro-BNP in making the diagnosis of heart
50% polymorphonuclear leukocytes, 1 6 (34%) had failure is superior to that of the BNP (50,5 1 ) . The
predominantly small lymphocytes, 22 (47%) had pre accuracy of NT-pro-BNP in making the diagnosis of
dominantly other mononuclear cells, and 3 (6%) had pleural effusion due to heart failure was attested to
C H APT E R 7 I C L I N ICAL M A N I F E STAT I O N S A N D U S E F U L TE STS 91
in a meta-analysis of 1 0 studies with a total of 1 , 1 20 Turbid pleural fluid can occur from either increased
patients in which the pooled sensitivity and specific cellular content or increased lipid content. These two
ity were 94% and 94%, respectively (52) . entities can be differentiated if the pleural fluid is cen
The pleural fluid NT-pro-BNP is also superior trifuged and the supernatant examined. If turbidity
to the BNP and the protein gradient in identifying remains after centrifugation, it is in all probability due
patients with heart failure who meet Light's criteria for to increased lipid content, and the fluid should be
exudates (50) . In one study of 20 patients with heart sent for lipid analysis (see the discussion later in this
failure who met Light's criteria for exudates, 1 8 had chapter) .
NT-pro-BNP levels above 1 ,300, 1 6 had NT-pro-BNP Alternately, if the supernatant is clear, the original
levels above 1 ,500, but only 1 0 had serum pleural fluid turbidity was due to increased numbers of cells or
protein gradient greater than 3 . 1 g/dL (50) . other debris. The discovery of pleural fluid that looks
Other workers have demonstrated that there is like chocolate sauce or anchovy paste is suggestive
a close relationship between the levels of NT-pro-BNP of amebiasis with a hepatopleural fistula (57) . This
in the pleural fluid and serum. Han et al. (53) measured appearance is due to the presence of a mixture of
the NT-pro-BNP levels in 240 patients and reported blood, cytolyzed liver tissue, and small solid particles
that the correlation coefficient between the pleural of liver parenchyma that have resisted dissolution.
and serum NT-pro BNP was 0.928. In a second study, A pleural fluid with a high viscosity is suggestive of
Kolditz et al. (54) measured the serum and pleural fluid malignant mesothelioma; the high viscosity is second
NT-pro-BNP levels in 93 patients including 25 with ary to an elevated pleural fluid hyaluronic acid level. Of
CHE They confirmed the results of the above study in course, the fluid from a pyothorax is also viscid because
that the levels of serum and pleural fluid NT-pro-BNP of the large amounts of cells and debris in the fluid.
were again closely correlated (r = 0.90) . From the The odor of all pleura fluids should be noted. One
latter two studies it appears that measurement of the can immediately establish two diagnoses by smelling
pleural fluid NT-pro-BNP levels provides no additional the pleural fluid. A feculent odor indicates that the
information beyond the serum measurements. patient has a bacterial infection of the pleural space
that is probably anaerobic. If the pleural fluid smells
like urine, the patient probably has a urinothorax.
G E N E RA L T E STS F O R
D I F F E R E N T I AT I N G CAU S E S O F
Red Blood Cell Count
E X U DAT E S
Only 5,000 to 1 0,000 RBCs per mm3 need be pres
Appearance of Fluid
ent to impart a red color to pleural fluid. If a pleural
The gross appearance of the pleural fluid frequently effusion has a total volume of 500 mL and the RBC
yields useful diagnostic information. The color, tur count in the peripheral blood is 5 million/mm3, a
bidity, viscosity, and odor should be described. Most leak of only 1 mL blood into the pleural space will
transudative and many exudative pleural effusions result in a blood-tinged pleural effusion. It is prob
are clear, straw colored, nonviscid, and odorless. Any ably for this reason that the presence of blood-tinged
deviations should be noted and investigated. or serosanguineous pleural fluid has little diagnostic
A reddish color indicates that blood is present, and significance. More than 1 5% of transudative and
a brownish tinge indicates that the blood has been more than 40% of all types of exudative pleural fluids
present for a prolonged period. If the pleural fluid is are blood tinged (4 1 ) ; that is, they have pleural fluid
blood tinged, the pleural fluid RBC count is between RBC counts between 5,000 and 1 00,000/mm3•
5 ,000 and 1 0,000/mm3• If the pleural fluid appears Occasionally, pleural fluid obtained by diagnostic
grossly bloody, a hematocrit should be obtained thoracentesis appears grossly bloody. In such cases, one
to determine whether the patient has a hemothorax can assume that the RBC count in the pleural fluid
(see Chapter 25) . is above 1 00,000/mm3• One should obtain a hemato
On rare occasions, the pleural fluid can be black. crit on such pleural fluids to document the amount of
Black pleural fluid has been reported with infection blood in the pleural fluid. If the hematocrit of the pleu
due to Aspergillus niger, infection due to Rhizopus ral fluid is greater than 50% of the peripheral hemato
oryzae, pigment laden macrophages following mas crit, a hemothorax is present, and one should consider
sive bleeding due to metastatic carcinoma (55) and inserting a chest tube (see Chapter 25). Usually, the
melanoma (56) . hematocrit of bloody pleural fluid is much lower than
92 P L E U RAL D I S EASES
what one would expect from its gross appearance. If a and differentials should be collected in a test tube with
pleural fluid hematocrit is not available, its estimate an anticoagulant (59) . If the pleural fluid is collected
can be obtained by dividing the pleural fluid RBC in plastic or glass tubes without an anticoagulant, the
count by 1 00,000. fluid may clot or the cells may clump, providing inac
The presence of bloody pleural fluid suggests one curate cell counts and differentials (59) .
of three diagnoses, namely, malignant disease, trauma, The pleural fluid WBC count is of limited diag
or pulmonary embolization. In a series of 22 bloody nostic use. Most transudates have WBC counts below
pleural effusions that I observed on medical wards, l ,OOO/mm3, whereas most exudates have WBC counts
1 2 were due to malignant disease, 5 to pulmonary above l ,OOO/mm3 ( 1 6) . Pleural fluid WBC counts
embolization, 2 to trauma, and 2 to pneumonia, and above 1 0 ,000/mm3 are most commonly seen with
1 was a transudative effusion secondary to cirrhosis parapneumonic effusions, but they are also seen with
(30) . The traumatic origin of the pleural effusion may many other diseases (4 1 ) , as shown in Table 7. 1 . I have
not be obvious, particularly when the patient is on seen pleural fluid WBC counts above 50,000/mm3
a medical ward. The patient may have broken a rib with both pancreatic disease and pulmonary embo
while coughing or suffered trauma during an episode lization. With grossly purulent pleural fluid, the
of inebriation that is not remembered. pleural fluid WBC count is frequently much lower
At times, it is unclear whether blood in the pleural than what one would anticipate because debris, rather
fluid resulted from or was present before the thora than cells, accounts for much of the turbidity.
centesis. If the blood is a result of the thoracentesis,
the degree of red discoloration of the fluid frequently Differential White Cell Count
is not uniform throughout the course of aspiration.
Examination of the fluid microscopically may also Examination of a Wright's stain of pleural fluid is
be useful. If the RBCs were present before the thora one of the most informative tests on pleural fluid.
centesis, the macrophages in the pleural fluid usually Because the pleural fluid WBC count is frequently
contain hemoglobin inclusions. Although the levels less than 5,000/mm3, it is useful to concentrate the
of D-dimer in the cerebrospinal fluid are useful in cells before staining. This is easily accomplished by
demonstrating whether blood in the fluid has a trau centrifuging approximately 1 0 mL of fluid and then
matic origin, pleural fluid levels of D-dimer are not resuspending the button of cells in approximately
useful in making this differentiation (58) . Crenation 0.5 mL of supernatant. After thorough mixing, slides
of the RBC in the pleural fluid rarely occurs because that are similar to those for examining peripheral blood
the osmotic pressure of the pleural fluid is similar to are made and stained in the usual way. Occasionally,
that of serum. large amounts of fibrinogen adhere to the cells. In
such cases, resuspension in saline solution, followed by
centrifugation, is indicated in order to evaluate cellular
White Blood Cell Count
morphologic features. Automatic cell counters do not
Although WBC counts on the pleural fluid were tra provide sufficiently accurate differential cell counts for
ditionally performed manually, we have shown that clinical use (59,60), presumably because of the high
the automated counters provide accurate pleural fluid number of mesothelial, lymphoid, and tumors cells in
WBC counts (59) . The pleural fluid for cell counts pleural fluid.
Pa ra p n e u m o n i c effusion 13 26 50
M a l i g na n t d i sease 3 43 7
P u l m o n a ry e m b o l ization 3 8 37
Tu berc u losis 2 14 14
Pa ncreatitis 2 5 40
Postmyoca rd i a l i nfa rcti o n Syn d ro m e 3 33
Syste m i c l u p u s e rythematosus 1 00
W B C , wh ite b l o o d ce l l .
C H APT E R 7 I C L I N ICAL M A N I F E STAT I O N S A N D U S E F U L TE STS 93
Although most laboratories divide pleural fluid pleural fluid eosinophilia is due to either air or blood
WBCs into polymorphonuclear leukocytes and in the pleural space and therefore does not contrib
mononuclear cells, I prefer to divide them into four ute any diagnostic information in these situations.
categories-polymorphonuclear leukocytes, lympho Charcot-Leyden crystals ( 66) , as well as Curschmann's
cytes, other mononuclear cells, and eosinophils spirals (67) , are occasionally found in the pleural fluid
because of the diagnostic significance of small lym of patients with pleural eosinophilia. Their presence
phocytes (see the discussion on lymphocytes later appears to have no diagnostic significance.
in this chapter) . The mononuclear cells include The factors responsible for pleural fluid eosino
mesothelial cells, macrophages, plasma cells, and philia have been intensively studied, but there is still no
malignant cells. Excellent color plates demonstrating unifying concept and it is likely that multiple factors
the morphologic and staining characteristics of the are involved. In animals, it has been shown that the
different cells in pleural effusions are contained in the intrapleural injection of stem cell factor (68) , platelet
monograph by Spriggs and Boddington (6 1 ) . activating factor (PAF) (69), endotoxin (70,71 ) , brady
kinin (72), and leukotriene B 4 (69) all result in pleural
fluid eosinophilia. The eosinophil influx is inhibited in
Neutrophils
some but not all situations by monoclonal antibodies
Because neutrophils are the cellular component of (MAb) directed against IL-5.
the acute inflammatory response, they predominate In humans, pleural fluid from patients with eosino
in pleural fluid resulting from acute inflammation philic pleural effusions stimulates bone marrow cells
as with pneumonia, pancreatitis, pulmonary embo to form colonies of eosinophils (73,74) . In addition,
lization, subphrenic abscess, and early tuberculosis. when eosinophils are incubated in the presence of
Although more than 1 0% of transudative pleural eosinophilic pleural fluid, their survival is prolonged
effusions contain predominantly neutrophils, pleural (73) . Peripheral blood from patients with eosinophilic
fluid neutrophilia in transudates has no clinical sig pleural effusions does not stimulate the bone marrow
nificance (4 1 ) . The significance of neutrophils in an to form eosinophil colonies and does not prolong
exudative pleural effusion is that they indicate acute survival of eosinophils. The factor responsible for the
inflammation of the pleural surface. increased colony-forming activity and the increased
Interleukin 8 (IL-8) appears to be one of the survival appears to be IL-5 (73,74) , although IL-3 and
primary chemotaxins for neutrophils in the pleural granulocyte or macrophage colony-stimulating fac
space (62,63) . The number of neutrophils in pleural tor (GM-CSF) may also play a role (73) . In patients
fluid is correlated with the IL-8 level, and empyemas with eosinophilic pleural effusions, the pleural fluid
have the highest levels of IL-8. The addition of IL-8 level of IL-5 is significantly correlated with the pleu
neutralizing serum decreases the chemotactic activity ral fluid eosinophil count ( r = 0.55) and the pleural
for neutrophils in empyema fluids (63) . The cellular fluid eosinophil percentage ( r = 0. 54) (75). Both
source of IL-8 is unknown (62) . bloody and nonbloody eosinophilic effusions have
Examination of the pleural fluid neutrophils in high levels of IL-5 (75). In patients with eosinophilic
patients with parapneumonic effusions is useful in pleural effusions, the pleural fluid levels of vascular
identifying those that are infected. If pleural infec cell adhesion molecule (VCAM)- 1 and eotaxin-3
tion is present, the neutrophils undergo a characteris are also significantly correlated with the eosinophil
tic degeneration. The nucleus becomes blurred and is count and percentage of eosinophils in the pleural
no longer stained purple. The cytoplasm shows toxic fluid (76) . VCAM- 1 is an adhesion molecule that is
granulation initially. Subsequently, the neutrophilic expressed on the endothelial surface and interacts with
granules become indistinct and are then lost. Finally, the /3 1 -integrins expressed on the eosinophil surface to
only a smear cell remains (6 1 ) . facilitate eosinophil tissue migration (76) .
The source of the IL-5 appears to be the CD4• lym
phocyte in the pleural fluid (74) , but the eosinophils
Eosinophils
in the pleural fluid may themselves also produce IL-5
Approximately 7% of pleural fluids are characterized (73) . The source of the eosinophils in eosinophilic
by pleural fluid eosinophilia (> 1 0%) (64,65) . Most pleural effusions appears to be the bone marrow; no
clinicians believe that significant numbers of eosino progenitor cells are present in the pleural space. It is
phils (> 1 0%) in pleural fluid should be a clue to the not known what stimulates the CD4• lymphocytes
origin of the pleural effusion. In most instances, the to produce the IL-5. However, it probably results
94 P L E U RAL D I S EASES
from another cytokine because the intrapleural injec week (6 1 ) . There is frequently an associated periph
tion of IL-2 into malignant pleural effusions results eral blood eosinophilia that does not disappear until
in an eosinophilic pleural effusion with a high level the pleural effusion is completely resolved (84) . The
of IL-5 (77) . There are factors other than IL-5 that pleural effusions associated with pulmonary emboli
recruit eosinophils to the pleural space. Antibodies to zation are frequently bloody and contain numerous
IL-5 eliminate the eosinophilic influx to an allergen eosinophils (85) . Bloody pleural fluids due to malig
but not to endotoxin in the mouse (78) . However, nant disease are not usually characterized by eosino
antibodies to gamma delta lymphocytes eliminate the philia (4 1 ,6 1 ) . In a study conducted by my colleagues
eosinophilic response to endotoxin (7 1 ) . and me of bloody pleural effusions, none of the
Th e most common cause o f pleural fluid eosino 1 1 cases of malignant pleural effusions with pleural
philia is air in the pleural space. In a series of 1 27 cases fluid RBC counts greater than 1 00,000/mm3 had
with more than 20% eosinophils in the pleural fluid, more than 1 0 % eosinophils (4 1 ) . Patients with pleu
8 1 (64%) were thought to have pleural fluid eosino ral effusions postcoronary artery bypass graft (CABG)
philia secondary to air in the pleural space (6 1 ) . In a surgery, frequently have bloody pleural effusions in
review of 343 pleural effusions with greater than 1 0% the first few weeks after surgery. In these effusions,
eosinophils, 95 (28%) had air in the pleural space the pleural fluid IL-5 level is higher than the corre
(79) . It is likely that the pleural fluid eosinophilia sponding serum level and there is a significant corre
in many other patients in this series was also due to lation between the pleural fluid and serum IL-5 level
the introduction of air into the pleural space during (86) . Moreover, the pleural fluid IL-5 levels are sig
a prior thoracentesis. On numerous occasions over nificantly correlated with the pleural fluid eosinophil
the past three decades, I have seen patients who had counts (86) . In addition, the pleural fluid eotaxin-3
no pleural fluid eosinophilia at the initial thoracen levels are significantly higher than the serum levels
tesis but who had many eosinophils at a subsequent and the pleural fluid eotaxin-3 levels significantly cor
thoracentesis. In each case, a small pneumothorax relate with the pleural fluid eosinophil count (86) .
resulted from the initial thoracentesis. When patients If the patient has neither blood nor air in their
with spontaneous pneumothorax undergo thora pleural space, what is the significance of an eosino
cotomy, a reactive eosinophilic pleuritis frequently philic pleural effusion? Kalomenidis and Light (87)
exists in the resected parietal pleura (80) . It should reviewed the etiology of 392 cases of eosinophilic
be noted, however, that in two recent series (64,8 1 ) pleural effusions when cases associated with pleural
the prevalence o f eosinophilic pleural effusion did not air and/or blood were excluded. They reported that
increase after a thoracentesis or a pleural biopsy. the most common diagnosis was idiopathic (39.8%),
The mechanism responsible for the pleural fluid followed by malignancy ( 1 7%), parapneumonic
eosinophilia in response to air in the pleural space is ( 1 2.5%), transudates (7.9%), tuberculosis (5.6%),
unknown but is probably related to IL-5. Smit et al. pulmonary embolism (4.3%), and others ( 1 2.8%). In
(82) measured the percentage of eosinophils and a recent study of 1 3 5 patients with eosinophilic pleural
the levels of IL-5 in 23 patients with pneumothorax effusions from a single institution, the following dis
and pleural fluid. They found that IL-5 level and the eases were responsible: malignancy 34.8%, infections
eosinophil concentration in the pleural fluid were 1 9.2%, unknown 1 4. 1 %, posttraumatic 8.9%, and
highly correlated ( r = 0.84) and that the eosino miscellaneous 23.0%. (64) . The incidence of malig
phil percentage tended to increase with time, with nancy was significantly higher in patients with a lower
a mean of less than 5% in the first 24 hours, 20% ( <40%) pleural fluid eosinophil percentage (64) .
at days 1 to 3, 40% at days 4 to 7, and 50% after day If neither air nor blood is present in the pleural
7 (82) . There was no relationship between the PAF space, several unusual diagnoses should be considered.
level or the monocyte chemotactic protein- I levels Pleural eosinophilia is common in patients with
and the eosinophils. In these fluids, IL-8 was not asbestos-related pleural effusions. In a review of
detectable. When air is injected into the pleural space eosinophilic pleural effusions (79) , 1 5 of 29 (52%)
of a mouse, pleural fluid eosinophilia occurs within asbestos-related pleural effusions had more than 1 0%
30 minutes and peaks at 48 hours (83) . eosinophils in the pleural fluid. Patients with eosino
The second most common cause of pleural fluid philic pneumonia frequently have pleural effusions
eosinophilia is blood in the pleural space. Follow and the mean eosinophil percentage in the pleural fluid
ing traumatic hemothorax, pleural fluid eosinophils was 38% in one study (88) . The pleural effusions sec
do not usually become numerous until the second ondary to drug reactions are frequently eosinophilic.
C H APT E R 7 I C L I N ICAL M A N I F E STAT I O N S A N D U S E F U L TE STS 95
Offending drugs include dantrolene, bromocriptine, Several papers have assessed the diagnostic utility
and nitrofurantoin (see Chapter 22) . Pleural effusions of separating pleural lymphocytes into T and B lym
secondary to parasitic diseases (see Chapter 1 5) such phocytes (94-97) . In general, this separation has not
as paragonimiasis (68) , hydatid disease (89), amebiasis been useful diagnostically. With most disease states, the
(6 1 ) , or ascariasis (6 1 ) frequently contain a large per pleural fluid contains a higher percentage of T lym
centage of eosinophils. Lastly, the pleural effusion phocytes (70%), a lower percentage of B lymphocytes
associated with the Churg-Strauss syndrome (see ( 1 0%), and a higher percentage of null cells (20%) than
Chapter 2 1 ) is eosinophilic (90) . the corresponding peripheral blood (94,95). The parti
If none of the foregoing rare diseases is causing the tioning of lymphocytes may be useful, however, when
pleural effusion, the following statements are perti chronic lymphatic leukemia or lymphoma is suspected.
nent to patients with eosinophilic pleural effusions. If In a report of four such patients, all had more than
the patient has pneumonia and pleural effusion, the 80% B lymphocytes in their pleural fluid (96) .
presence of pleural fluid eosinophilia is a good prog The development of MAb has permitted a fur
nostic sign because such an effusion rarely becomes ther subdivision of T lymphocytes. In comparison
infected. The origin of approximately 40% of eosino to peripheral blood, in pleural fluid the ratio of the
philic effusions is not established, and these effusions helper and inducer cells (CD4•) to the suppressor
resolve spontaneously. and cytotoxic cells (CDS•) is higher, regardless of the
etiology of the pleural effusion (98-1 00) . Therefore,
this subdivision is not useful diagnostically.
Basophils
Natural killer (NK) cells are lymphocytes derived
Basophilic pleural effusions are distinctly uncom from an unimmunized host that lyse certain tumor
mon. I have not seen a pleural effusion that contained cell lines and virus-infected cells. In general, the
more than 2% basophils. A few basophils are usually percentage of T lymphocytes in pleural fluid that
present in pleural effusions with eosinophils. Baso are identified as NK cells is approximately the same
phil counts greater than 1 0% are most common with (� 1 5%) as in the peripheral blood ( 1 0 1 ) . There are
leukemic pleural involvement (6 1 ) . Basophil counts two types of NK cells, CD 5 6brighr and CD5 6dim. In
greater than 5% are most common with pneumotho the pleural fluid, there is a much higher percentage of
rax, pneumonia, and transudates (9 1 ) . CD5 6brighr cells than C5 6dull cells ( 1 0 1 ) . The CD56dim
subset has cytotoxic activity that is superior to that
of the CD56brighr subset. However, there is a discrep
Lymphocytes
ancy between the number of NK positive cells and
The discovery that more than 50% of the WBCs in the NK activity of the cells when patients with tuber
an exudative pleural effusion are small lymphocytes culosis are compared with patients with malignancy.
is diagnostically important because it means that the Although the number of NK cells is comparable in
patient probably has a malignant disease, tuberculous the two populations, there is much more NK activ
pleuritis, or a pleural effusion after CABG surgery. ity in the tuberculous pleural effusions ( 1 02) . Differ
In two series (4 1 ,92) studied before the advent of ences in the NK subset percentages may explain the
CABG surgery, 96 of 2 1 1 exudative pleural effusions variation in the NK activity ( 1 03).
had more than 50% small lymphocytes. Of these
96 effusions, 90 (94%) were due to tuberculosis or
Mesothelial Cells
malignant disease.
When the foregoing series are analyzed, almost Mesothelial cells line the pleural cavmes. They
all of the effusions secondary to tuberculosis (43 of frequently become dislodged from the pleural surfaces
46) , but only two thirds of the effusions secondary and are present in the small amount of normal pleural
to malignant disease (47 of 70) , had predominantly fluid ( 1 04) . These cells are usually 1 2 to 30 µm in
small lymphocytes. In one series of 26 patients with diameter, but multinucleated forms may have diam
chronic pleural effusions post-CABG, the mean per eters up to 75 µm. Their cytoplasm is light blue
centage of lymphocytes in the pleural fluid was 6 1 % (Fig. 7 . l A) and often contains a few vacuoles. The
(93) . Approximately one third of transudative pleural nucleus is large (9-22 µm) and stains purplish with
effusions contain predominantly small lymphocytes a uniform appearance. The nucleus usually contains
(4 1 ) , and a lymphocytic transudative effusion is not one to three bright blue nucleoli (6 1 ) . Mesothelial
an indication for pleural biopsy. cells are discussed in more detail in Chapter 1 .
96 P L E U RAL D I S EASES
A B
c D
Mesothelial cells are significant for two reasons. coated with fibrin. It is also common with malignant
First, their presence or absence is often useful diag effusions after sclerosing agents have been injected to
nostically because these cells are uncommon in effect a pleurodesis. Second, mesothelial cells, partic
tuberculous effusions. Spriggs and Boddington (6 1 ) ularly in their activated form, may be confused with
analyzed 65 tuberculous effusions and found that malignant cells. Frequently, an experienced patholo
only one effusion had more than a single mesothelial gist is required to make the differentiation. Immu
cell per 1 ,000 cells (6 1 ) . Light et al. have confirmed nohistochemistry is useful in making this distinction
the paucity of mesothelial cells in tuberculous pleural (see discussion later in this chapter) .
effusions (4 1 ) , as have Yam (92) and Hurwitz et al.
( 1 05). The exception to this observation is the patient
Macrophages
with acquired immunodeficiency syndrome (AIDS) .
Patients with AIDS having a low CD4• count who In general, the presence of macrophages in pleural
have tuberculous pleuritis may have numerous meso fluid is oflimited diagnostic use. By definition, macro
thelial cells in their pleural fluid ( 1 06) . The lack of phages are cells that store vital dyes. It appears that the
mesothelial cells is not diagnostic of tuberculosis, origin of the pleural fluid macrophages can be either
however. It simply indicates that the pleural surfaces the circulating monocyte or the mesothelial cell ( 1 07) .
have become extensively involved by the disease pro Macrophages vary in diameter from 1 5 to 50 µm and
cess so that the mesothelial cells cannot enter the have irregular nuclei. Their cytoplasm is gray, cloudy,
pleural space. The absence of mesothelial cells is com and full of vacuoles. At times, the macrophage may
mon with complicated parapneumonic effusions and become engorged with debris, taking on the appear
with other conditions in which the pleura becomes ance of a "signet-ring" cell, with the nucleus flattened
C H APT E R 7 I C L I N ICAL M A N I F E STAT I O N S A N D U S E F U L TE STS 97
Dendritic cells are human leukocyte antigen (HLA) Measurement o f the pleural fluid glucose level is use
DR-positive accessory cells that play a critical role in ful in the differential diagnosis of exudative pleural
the development of cell-mediated immune reactions. effusions because a low pleural fluid glucose level
Dendritic cells have been identified in the pleural ( <60 mg/dL) indicates that the patient probably has
fluid ( 1 1 1 ) . It is unknown whether the dendritic cells one of four disorders, namely, a parapneumonic effu
in pleural fluid are resident in the pleural space or sion, malignant disease, rheumatoid disease, or tuber
reach the pleural space from the blood. Light micros culous pleuritis. Other rare causes of a low glucose
copy reveals that dendritic cells from pleural fluid are pleural effusion include paragonimiasis, hemothorax,
intermediate in size between lymphocytes and mac Churg-Strauss syndrome, and, occasionally, lupus
rophages, do not contain intracytoplasmic inclusions, pleuritis. The pleural fluid glucose level of all transu
and have an eccentric nucleus ( 1 1 1 ) . dates and of most exudates parallels that of the serum.
98 P L E U RAL D I S EASES
OTH OTH
CHF TRAN MALIG TB PNEU EXUD
•
.. I
6.0 .... • •
-
•
•
I
• •
5.0
.... .. ..
•
..
-•
i
.. •
•
_J
E
..
..
-..
0
FIGURE 7.2 • P l e u ra l fl u i d p rote i n ...
l eve l s i n effus i o n seco n d a ry to
cong estive h e a rt fa i l u re (CH F), oth e r
0
�
Cl
c
4.0 -
•
•
•
•
•
•
T
•
'
"iii
tra n s u d ates (OTH TRAN), m a l i g n a nt 0 3.o - I
..
I •
d isease (MALIG), tu bercu losis (TB), c.
..
:"Q I ..
::i •
p n e u m o n i a (PN E U), a n d oth e r
-= •
exud ates (OTH E X U D ) . Each p o i n t
represe nts o n e p l e u ra l fl u i d . N ote � 2.0 -
::i
Q)
t •
•
i:L •
that the d istri b u t i o n of p rote i n
l eve l s for a l l categ ories of exud ative I
p l e u ra l effusions is s i m i l a r. (From Light
RW, MacGregor Ml, Luchsinger PC, et al.
Pleural effusions: the diagnostic separation of
1 .0 -
t I
•
transudates and exudates. Ann I ntern M e d .
1 972;77"507-5 1 3, with permission.)
In my experience, it is not necessary to obtain pleu extent of the tumor at thoracoscopy was significantly
ral fluid glucose levels with the patient fasting or to higher in those 1 6 patients in whom the pleural fluid
take the serum glucose level into consideration when glucose was less than 60 mg/dL. In addition, patients
evaluating the pleural fluid glucose level. with a low pleural fluid glucose are more likely to have
The pleural fluid glucose level is low in some positive pleural fluid cytology and a positive pleural
patients with parapneumonic effusions or empyema biopsy ( 1 2 1 ) , are less likely to have a good result from
( 1 1 6 , 1 1 7) . The lower the pleural fluid glucose level, chemical pleurodesis ( 1 20, 1 22) , and have a shortened
the more likely that one is dealing with a complicated life expectancy ( 1 22, 1 23).
parapneumonic pleural effusion. If the pleural fluid The pleural fluid glucose level is also reduced in
is thick and purulent, the pleural fluid glucose level some patients with tuberculous pleuritis. Indeed, early
is frequently close to zero ( 1 1 6) . Even in more serous repons indicated that low pleural fluid glucose levels
fluid, the glucose level may be reduced. The presence were seen only with tuberculous pleural effusions
of a low pleural fluid glucose is a poor prognostic sign ( 1 24, 1 25). Subsequent studies (4 1 , 1 1 6, 1 1 9, 1 26) ,
in patients with parapneumonic effusion and serves however, revealed that low pleural fluid glucose levels
as an indicator that more aggressive therapy such as also occurred with malignant and rheumatoid disease
tube thoracostomy or thoracoscopy with the break and parapneumonic effusion.
down of loculations is necessary ( 1 1 8) . The distribution of pleural fluid glucose levels for
Approximately 1 5% to 2 5 % o f patients with tuberculous and malignant pleural effusions is, in fact,
malignant pleural effusions have pleural fluid glu similar (42) . Most patients with tuberculous pleuritis
cose levels below 60 mg/dL (42, 1 1 9, 1 20) and the have a pleural fluid glucose level above 80 mg/dL
level may be less than 1 0 mg/dL. Patients with malig (42) . Accordingly, a low pleural fluid glucose level is
nant pleural effusions and a low glucose level have a compatible with the diagnosis of tuberculous pleuri
greater tumor burden in their pleural space than do tis, but it is not necessary for the diagnosis.
those with normal pleural fluid glucose levels. In one Pleural effusions due to rheumatoid disease (see
report of 77 patients with malignant pleural effusion Chapter 2 1 ) classically have a low pleural fluid glu
who underwent thoracoscopic examination ( 1 20), the cose level. Carr and Power ( 1 26) first reported that
C H APT E R 7 I C L I N ICAL M A N I F E STAT I O N S A N D U S E F U L TE STS 99
rheumatoid pleural effusions had a low pleural fluid approximately 50% of patients with malignant pleu
glucose level. In a subsequent review of 76 cases of ral effusions and an elevated pleural fluid amylase.
rheumatoid pleural effusions ( 1 27) , 42% had pleural The pleural fluid amylase level in malignant pleural
fluid glucose levels below 1 0 mg/dL, and 78% had effusions is usually only minimally to moderately
levels below 30 mg/dL. The explanation for the low elevated, in contrast to the marked elevations with
pleural fluid glucose level in this condition appears to pancreatitis or esophageal rupture. The primary site
be a selective block to the entry of glucose into the of the tumor in patients with neoplastic pleural effu
pleural effusion ( 1 28). The pleural fluid glucose level sions and elevated pleural fluid amylase levels is usu
in effusions secondary to LE is usually normal. In one ally not the pancreas ( 42, 1 36) . Because the amylase
report of nine patients, the pleural fluid glucose; the in malignant pleural effusions is of the salivary type
pleural fluid glucose level was below 50 mg/dL in 2 ( 1 37), amylase isoenzyme determinations are useful
of 1 4 ( 1 4%) patients with lupus pleuritis. in distinguishing between malignant and pancreas
related pleural effusions with high amylase levels.
The pleural fluid amylase level is also elevated with
A M Y LA S E D ET E R M I N AT I O N esophageal rupture (42, 1 38). The origin of the amy
lase with esophageal rupture has been shown to be the
Pleural fluid amylase determinations are useful i n the
salivary gland rather than the pancreas ( 1 39) . With
differential diagnosis of exudative pleural effusions
the tear in the esophagus, the swallowed saliva with
because a pleural fluid amylase level above the upper
its high amylase content passes into the pleural space.
normal limits for serum indicates that the patient has
Because the early diagnosis of esophageal perforation
one of three problems: pancreatic disease, malignant is imperative, owing to the high mortaliry rate without
tumor, or esophageal rupture (42) . However, it is not rapid operative intervention, the pleural fluid amylase
cost effective to obtain an amylase measurement on
determination should be performed promptly when
every undiagnosed pleural fluid; rather, pleural fluid
this diagnosis is suspected. In animal experiments, the
amylase levels should be determined only when esoph
pleural fluid amylase concentration is elevated within
ageal rupture or pancreatic disease is suspected ( 1 3 1 ) .
2 hours of esophageal rupture ( 1 40).
Approximately 5 0 % o f patients with inflamma
tory pancreatic disease have an accompanying pleural
effusion ( 1 32) . In such patients, the pleural fluid amy
LACT I C AC I D D E H Y D R O G E N A S E
lase level is usually raised well above the normal upper
M EA S U R E M E NT
limits for serum and is also higher than the simulta
neously sampled serum (42, 1 33) . On rare occasions, The pleural fluid LDH level is used to separate tran
the pleural fluid amylase level is normal at the time sudates from exudates (see discussion earlier in this
of the original thoracentesis, only to become elevated chapter) . Most patients who meet the criteria for
at the time of a subsequent thoracentesis. In some exudative pleural effusions with LDH but not with
patients with acute pancreatitis with pleural effusion, protein levels have either parapneumonic effusions or
the chest symptoms of pleuritic chest pain and dys malignant pleural disease. Although initial reports sug
pnea may overshadow the abdominal symptoms. In gested that the pleural fluid LDH level was increased
such instances, an elevated pleural fluid amylase level only in patients with malignant pleural disease ( 1 4 1 ) ,
may be the first hint of a pancreatic problem (42) . subsequent reports demonstrated that the pleural
Patients with chronic pancreatic disease may also fluid LDH was elevated in most exudative effusions
present with a pleural effusion with a high amylase regardless of origin (Fig. 7.3), and therefore, this deter
content ( 1 34) . The effusion results when a sinus tract mination is of no use in the differential diagnosis of
connects the pancreatic pseudocyst and the pleural exudative pleural effusions ( 1 6) .
space. The patients typically appear chronically ill with Nevertheless, every time that I perform a thoracen
out abdominal symptoms and appear to have cancer. If tesis, I obtain a pleural fluid LDH level. This is because
a pleural fluid amylase level is not measured, the cor the level of the pleural fluid LDH is a reliable indicator
rect diagnosis may never be established. The key to this of the degree of pleural inflammation; the higher the
diagnosis is a markedly elevated (> 1 ,000 U/L) pleural LDH, the more inflamed the pleural surfaces. Serial
fluid amylase level ( 1 35). measurement of the pleural fluid LDH levels is infor
The pleural fluid amylase level is elevated in mative when one is dealing with a patient with an undi
approximately 1 0% of malignant pleural effusions agnosed pleural effusion. If with repeated thoracenteses
(42, 1 3 5). The serum amylase level is also elevated in the pleural fluid LDH level becomes progressively
1 00 PLE U RAL D I S EASES
FIGURE 7.3 • P l e u ra l f l u i d l a ctic d e hydrogenase (LDH) leve l s . N ote the s i m i l a r d i stributions of the L D H l eve l s
f o r a l l categories o f exudative p l e u r a l effusions. C H F, cong estive h e a rt fa i l u re; O T H TRAN, oth e r tra n s u d ates;
MALIG, m a l i g n a n t d i sease; TB, tu bercu l os i s; P N E U , p n e u m o n ia; OTH EXUD, oth e r exudates. (From Light RW,
MacGregor Ml, Luchsinger PC, et al. Pleural effusions: the diagnostic separation of transudates and exudates. An n Intern M e d .
1 972;77:507-5 1 3, with permission.)
higher, the degree of inflammation in the pleural space effusion was the LOH- 1 more than 5% above that
is increasing and one should be aggressive in pursuing in the serum, and the total pleural fluid LOH in that
a diagnosis. Alternatively, if the pleural fluid LOH level effusion was only 1 07 IU/L (40) .
decreases with time, the process is resolving and one Although the total pleural fluid LOH level is
need not be as aggressive in the approach to the patient. not useful in distinguishing among various exuda
It should be noted that needle biopsy of the pleura tive pleural effusions, one might suppose that LOH
results in an increase in the pleural fluid LOH level of isoenzymes would be useful in the differentiation.
slightly more than 10% 48 hours after the biopsy (8 1 ) . Three studies have shown that LOH isoenzymes have
When bloody pleural fluid i s obtained, one might limited value in the differential diagnosis of exuda
wonder whether the LOH measurement would be tive pleural effusions (40, 1 42, 1 43) . All benign effu
useful because RBCs contain large amounts of LOH. sions with elevated pleural fluid LOH levels and most
The presence of blood in the pleural fluid, however, malignant effusions are characterized by a higher per
usually does not adversely affect the measurement of centage of LOH-4 and LOH-5 in the pleural fluid
the LOH. In one study, LOH isoenzyme analysis was than in the corresponding serum (40) .
performed on 1 2 pleural fluids that had contained The increased amounts of LOH-4 and LOH-5
more than 1 00,000 erythrocytes/mm3• In only one are thought to arise from the inflammatory WBCs
C H APT E R 7 I CLI N I CAL M A N I F E STAT I O N S A N D U S E F U L TESTS 101
in the pleural effusion (40) . Approximately, one third decrease of only 1 mEq in the bicarbonate concentra
of malignant pleural effusions have a different pleu tion ( 1 46) , pleural fluid acidosis is characterized by a
ral fluid LDH isoenzyme pattern that is character Pco 2 that is increased proportionately more than the
ized by large amounts (>35%) of LDH-2 and less bicarbonate is reduced (43 , 1 44) .
LD H-4 and LD H-5 . None of the 3 1 benign exudates The increased pleural fluid Pco 2 could result from
in one series had more than 35% LDH-2 (40) . No either an increased production of CO 2 or a decreased
relationship exists between the histologic type of the diffusion of co 2 from the pleural fluid to blood or
malignant pleural disease and the pleural fluid LDH a combination of these factors. It is my belief that
isoenzyme pattern (40) . At present, the only situa limited diffusion of co 2 out of the pleural space is
tion in which we obtain LDH isoenzyme analysis of the predominant mechanism. In Figure 7.4A, it can
pleural fluid is when there is a bloody pleural effusion be seen that changes in the arterial Pco2 of a patient
in a patient who is clinically thought to have a tran with a malignant pleural effusion and mild pleural
sudative pleural effusion. If the LDH is in the exuda fluid acidosis were not associated with changes in the
tive range and the protein is in the transudative range, pleural fluid Pco 2 •
the demonstration that most of the pleural LDH is Similarly, in a second patient, the administration
LDH-1 indicates that the increase in the LDH is due of bicarbonate with an increase in the arterial pH
to the blood. from 7.40 to 7.59 did not change the pleural fluid pH
or bicarbonate level (Fig. 7.4B) . When pleural fluids
are incubated at 37°C in vitro, no correlation exists
pH A N D Pco2 M E A S U R E M E N T
between the rate of acid accumulation in vitro and the
Measurement of the pleural fluid pH and Pco2 is pleural fluid pH in vivo ( 1 46, 1 47) , with the possible
useful in the differential diagnosis of exudative pleu exception of patients with complicated parapneu
ral effusions. If the pleural fluid pH is less than 7.2, monic effusions in whom the rate of acid accumula
it means that the patient has 1 of 10 conditions: tion is high ( 1 47) .
(a) complicated parapneumonic effusion, (b) esopha The pleural fluid pH is frequently not measured
geal rupture, (c) rheumatoid pleuritis, (d) tuberculous correctly ( 1 48-1 50) . Chandler et al. ( 1 48) surveyed
pleuritis, (e) malignant pleural disease, (f) hemothorax, the methods by which pleural fluid pHs were mea
(g) systemic acidosis, (h) paragonimiasis, (i) lupus sured at 277 acute care institutions in the southeast
pleuritis, or (j) urinothorax. ern part of the United States in 1 998. They reported
The pleural fluid pH is obviously influenced by the that the pleural fluid pH was measured with the
arterial pH. With transudative pleural effusions, the blood gas machine in only 32% of the institutions,
pleural fluid pH is usually higher than the simulta whereas it was measured with dip stick or pH indica
neous blood pH (43) , presumably because of active tor paper in 56% and by a pH meter in 1 2% ( 1 48).
transport of bicarbonate from the blood into the A survey of 267 pulmonologists in 2008 from the
pleural space (44) . If a low pleural fluid pH is dis United States revealed that 39% of the physicians
covered, the arterial pH should be checked to ensure who use the pleural fluid pH in the management
that the patient does not have systemic acidosis. With of parapneumonic effusions were wrong in their
certain exudative effusions, the pleural fluid pH assumption that their laboratory used the blood gas
falls substantially below that of the arterial pH. The machine to measure pleural fluid pH ( 1 50) . It has
explanation for the relative pleural fluid acidosis is as been shown that neither pH indicator strip paper
follows. The relationship between the pleural fluid ( 1 48, 1 5 1 , 1 52) nor pH meters ( 1 5 1 ) are sufficiently
pH and the blood pH depends on the extent to which accurate for clinical use. Bowling et al. ( 1 53) recently
the blood and pleural fluid Pco2 and bicarbonate are reported similar results from North Carolina. In this
in equilibrium. In conditions associated with pleural study, only 2 of 1 1 hospitals measured pleural fluid
fluid acidosis, lactic acid accumulates in the pleural pH with a blood gas analyzer. In a second study
fluid ( 1 44, 145), presumably from anaerobic glycoly ( 1 53), 43% of 22 1 pulmonologists who use the pleu
sis in the pleural fluid or tissues. ral fluid pH were not aware that only pH's obtained
The hydrogen ions associated with the lactic with the blood gas machine are sufficiently accurate.
acid combine with bicarbonate to form water and The above studies demonstrate that it is important
carbon dioxide. Accordingly, the pleural fluid Pco 2 for physicians who order pleural fluid pH to know
increases and the pH decreases. Because the addition how their hospital measures the pleural fluid pH.
of 1 mEq of fixed acid to 1 L of pleural fluid results The pH meter gives a reading that is approximately
in an increase of 33 mm Hg in the Pco 2 but in a 0.20 to 0.30 too high because it measures the pH at
1 02 PLE U RAL D I S EASES
70
60
Pco2
50
r
35
40
HC03
7.5
25
0 30 60 90 1 20
l<---- 02------.i l� I PPB-->i
A Minutes
Pco2
50 j
40
40
30 HC03
7.6
20
7.5
PT 2 Tuberculous
pH pleu ritis
7.4 • Arterial blood
• Pleural fluid
7.3
B 0 30 60 1 20 1 50
02 02 NaHC03
Minutes
room temperature (at which the Pco2 is lower) and whether invasive procedures on the pleura are neces
the sample comes in contact with room air, which sary for patients with parapneumonic effusions ( 1 58).
allows the C0 2 to escape again lowering the Pco2 If the patient has an infection with Proteus organisms,
( 1 5 1 ) . At times, laboratory personnel object to inject the pleural fluid pH may be elevated because these
ing the pleural fluid through blood gas machines for organisms produce ammonia by their urea-splitting
fear of the development of clots. This objection can ability, which can increase the pH ( 1 59). In patients
be overcome if a clot-catching apparatus is inserted with parapneumonic effusions, the pleural fluid pH
between the syringe and blood gas machine ( 1 5 1 ) . may fall before the pleural fluid glucose level becomes
Th e pleural fluid p H can also b e measured accurately depressed ( 1 1 7, 1 60) . It should be noted that the pH
with a handheld analyzer such as the I-STAT Portable can vary markedly from locule to locule in patients
Clinical Analyzer that is used in intensive care units with parapneumonic effusions ( 1 6 1 ) .
for measuring blood gases ( 1 49) . Th e pleural fluid p H is also decreased with esopha
When the pleural fluid pH is used as a diagnostic geal rupture ( 1 38 , 1 62) . In fact, Dye and Laforet ( 1 62)
test, it must be measured with the same care as arte concluded that a pleural fluid pH of less than 6.0 was
rial pH. The fluid should be collected anaerobically in highly suggestive of esophageal rupture. These work
a heparinized syringe (see Chapter 28) . If the fluid is ers attributed the low pleural fluid pH to the reflux
opened to room air, carbon dioxide will leave the fluid of gastric acid through the rent in the esophagus into
and the recorded pH will be falsely high ( 1 5 1 , 1 54) . It the pleural space. Subsequent studies in rabbits ( 1 63),
appears that it is not necessary to put the fluid on ice however, have demonstrated that the pleural fluid pH
if the pH is measured within an hour or so ( 1 54, 1 5 5). becomes just as acidic afrer esophageal rupture if the
Accurate results are also obtained when pleural fluid esophagogastric junction is ligated. It appears that the
is transferred from a large syringe to a small heparin low pleural fluid pH is due to infection in the pleural
ized syringe ( 1 56) . If the effusion is small, injection space rather than to acid reflux. Over the past few
of a local anesthetic into the pleural fluid may falsely years, we have seen several patients with pleural infec
lower the pH ( 1 57) as can residual lidocaine in the tion without esophageal rupture in whom the pleural
blood gas syringe ( 1 54). If frank pus is obtained at fluid pH was below 6.0.
thoracentesis, one should not submit it for pH deter In summary, esophageal rupture is associated with
mination because the thick, purulent fluid may clog a low pleural fluid pH because of the concomitant
the blood gas machine and laboratory personnel may pleural infection and not acid reflux. A pleural fluid
hesitate to analyze subsequent pleural fluids. pH below 6.0 is consistent with but not diagnostic of
In general, pleural fluids with a low pH also have esophageal rupture.
low glucose and high LDH levels ( 1 44) . If the labora Patients with pleural effusions secondary to both
tory reports a low pH with normal glucose and low malignant disease and tuberculosis may have a low
LDH levels, the pH measurement is probably in error. pleural fluid pH (43 , 1 44, 1 64) . When Light et al.
In a similar manner, a low glucose level with a normal wrote their first paper on pleural fluid pH (43) , they
pH and a low LDH is probably a laboratory error. concluded that the pleural fluid pH was useful in
The only reason to measure the pleural fluid Pco 2 distinguishing tuberculous pleural effusions from
is to verify the pleural fluid pH because a low pleural malignant pleural effusions; a pleural fluid pH below
fluid pH is almost always associated with a high Pco 2 7.3 suggested tuberculosis, whereas a pleural fluid pH
(43, 1 44) . The pleural fluid Pco 2 adds no diagnostic above 7.4 suggested malignant disease. Subsequent
value. studies by others ( 1 45, 1 64, 1 65), however, and my
The pleural fluid pH is most useful in indicating own observations have not supported this conclusion.
the prognosis of patients with pneumonic effusions At present, I consider the pleural fluid pH valueless
(see Chapter 9) . If the pleural fluid pH is below 7.0, in distinguishing tuberculous pleural effusions from
the patient invariably has a complicated parapneu malignant pleural effusions. The pleural fluid pH does
monic effusion, and attempts should be made to provide information about malignant pleural effusions
remove all the pleural fluid with therapeutic thora because patients with a low pleural fluid pH have a
centesis or tube thoracostomy ( 1 1 8) . If the pleural shorter life expectancy and are less likely to have a
fluid pH is above 7.2, the prognosis of the patient is favorable response to pleurodesis ( 1 2 1 , 1 65).
excellent and the pleural fluid need not be removed. The pleural fluid pH is almost always less than
The American College of Chest Physicians recom 7.2 with rheumatoid pleural effusions ( 1 29). The
mends using the pleural fluid pH to determine pleural fluid pH with lupus pleuritis is usually above
1 04 PLE U RAL D I S EASES
7.35 ( 1 29), but occasionally, it is less than 7.2 ( 1 30) . the nuclear size of the cells in pleural effusions has
The pleural fluid pH tends to be low with both para been used diagnostically. Marchevsky et al. ( 1 70) per
gonimiasis ( 1 66) and Churg-Strauss syndrome (90) , formed a computer-assisted morphometric study of 48
and these are the only two conditions in which a low pleural fluids including 20 benign fluids, eight meso
pleural fluid pH is associated with pleural eosino theliomas, and 20 carcinomas. If the mean nuclear
philia. Another situation in which the pleural fluid diameter exceeded 1 0 . 5 µm or the mean nuclear diam
pH may be decreased is with a large hemothorax (43) . eter exceeded 9 .3 µm, and the mean nuclear diameter
The metabolism of the many RBCs in this condition, divided by the cytoplasmic diameter exceeded 0.74,
in conjunction with the atelectatic underlying lung, the patient had a malignant pleural effusion. This mor
is the probable explanation for the decreased pH. phologic analysis was not able to separate carcinomas
Finally, the pleural fluid pH may be reduced with uri from mesotheliomas.
nothorax ( 1 67) . This is the only situation in which a There are, however, cytologic characteristics that
transudative pleural fluid has a low pH without con tend to be different for mesothelioma and adenocar
comitant systemic acidosis. cinomas. Stevens et al. ( 1 7 1 ) compared the cytologic
characteristics of 44 cases of malignant mesothelioma
and 46 cases of metastatic adenocarcinomas. They
T E STS F O R D I AG N O S I N G P L E U RA L concluded that the following five features separate
M A L I G N A N CY malignant mesothelioma from adenocarcinoma with
more than 95% accuracy. Mesotheliomas tend to
Cytologic Examination of Pleural Fluid
have true papillary aggregation, multinucleation with
Cytologic examination of pleural fluid is one of the atypia, and cell-to-cell apposition, whereas adeno
most informative laboratory procedures in the diagno carcinomas tend to have acinus-like structures and
sis of pleural effusions because with it a definitive diag balloon-like vacuolation ( 1 7 1 ) .
nosis can be made in more than 50% of patients with Malignant cells sometimes aggregate, and large
malignant disease involving the pleura. It is important balls or clumps of cells are characteristic of adeno
to process pleural fluid specimens expeditiously when carcinoma (Fig. 7. l D) . Although aggregates of 20 or
they are submitted for cytology. Specimens maintained more benign mesothelial cells occasionally occur, the
at room temperature deteriorate markedly within bizarre, large, vacuolated cells with adenocarcinoma
48 hours as do refrigerated specimens maintained for allow for a differentiation between these entities.
96 hours ( 1 68). The optimal amount of pleural fluid Both malignant cells and macrophages may have vac
to submit for direct smear and cell block preparations uolation. Small numbers of mitotic figures frequently
appears to be 1 50 ml although the diagnosis of malig occur in benign effusions, and, accordingly, the pres
nancy can be established in most patients with 1 0 ml ence of such figures is not indicative of malignant
pleural fluid ( 1 69) . disease.
Malignant cells have several characteristics that The accuracy of the cytologic diagnosis of malig
differentiate them from other cells in the pleural nant pleural effusions has been reported to be anywhere
fluid (6 1 ) . Malignant cells in a given pleural effu between 40% and 87% ( 1 72-1 74) . Several factors
sion are recognizably similar to each other and are influence the percentages in the various reports. First,
different from any nonmalignant cells in pleural fluid in many patients with proven malignant disease and
(Fig. 7 . 1 C) . Although the overall appearance of the pleural effusion, the effusion is not related to malig
malignant cells is similar, sometimes there is a marked nant involvement of the pleura but is rather second
variation in their sizes and shapes; one cell may have ary to other factors such as CHF, pulmonary emboli,
many times the diameter of its twin. pneumonia, lymphatic blockade, or hypoproteinemia.
Frequently, malignant cells are large. The nuclei of In such patients, one cannot expect the result of the
malignant cells may exceed 50 µm in diameter, in con pleural fluid cytologic test to be positive. For example,
trast with mesothelial cell nuclei, which rarely exceed it is unusual for the results of pleural fluid cytologic
20 µm in diameter. Small lymphocytes, by comparison, tests to be positive in patients with squamous cell
have a diameter of approximately 10 µm. The nucleoli carcinoma (4 1 ,6 1 , 1 75) because the pleural effusions
of malignant cells are often large, exceeding 5 µm in are usually due to bronchial obstruction or lymphatic
diameter, whereas the nucleoli of nonmalignant cells blockade. Second, the frequency of positive cytologic
in pleural fluid usually do not exceed 3 µm. Malig results depends on the tumor type. For example, with
nant cells have a high nucleocytoplasmic ratio. Indeed, lymphoma, the cytologic examination was positive in
C H APT E R 7 / CLI N I CAL M A N I F E STAT I O N S A N D U S E F U L TESTS 1 05
75% of patients with diffuse histiocytic lymphoma the neoplasm accurately into its histologic type such
but in only 25% of patients with Hodgkin's disease in as adenocarcinoma. Only occasionally is it possible to
one series ( 1 76) . The cytologic test is more frequently suggest with confidence the primary site of the neo
positive with adenocarcinomas than with sarcomas plasm ( 1 75).
( 1 75). Third, the accuracy of the results depends on
the way in which the specimens are examined. If both Nucleolar Organi zer Reg ions
cell blocks and smears are prepared and examined, the
Nucleolar organizer regions (NOR) are loops of DNA
percentage of positive diagnoses will be greater than
in the nucleus that code for ribosomal RNA and are
if only one method is used ( 1 77) . It is recommended
important in the synthesis of protein ( 1 80) . These
that the standard preparation of an effusion should
regions are associated with acidic nonhistone pro
include the preparation of a cellblock, and cytospins
teins that can be visualized by argyrophilic staining
stained with Diff-Quik and Papanicolaou stains
(AgNOR) . In general, malignant cells have more
( 1 78). Fourth, the more separate specimens submitted
AgNOR staining than benign cells ( 1 80) . Although
for cytologic examination, the higher the percentage
several papers have concluded that AgNOR staining
of positive repons (4 1 , 1 76) . In my own experience in
is useful in separating benign and malignant effusions
patients with proven malignant disease involving the
( 1 8 1- 1 84) , there has been no standardization of the
pleural space, the initial pleural fluid cytologic exami
technique and there is overlap between benign and
nation is positive in approximately 60% of patients,
malignant cells. Until more research is conducted on
and if three separate specimens are submitted, nearly
AgNOR, it cannot be recommended.
80% of the patients will have positive results (42) . The
third specimen frequently contains fresher cells that
lmmunohistochemical Studies
allow the diagnosis to be made. Fifth, the incidence of
positive diagnoses is obviously dependent on the skill With the development of the necessary technology for
of the cytologist. Sixth, the incidence of positive diag MAb, numerous papers have been published in the
noses is related to the tumor burden in the pleural last 25 years that have assessed the diagnostic utility of
space. A patient with a large tumor burden is more MAb in the diagnosis of pleural malignancy.
likely to have a positive pleural fluid cytology than is a The basis for this approach is the belief that there
patient with a small tumor burden. are antigens that are unique for benign mesothelial
Most tumors that are metastatic to the pleura have cells, adenocarcinoma cells, and malignant mesotheli
an epithelial origin whereas the cells that are normally oma cells. If MAb are developed against these specific
in the pleural space (neutrophils, lymphocytes, meso antigens, then positive identification of these cells can
thelial cells and macrophages) do not have an epithe be made when tissue samples or cytologic prepara
lial origin. Kielhorn et al. ( 1 79) reported that the yield tions are incubated with the antibody and then coun
with cytology was increased if epithelial cells were terstained with immunoalkaline phosphatase or using
selected by immunomagnetic selection of cells bind some similar method.
ing to EpCAM antibodies. In their study of 59 effu Many studies have compared the usefulness of
sions, the cytology alone was positive in 1 2 patients, the different antibodies in distinguishing the three
but when the epithelial cells were identified with the different cells and these studies are summarized in a
EpCAM antibodies, the cytology became positive in recent article by Ordonez ( 1 85). At the present time,
1 6 patients ( 1 79). the best markers for adenocarcinoma appear to be
In summary, when three separate pleural fluid carcinoembryonic antigen (CEA), MOC-3 1 , B72.3,
specimens from a patient with malignant pleural dis Ber-EP4, BG8, and TTF- 1 , whereas the best markers
ease are submitted to an experienced cytologist, one for mesothelioma appear to be calretinin, keratin
should expect a positive diagnosis in approximately 516, podoplanin, and WT l ( 1 85). TTF-1 has high
80% of patients. Because it is important to prevent specificity for lung carcinoma ( 1 85). It is important
the pleural fluid specimen from clotting, about 0.5 mL to realize that nonmalignant mesothelial cells will also
heparin should be added to the syringe during a diag stain positive for calretinin and cytokeratin 5/6 ( 1 86) .
nostic thoracentesis (see Chapter 28) . If a larger vol Moreover, the sarcomatous type of mesothelioma
ume of pleural fluid is obtained during a therapeutic is positive with calretinin less than 1 0% of the time
thoracentesis for submission to a cytologist, additional ( 1 87) . Ordonez concluded that D2-40 and podo
heparin should be added. From the examination of the planin are the best immunohistochemical markers for
exfoliated cancer cells, it is usually possible to classify epithelioid mesotheliomas ( 1 88) .
1 06 PLE U RAL D I S EASES
mucins, which are diagnostic of adenocarcinomas. effusion, 77 with probable malignant effusions, and
In one study the PAS-D stain was positive in 27 of 1 73 with benign effusions. Cutoff levels were estab
44 (6 1 %) patients with adenocarcinoma but in no lished such that the levels in all 1 73 benign effu
patients with mesothelioma (200) . In summary, if sions were below the cutoff level. Using these cutoff
the cells stain positive with PAS-D, the patient in all levels, 54% of the malignant effusions were clas
probability has an adenocarcinoma. If the cells stain sified as malignant and more than one third of the
positive with Aldan blue, the patient in all probabil cytology-negative malignant pleural effusions could
ity has a mesothelioma. If the cells stain positive with be identified by at least one marker (206) . These
neither, no conclusion can be made. authors suggested that the main use of measuring
tumor markers in the pleural fluid would be to select
patients (those with the higher levels) for more inva
Tumor Markers in Pleural Fluid
sive studies (206) .
The possibility of establishing the diagnosis of pleural
malignancy by demonstrating an elevated level of a
Soluble Mesothelin Related Protein (SMRP)
tumor marker in the pleural fluid has been the subject
of many publications. The tumor markers that have It has been suggested that elevated pleural fluid levels
been evaluated have included (20 1-2 1 2), carbohy ofSMRP are useful in the diagnosis of malignant meso
drate antigens CA 1 5-3 (20 1 ,205-209,2 1 2-2 1 4), CA thelioma (224-226) . Creaney et al. (224) measured
1 9-9 (20 1 ,205,207,208,2 1 5), CA 549 (2 1 2), CA 72-4 the SMRP levels in pleural fluids from 1 92 patients
(20 1 ,205,2 1 2), cytokeratin 1 9 fragments (CYFRA presenting to a respiratory clinic including 52 with
2 1 - 1 ) (20 1 ,202,206,207,2 1 6), cancer antigen 125 malignant mesothelioma, 56 with nonmesothelioma
(CA 1 25) (206,207) , sialyl stage-specific antigen tous malignancies and 84 benign effusions. The pleu
(2 1 7,2 1 8), neuron-specific enolase (20 1 ,203,2 1 9), ral effusions from the patients with mesothelioma
squamous cell carcinoma antigen (20 1 ,203), HER-2/ had significantly higher concentrations of SMRP
neu (220) and telomerase (22 1 ,222) . than did the other patients (224) . However, the
In general, I do not recommend that tumor markers SMRP in patients with sarcomatoid mesothelioma
be used in the evaluation of patients with undiagnosed did not differ significantly from nonmalignant
pleural effusions (223) . Although an elevated level of effusions (224) . Davies et al. (225) measured pleu
any of the tumor markers is very suggestive of malig ral fluid SMRP in 24 patients with mesothelioma,
nancy, specificity is not high enough to establish the 67 patients with pleural metastases and 75 patients
diagnosis. Although there is no doubt that the median with benign conditions. Using ROC curve analysis,
levels of the different tumor markers are significantly pleural fluid SMRP had an AUC of0.878 in its ability
higher in patients with malignancy than in patients to differentiate between patients with mesothelioma
with benign pleural effusions, there is always some and all other diagnoses at an optimal cutoff value of
overlap in the values. If the cutoff level for a tumor 20 nM (225) . At this cutoff, the diagnostic sensitive
marker is set high enough that the level is exceeded and specificity were 0.71 and 0.90, respectively (225) .
by none of the benign effusions, then the test tends to Again, the levels of SMRP were lower in patients who
be very insensitive. The same criticism could be made had sarcomatoid mesothelioma (225) . Adenocar
of tests on the pleural fluid for tuberculosis, such as cinomas accounted for 1 2 of the 1 3 false positives.
adenosine deaminase (ADA) or interferon-gamma. The above two studies demonstrate that pleural fluid
Nevertheless, I rely heavily on these tests to establish SMRP measurement provide additional information
the diagnosis of tuberculosis. The primary difference to cytology. However, tissue confirmation of meso
in the two situations is that it would be disastrous thelioma is indicated in most situations.
to wrongly make the diagnosis of malignancy with
a tumor marker because the patient is essentially told
Oncogenes
that he or she has only 90 days to live. If the diagnosis
of tuberculosis is wrongly established, the patient is The development of cancer is a multistep process in
not sentenced to death but rather to taking antituber which multiple genetic alterations must occur. The
culous medications for 6 to 9 months. transforming genes are collectively called oncogenes.
The best study to date was reported by Poree! et al. The oncogenes may be related to viruses, environ
(206) . They measured the pleural fluid levels of CEA, mental carcinogens, or spontaneous mutations.
CA- 1 25, CA 1 5-3 and cytokeratin 1 9 fragments in Because the oncogenes are associated with the devel
4 1 6 patients including 1 66 with definite malignant opment of malignancy, one might hypothesize that
1 08 PLE U RAL D I S EASES
pleural effusions (244,245) . Malignant cells have more to have tuberculosis. Liang et al. (255) performed a
chromosomes and marker chromosomes, which are meta-analysis of 63 articles evaluating the diagnostic
chromosomes with structural abnormalities (transloca usefulness of ADA that included 2,796 patients with
tion, deletion, acentric, dicentric, inversion, isochro tuberculous pleuritis and 5,297 patients with other
mosome, or ring) (244) . It remains to be demonstrated diseases. They reported that the mean sensitivity was
that there is a place for chromosomal analysis in the 0.92, the mean specificity was 0.90, the mean posi
routine examination of pleural fluid. However, there tive likelihood ratio was 0.903 and the mean negative
may be a place for fluorescence in situ hybridization likelihood ratio was 0. 1 0 (25 5). In the largest series
(FISH) to demonstrate chromosomal abnormalities. from a single institution, Poree! et al. (256) reported
In the last few years, a new methodology called measurements of the pleural fluid ADA. In their study
FISH has been introduced (246) . With this technique of 2, 1 04 pleural effusions including 22 1 with tuber
prespecified chromosomal aberrations can be visual culous pleuritis, the sensitivity was 0.93, the specific
ized in cytologic smears (246) . Fiegl et al. (247) used ity was 0.90, the positive likelihood ratio was 1 0.05
cytology and FISH techniques on 1 94 effusions from and the negative likelihood ratio was 0.07 (245) .
patients with malignancy. They reported that cytology Valdes et al. (254) reported their results on pleural
was positive in 99 (5 1 % ) whereas FISH was positive in fluids from 405 patients, including 9 1 due to tuber
1 1 6 (60%) (247) . One of the two tests was positive in culosis, 1 1 0 due to malignancy, 5 8 due to pneumonia,
1 33 (69%) of the effusions (247) . They concluded that 1 0 due to empyema, 88 transudates, and 48 miscella
the FISH technique was complimentary to cytology neous. Their results were very similar to those of pre
in establishing the diagnosis of malignant pleural effu vious workers with the exception that empyemas also
sion (247) . The FISH technique on the pleural fluid had very high ADA levels (Fig. 7.5). Measurement
is also useful in distinguishing mesothelioma from of the ratio of the pleural fluid to the serum ADA is
benign effusions (248) . much less useful diagnostically (254) .
The pleural fluid ADA is also elevated in patients
Proteomics with tuberculous pleuritis who are immunosup
pressed. The levels of ADA in patients with and with
Proteomics involves identifying all the proteins pres out AIDS are comparable (257) and renal transplant
ent in a given specimen and then identifying those patients who develop a tuberculous pleural effusion
proteins that are unique to a certain condition. There have an elevated pleural fluid ADA level (258).
have been a couple of preliminary articles exploring Some caution must be used in relying on ADA
the feasibility of using proteomics for the diagnosis levels exclusively to establish the diagnosis of tuber
of malignant pleural effusions (249,250) . Unique culous pleuritis. The two main diseases that cause an
proteins have been identified but it is still too early elevated ADA in addition to tuberculosis are rheuma
to determine whether proteomics will turn out to toid pleuritis and empyema (254,259) . If the diag
be diagnostically useful. Another preliminary study nostic criteria for tuberculous pleuritis also include a
(25 1 ) used proteomics to analyze the proteins from pleural fluid lymphocyte-to-neutrophil ratio greater
exosomes isolated from pleural effusions and reported than 0.75, the specificity of the test is increased (260) .
that they were able to identify some proteins that High pleural fluid ADA levels have also been reported
were not previously reported. with a very small percent of other neoplasms (26 1 ) ,
with Q fever (262) , with brucellosis (263) and with
T E STS F O R D I AG N O S I N G P L E U RA L Legionnaire's disease (264) .
T U B E R C U LO S I S The pleural fluid ADA level can be used to exclude
the diagnosis of tuberculous pleural effusions in
Adenosine Deaminase Measurement (ADA)
patients with undiagnosed lymphocytic pleural effu
Measurement of the ADA level in pleural fluid is sions. Two series (265 ,266) measured the pleural fluid
diagnostically useful because ADA levels tend to be ADA levels in 506 patients with lymphocytic pleural
higher in tuberculous pleural effusions than in other effusions not due to tuberculosis. Only 1 0 patients
exudates (252-256) . ADA is the enzyme that cata (2%) had an ADA level above 40 U/L. The diagnoses
lyzes the conversion of adenosine to inosine. In gen in these 1 0 patients included 3 cases of lymphoma,
eral, a cutoff level of between 40 and 45 U/L is used 3 cases of parapneumonic effusions, 2 bronchogenic
with levels above this being indicative of tuberculosis. carcinomas, 1 mesothelioma, and 1 idiopathic pleural
The higher the level, the more likely the patient is effusion (265 ,266) . The pleural effusions that occur
1 10 PLE U RAL D I S EASES
.
* 1 , 900
}
1 70 •
•
· ·- * 1 ,500
• *
* 800
.
1 50 ' •
• *
* 700
k:
*
1 30 ' t.. * •
*
x- * 500
: *
***
***
f.·
110
*
-.
**
** 300
t
I•
90 t
*
.
• • 200
II • *
•
---
70 t •
...
� ... ...
·t - �:
t *
1•
1 40
�- *
... *
t ...
••
47 •
*J t ... 1 00
•
•
*
•
* •
30 '
r
- !� ,
••
* 50
*
10 ' •
• .
•
after CABG surgery are typically lymphocytic, but than the cells in the pleural fluid. There is one report
their ADA levels are below 40 U/L (265). in which the use of an ADA l -to-ADA total ratio of
Several papers have been written on the diagnostic less than 0.42 increased the accuracy with which the
utility of ADA isoenzymes. ADA has two isoenzymes, diagnosis of tuberculous pleuritis was established
ADAl and ADA2 (267-270) . ADAl is ubiquitous (269) . In most cases, ADA isoenzymes are not needed
and is produced by lymphocytes, neutrophils, mono to establish the diagnosis of tuberculosis. However, in
cytes, and macrophages (268) . In contrast, ADA2 certain instances they can be quite useful.
exists only in monocytes and macrophages. The An added advantage of ADA in the diagnosis
increase in ADA activity with tuberculous pleuritis of pleural effusions is that if 0.9 mL pleural fluid is
is mainly due to ADA2, which is surprising because added to a test tube containing 0. 1 0 mL of a mixture
most of the cells in the pleural fluid are lymphocytes. of 50% glycerol and 50% ethylene glycol, the pleural
These observations suggest that the origin of the pleu fluid can be mailed by regular mail with no loss of
ral fluid ADA is probably in the pleural tissues rather ADA activity (27 1 ) . If the pleural fluid is stored at
C H APT E R 7 I CLI N I CAL M A N I F E STAT I O N S A N D U S E F U L TESTS 111
4 degrees C or -20 degrees C, the level ofADA remains In view of the information mentioned in the
stable up to 28 days (272) . preceding text, it appears that measurement of the
In view of the foregoing information, it is recom interferon-gamma level is very useful in the diag
mended that facilities in which a sizable percentage of nosis of tuberculous pleuritis. One must be careful
the cases of pleural effusions are due to tuberculosis in interpreting the results from a given laboratory,
develop the faculty to perform ADA assays on pleural however, as different laboratories report their results
fluid. An ADA level above 70 U/L in a patient who in different units. In the United States, pleural fluid
does not have an empyema or rheumatoid arthritis interferon-gamma levels can be obtained from Immu
(RA) is essentially diagnostic of tuberculous pleuritis. noscience Laboratory in Los Angeles, 3 1 0-667 - 1 077,
An ADA level above 40 is suggestive of tuberculosis, www. immuno-sci-lab.com.
and the higher the ADA, the more likely the diagno Which test should be used to establish the diagnosis
sis of tuberculous pleuritis. of tuberculous pleuritis? Greco et al. (279) reviewed
all English language studies from 1 978 to November
2000. The studies included 4,738 patients on whom
Interferon-Gamma ADA was measured and 1 , 1 89 patients on whom
interferon-gamma was measured. These researchers
Measurement of the pleural fluid levels of interferon reported that the maximum joint sensitivity and
gamma is useful in the diagnosis of tuberculous specificity for ADA was 93%, whereas it was 96% for
pleuritis because patients with tuberculous pleuritis interferon-gamma (279) . Because there is not much
tend to have higher levels of interferon-gamma than difference in the performance of the two tests and as
do other types of exudates (273-278) . Jiang et al. ADA is much less expensive, ADA appears to be the
(278) performed a meta-analysis on the diagnostic preferred test.
usefulness of pleural fluid interferon-gamma levels in
22 studies with 782 patient with tuberculous pleuritis
and 1 ,3 1 9 patients with pleural effusions due to other Interferon-gamma release assays (IG RA)
diseases. They reported that the mean sensitivity was IGRA on the blood were designed to diagnose latent
0.89, the mean specificity was 0.97, the mean posi tuberculosis (280) . There are two commercially avail
tive likelihood ratio was 23.45 and the mean negative able assays the QuantiFERON TB Gold (Qtf Gold;
likelihood ratio was 0. 1 1 (278) . Cellestis Ltd., Carnegie, Victoria, Australia) and the
Individual studies have reported similar results. T SPOT.TB (Oxford Immunotec Ltd., Abington,
Pleural fluid interferon-gamma levels from the UK) . These tests incubate whole-blood or isolated
145 patients reported by Valdes et al. are shown in peripheral blood mononuclear cells with Mycobac
Figure 7.5. As can be seen from this figure, 26 of terium tuberculosis specific antigens and assay the
35 patients (74%) with tuberculous pleurisy had amount of interferon-gamma released.
interferon-gamma levels above 200 pg/mL, whereas The IGRA are inferior to the pleural fluid
only 1 out of 1 1 0 other effusions that were not empy interferon-gamma levels (280-28 1 ) . Zhou et al. (28 1 )
emas had an interferon-gamma level that exceeded performed a meta-analysis o f the available reports
this. Villena et al. (277) reported comparable results on IGRA on the pleural fluid for diagnosing tuber
in the pleural fluids of 595 patients, including 82 culous pleuritis. They analyzed seven reports with a
with tuberculous effusions. She reported that an total of 2 1 3 patients with tuberculous pleuritis and
interferon-gamma level of 3.7 IU/mL (as measured 1 53 patients with pleural effusions of other etiologies
by radioimmunoassay) had a sensitivity of 0.98 and and reported that the mean sensitivity was 0.75, the
a specificity of 0.98 (277) for the diagnosis of tuber mean specificity was 0.82, the mean positive likeli
culous pleuritis. hood ratio was 3 .49 and the mean negative predictive
Interferon-gamma is produced by the CD4• lym ratio was 0 .24 (28 1 ) . Since these results are markedly
phocytes from patients with tuberculous pleuritis inferior to those with ADA or interferon-gamma,
(275) . The production of interferon-gamma appears the IGRA should not be used in assessing whether
to be a useful defense mechanism. Interferon-gamma patients have tuberculous pleuritis.
enhances polymyristate acetate-induced hydrogen
peroxide production in macrophages, facilitating
Polymerase Chain Reaction
elimination of intracellular parasites. This lympho
kine also inhibits mycobacterial growth in human The role of the polymerase chain reaction (PCR) and
monocytes (275) . other nucleic acid amplification (NAA)-based tests
1 12 PLE U RAL D I S EASES
Two reports have evaluated the diagnostic utility of found that the pleural fluid RF was elevated in 1 1
measuring the levels of different tuberculous anti of 1 1 seropositive patients with rheumatoid pleural
gens in the pleural fluid (292,293) . Although the effusions. In each patient, the pleural fluid RF titer
mean levels of tuberculous antigens were higher in was equal to or greater than 1 :320 and was equal
the pleural fluid of patients with tuberculous pleuritis to or greater than that in the serum. In view of the
than in the pleural fluid of other patients, there was last-mentioned study, I recommend that RF titers
so much overlap that the test was of little diagnostic be determined in pleural fluid when the diagnosis of
use. In a similar vein, there have been at least eight rheumatoid pleuritis is considered. The demonstra
separate reports (294-30 1 ) that have indicated that tion of a pleural fluid RF titer equal to or greater than
patients with tuberculous pleural effusions tend to 1 :320 and equal to or greater than the serum titer
have higher levels of specific antituberculous anti is strong evidence that the patient has a rheumatoid
bodies in their pleural fluid than do patients with pleural effusion.
other types of exudative effusions. The source of the
antibodies, however, is apparently the serum rather
than local antibody production in the pleural space Antinuclear Antibodies
in most instances (296) , although in one study, the
In the earlier editions of this book, I stated that mea
pleural fluid levels of lgM antibodies against myco
surement of the antinuclear antibody (ANA) levels in
bacterial antigen A60 were higher in the pleural fluid
pleural fluid appeared to be the best test for establish
than in the serum and were higher in patients with
ing the diagnosis of lupus pleuritis. Two subsequent
tuberculous pleuritis than in those without tubercu
studies (305 ,306) have cast doubt on this statement.
lous pleuritis (30 1 ) . If these results can be confirmed,
Khare et al. (305) measured pleural fluid ANA
this test may prove useful in identifying patients with
levels using the Hep-2 cell line on 82 pleural fluids
tuberculous pleuritis.
including 8 with SLE. Six of the eight patients (75%)
with SLE had high (> 1 :320) titers of ANA in their
I M M U N O LO G I C STU D I E S pleural fluid with a homogeneous staining pattern.
Because nearly 5% o f patients with RA (302) and In none did the pleural fluid ANA titer differ by
50% of patients with systemic lupus erythematosus more than one dilution from the serum titer. The
(SLE) have pleural effusions sometime during the other two patients with SLE had alternative explana
course of their disease, and because such effusions tions for their pleural effusions. However, 8 of the
may be present before the underlying disease is obvi remaining 74 patients ( 1 0.8%) had positive pleural
ous ( 1 29,302), it is important to consider these diag fluid ANA titer (> 1 :40) and 2 had a homogeneous
nostic possibilities in patients with exudative pleural pattern. In the patients with SLE, pleural fluid analy
effusions of undetermined origin. Numerous papers sis of anti-ssDNA, anti-dsDNA, anti-Sm, anti-SSA,
have assessed the diagnostic utility of various immu and anti-SSB antibodies reflected the findings in the
nologic measurements of the pleural fluid in estab serum (305) .
lishing these diagnoses. A more recent study evaluated the ANA titers in
1 26 pleural fluids using the HEP-2 cell line as sub
strate (306) . Again all of the pleural fluids from the
Rheumatoid Factor
patients with SLE had a high ANA titer (> 1 : 1 60),
Berger and Seckler (303) first reported that rheu but in this study, the pattern could be either homo
matoid factor (RF) was elevated in the pleural fluid geneous or speckled (306) . Moreover, pleural fluid
of patients with rheumatoid pleuritis. Subsequently, ANA titers were greater than 1 : 1 60 in 1 3 other
Levine et al. (304) studied pleural fluid RF levels in patients, 1 1 of whom had malignant pleural effu
65 patients with pleural effusions and found that sions. The pleural fluid and the serum ANA titers
4 1 % of patients with bacterial pneumonia and 20% were closely correlated in all patients. The explana
of patients with carcinoma had pleural fluid RF titers tion for the discrepancy between these two recent
equal to or greater than 1 : 1 60. In seven of the 65 studies and two older studies ( 1 30,307) , which sug
fluids, the pleural fluid RF titers were greater than gested that the pleural fluid ANA levels were very use
the serum titers, but the pleural fluid titer was greater ful, is not known but it may be related to the cell line
than 1 :640 in only one of the patients. These work used for the assays.
ers concluded that the RF titers in pleural fluid were Nevertheless, the two more recent studies suggest
not useful diagnostically. Halla et al. ( 1 29), however, that measurement of the pleural fluid ANA titer adds
1 14 PLE U RAL D I S EASES
essentially nothing to measuring the serum ANA titer this test. It should also be noted that false-positive LE
in the differential diagnosis of pleural effusions. cell results have been reported (309) .
FIGURE 7.6 • M u lt i n u c l eated tear d ro p-s h a ped ce l l from a patient with r h e u matoid p l e u ritis, w h i c h i s
t h o u g ht to be pathog n o m o n i c o f t h i s d i sease.
C H APT E R 7 I CLI N I CAL M A N I F E STAT I O N S A N D U S E F U L TESTS 1 15
L I P I D STU D I E S
regardless of whether CH50, C3, or C4 is measured. Pleural fluid is occasionally milky o r opalescent. This
Nevertheless, a CH 50 level below 1 0 U/mL (3 1 2) or opalescence is sometimes mistakenly attributed to
a C4 level below 1 0 times 1 0-5 U/g protein ( 1 29) is myriad WBCs in the pleural fluid, and the patient is
seen with most patients with either RA or SLE and treated for an empyema. This mistake is not made
rarely with other diseases. Because the serum ANA if the supernatant of the centrifuged pleural fluid
levels and the RF titers appear to be more specific is examined. In empyema, the supernatant is clear,
and more sensitive at identifying individuals with whereas in a chylous or chyliform effusion, the super
pleural effusions due to RA or SLE, measurement natant remains cloudy or milky. Some pleural fluids
of pleural fluid complement levels is not routinely with high lipid content also contain numerous RBCs,
recommended. and their color is accordingly red or brown. The
There have been several articles that have evalu supernatants of all pleural fluids should be examined
ated the diagnostic utility of complement activation for turbidity. Lipid studies of the pleural fluid should
products in pleural fluid. Two studies have demon be ordered whenever the supernatant is turbid.
strated that the pleural fluid levels of SC5b-9, which The persistent cloudiness of these pleural fluids
is a product of C3 activation, are elevated in tuber after centrifugation is due to their high lipid content,
culous pleural fluid as opposed to malignant pleu which can result from one of three mechanisms. First,
ral effusions or transudates (3 1 4,3 1 5) and that the the lymphatic duct may be disrupted so that chyle
SC5b-9 levels are highest in patients with rheumatoid accumulates in the pleural space. The patient is then
disease (3 1 5) . Another study (3 1 6) demonstrated that said to have a chylothorax, and the pleural effusion is
1 16 PLE U RAL D I S EASES
called a chylous pleural effusion. Second, large amounts glucose and the potassium levels are usually exceed
of cholesterol or lecithin-globulin complexes can ingly high if the fat emulsion is entering the pleural
accumulate for unknown reasons in the pleural fluid. space (32 1 ) .
The patient is then said to have a pseudochylotho Th e diagnosis o f chylothorax is best made by mea
rax and a chyliform pleural effusion. Although some suring the triglyceride levels in the pleural fluid. If the
authors have separated pseudochylothoraces into pleural fluid triglyceride level exceeds 1 1 0 mg/dL,
chyliform pleural effusions characterized by high the patient probably has a chylothorax; if the triglyc
lecithin-globulin levels and pseudochylous effusions eride level is below 50 mg/dL, the patient does not
characterized by the presence of cholesterol crystals have a chylothorax (Fig. 7.8) . If the triglyceride level is
(3 1 9 ,320) I see no advantage in making this differ between 50 and 1 1 0 mg/dL, the patient may or may
entiation. Third, if the patient is receiving parenteral not have a chylothorax (322) . When the diagnosis is
nutrition through a central line and the superior vena uncertain, lipoprotein analysis of the fluid should be
cava is perforated, the fat emulsion can collect in the ordered. The demonstration of chylomicrons in the
pleural space. It is important to distinguish these pleural fluid with lipoprotein analysis is diagnostic
three entities because the treatment is different for of chylothorax (322,323) . Most patients with chy
each one. lothoraces and pleural fluid triglyceride levels below
When milky pleural fluid is discovered, the first 1 1 0 mg/dL are malnourished. Except for diagnosing
question that should be asked is whether the patient chylothorax, pleural fluid triglyceride levels have no
is receiving parenteral nutrition through a central role in the diagnosis of pleural effusions (324) .
line. If the superior vena cava is perforated in such a Romero et al. (325) have suggested that the diag
setting, the intravenous infusion fluid may accumu nosis of chylothorax should also require a pleural
late rapidly in the pleural space (32 1 ) . If the patient is fluid-to-serum cholesterol ratio lower than 1 and a
receiving a lipid emulsion, the pleural fluid triglycer pleural fluid-to-serum triglyceride ratio less than 1 .
ide level can be very high. However, the pleural fluid The first requirement is reasonable because patients
•
• •
• •
1 000 •
. "' •
• •
•
• •
• •
•
=6 • •
'- ••
•
0) •
E •
•
•
a)
-c 0 • 0 0
·c:
110
0 0
Q) •
0 . 8 •
()
>-
• 0
•
oO 0 0 o'O 0 •
Ci 00 � 00
0
0 0
50
� 0 r:P' 0 0 0
o� 0 0
0 0
8 ( g oo oo
8 0 :Jo 0
0
Oo 0 0 • Chylous effusions
0 0
10 o Nonchylous effusions
50 1 00 1 50 200 250 300
Cholesterol, mg I dl
FIGURE 7.8 • Scatte rgram of c h o l este rol a n d trig lyce r i d e va l ues in p l e u r a l fl u i d . So l i d dots rep resent
patients whose eff us io n s conta i ned chy l o m icrons and were c o n s i d e red chylous. Open c i rcles represent
patients without chy l o m icrons wh ose effus i o n s were con s i d e red n o n c h y l o u s . All patie nts with trig lyce r i d e
l eve l s a bove 1 1 0 m g/d L h a d chylous effu s i ons, w h e reas n o p a t i e n t w i t h trig lyce ride l eve ls b e l ow S O mg/d L
h a d a chyl o u s effus i o n . (From Staats BA, Ellefson RD, Budahn LL, e t al. The lipoprotein profile o f chylous and nonchylous pleural
effusions. M ayo C l i n Proc. 1 980;55:700-704, with permission.)
C H APT E R 7 I CLI N I CAL M A N I F E STAT I O N S A N D U S E F U L TESTS 1 17
with chyliform pleural effusions often have pleural and faster results than do conventional methods. In
fluid cholesterol levels above 250 mg/dL (326) and one study, the median time for the BACTEC cultures
this requirement should eliminate them. It should be to become positive was 1 8 days (range 3-40 days) ,
noted that the clinical picture with chyliform effu whereas the median time for conventional cultures
sion of a long-standing pleural effusion with thick was 33.5 days (range 2 1 -48 days) (33 1 ) . Routine
ened pleura should be easy to differentiate from that smears for mycobacteria are not indicated because
of an acute pleural effusion with normal pleural sur they are almost always negative, unless the patient
faces seen with chylous effusion. If any doubt exists, has a tuberculous empyema or unless the patient is
lipoprotein analysis of the pleural fluid should be per HIV positive.
formed. The second criteria proposed by Romero et al. It has been suggested that microbiologic testing of
(325) is probably not necessary because in patients pleural fluid should be ordered more selectively (332) .
without chylothorax, there is no relationship between The basis for this recommendation was the observa
the serum and the pleural fluid triglyceride level, tion at the Mayo Clinic that only 1 5 of 476 patients
and the mean pleural fluid triglyceride level is only (3%) had positive cultures (332) . The results of the
approximately 25% of the serum level (32) . study by Ferrer et al. (333) tend to go against this
Other studies useful for diagnosing turbid pleural conclusion. They reviewed the results of 245 pleural
fluids are the total lipid content, the cholesterol con fluid specimens inoculated into blood culture vials
tent, and microscopic examination of the sediment. and found that 1 5% of the specimens were positive
Most pleural effusions that are cloudy secondary to and 60% of the positive specimens occurred in fluid
high lipid levels have a total fat content greater than that was nonpurulent (333) . Certainly, it is not cost
400 mg/dL (327) . The cholesterol levels in the pleu effective to obtain cultures from transudative pleural
ral fluid are elevated in high-lipid pleural effusions fluids, but I believe cultures should be obtained from
because of high numbers of cholesterol crystals or most patients with exudative effusions of unknown
lecithin-globulin complexes (320) , but cholesterol etiology.
levels may also be elevated in chylous pleural effu Culturing pleural fluid from chest tube drainage
sions (322) , as demonstrated in Figure 7.8. When the should be discouraged. Cultures from chest tubes
turbidity is due to high numbers of cholesterol crys yield inaccurate culture results when compared with
tals, examination of the pleural fluid sediment reveals direct aspirates (330) .
the cholesterol crystals, which are large, rhombic, or
polyhedric, as illustrated in Figure 26. 1 . Nucleic Acid Amplification (NAA)
It has been shown that using NAA to amplify and
M I C R O B I O LO G I C STU D I E S O N sequence the bacterial ribosomal RNA gene, that
P L E U RA L F L U I D the bacteria responsible for a complicated parapneu
Cultures monic effusion can frequently be identified (334) .
Maskell et al. (334) performed this procedure on
Pleural fluid from patients with undiagnosed exudative 404 pleural fluid specimens obtained during the First
pleural effusions should be cultured for bacteria (both Multicenter Intrapleural Sepsis Trial (335) . They
aerobically and anaerobically), mycobacteria, and fungi. reported that the NAA technique identified bacteria
For aerobic and anaerobic bacterial cultures, the pleural in 70 samples which were negative on culture (335).
fluid should be inoculated directly into blood culture The standard cultures and the nuclear amplification
media at the bedside because the number of positive technique were complimentary in identifying differ
cultures will increase with this method (328,329) . In ent organisms when they were both positive (334) .
one study (329) of 62 patients with suspected pleural
infection, the addition of the blood culture bottle cul
Countercurrent lmmunoelectrophoresis
ture to the standard culture increased the proportion
of patients with identifiable pathogens from 37.7% to The goal of countercurrent immunoelectrophoresis
58.5%. If the fluid is not inoculated at the bedside, it (CIE) is to identify bacterial antigens in pleural fluid,
should be sent to the laboratory in an anaerobic trans and to thereby establish a presumptive bacteriologic
port container at room temperature (330) . A Gram's diagnosis in patients with parapneumonic pleural
stain of the fluid should also be obtained. effusions. CIE depends on the interaction of an anti
For mycobacterial cultures, use of a BACTEC sys gen with a negative charge and a specific antibody
tem with bedside inoculation provides higher yields with a positive charge in an electrical field to form a
1 18 PLE U RAL D I S EASES
distinct precipitin line (336,337) . The advantage of Fibrinogen and Fibrin Degradation
CIE over bacterial cultures is that results are avail Products
able within hours rather than days, so that appropri
The fibrinogen levels in pleural fluid are low in com
ate antibiotics can be administered sooner. CIE is
parison to those in plasma (339,340) . The levels of
most useful in the diagnosis of pleural effusions in
fibrinogen tend to be low in patients with CHF and
children, in whom most pleural effusions are due to
malignancy, and high in patients with tuberculosis
bacterial infections with Streptococcus pneumoniae, S.
and empyema but there is much overlap. The level of
aureus, or Hemophilus influenza. These are the three
D-dimer is comparable in all pleural effusions (340) .
bacteria for which antigens are available to perform
CIE studies. In a series of 87 pediatric patients with
pleural effusions (336) , pleural fluid cultures were M I S C E L LA N E O U S T E STS O N
positive for one of these three bacteria in 34 patients, P L E U RA L F L U I D
and CIE correctly identified the offending organisms
Other Proteins
in 33 (97%) . In approximately 25% of those patients,
the Gram's stain of the pleural fluid was negative. In Numerous studies have evaluated the diagnostic
an additional 23 patients, CIE identified antigens in usefulness of measuring other proteins in the pleural
the pleural fluid when the pleural fluid cultures were fluid, including, ferritin (34 1 ) , mucoproteins ( 1 1 2) ,
negative (336) . fibronectin (342) , acid glycosaminoglycans (muco
Another advantage of CIE is that it remains posi polysaccharides) (343), /3 2-microglobulins (344) , and
tive for several days after antibiotic therapy is initiated a-fetoprotein (344) . The pleural fluid levels of these
(332,333) . To my knowledge, examination of pleural proteins are not useful in the differential diagnosis of
fluid with CIE has not been performed on a large series exudative pleural effusions.
of adult patients with pleural effusion. The disadvan
tage of CIE in the adult patient with a complicated
Other Enzyme Determinations
parapneumonic effusion is that many such effusions
are due to anaerobic bacteria ( 1 1 7) , and antigens for Many other enzymes, including aldolase (345), glu
all the anaerobic organisms are not yet available for tamic oxaloacetic transaminase, glutamic pyruvic
routine use. Certainly, if CIE is readily available, it transaminase (345), phosphohexose isomerase (345),
should be performed on the pleural fluid from patients malic dehydrogenase (345), isocitric dehydrogenase
with an acute febrile illness and pleural effusion. (345), glutathione reductase (345), alkaline phos
phatase (346) , angiotensin-converting enzyme (347) ,
and transketolase, have been measured in the pleural
Direct Gas-Liquid Chromatography
fluid and have been found to give no useful diagnos
Most anaerobic bacteria produce volatile fatty acids. tic information. One report suggested that elevation
Demonstration of such fatty acids by direct gas of the acid phosphatase level in the pleural fluid was
liquid chromatography of the pleural fluid has been diagnostic of metastatic prostatic carcinoma (348) ,
proposed as a means to establish the diagnosis of but a second report indicated that approximately
anaerobic pleural infections. In one report, pleural 1 0 % of all pleural effusions have acid phosphatase
fluid from 52 patients, including 1 4 with anaerobic levels above the upper normal limit for serum and
infections, were analyzed with direct gas-liquid chro higher than those in the corresponding serum (349) .
matography (338) . Multiple volatile fatty acids or
succinic acid was present in 13 of the 14 (93%) fluids
from patients with anaerobic infections. Succinic acid REFERENCES
was the major product in 1 0 patients, and 9 of them
1 . Mare! M, Stastny B, Melinova L, et al. Diagnosis of pleural
had infections due to Bacteroides. In contrast, pleural effusions-experience with clinical studies 1 9 86- 1 990. Chest.
fluid from other patients did not contain multiple 1 995; 1 07: 1 598-1 603.
volatile fatty acids or succinic acid (338) . Direct gas 2 . Anthonisen NR, Martin RR. Regional lung function in pleural
liquid chromatography is a difficult and expensive effusion. Am Rev Respir Dis. 1 977; 1 1 6:20 1 -207.
3. Light RW, Stansbury DW, Brown SE. The relationship between
procedure, however, and it requires specially trained
pleural pressures and changes in pulmonary function after ther
technicians. The resources required for gas-liquid apeutic thoracentesis. Am Rev Respir Dis. 1 9 86; 1 33:658-66 1 .
chromatography are probably better spent in upgrad 4 . Wang JS, Tseng CH. Changes i n pulmonary mechanics and
ing anaerobic culture techniques. gas exchange after thoracentesis on patients with inversion of
C H APT E R 7 / CLI N I CAL M A N I F E STAT I O N S A N D U S E F U L TESTS 1 19
a hemidiaphragm secondary to large pleural effusion. Chest. comparative analysis with other biochemical parameters.
1 995; 1 07: 1 6 1 0- 1 6 1 4. Respiration. 2002;69:20 1 -206.
5. Vaska K, Wann LS, Sagar K, et al. Pleural effusion as a 26. Uzun K, Vural H, Ozer F, et al. Diagnostic value of uric acid to
cause of right ventricular diastolic collapse. Circulation. differentiate transudates and exudates. Clin Chem Lab Med.
l 992;86:609-6 1 7. 2000;3 8:66 1 -665.
6. Brandstetter RD , Cohen RP. Hypoxemia after thoracente 27. Garcia-Pachon E, Padilla-Navas I, Sanchez JF, et al. Pleural
sis. A predictable and treatable condition. JAMA. 1 979;242: fluid to serum cholinesterase ratio for the separation of transu
1 060- 1 06 1 . dates and exudates. Chest. 1 996; 1 1 0 :97-1 0 1 .
7. Diaz-Guzman E , Budev MM. Accuracy o f the physical 28. Romero S , Candela A, Martin C , et al. Evaluation o f different
examination in evaluating pleural effusion. Cleve Clin ] Med. criteria for the separation of pleural transudates from exudates.
2008;75 :297-303. Chest. 1 993; 1 04:399-404.
8. Pavlin J, Cheney FW Jr. Unilateral pulmonary edema in 29. Burgess LJ, Maritz FJ, Taljaard JJ. Comparative analysis of the
rabbits after re-expansion of collapsed lung. ] Appl Physiol. biochemical parameters used to distinguish between pleural
1 979;46:3 1 -3 5 . transudates and exudates. Chest. 1 99 5 ; 1 07: 1 604- 1 609.
9. Gilbert VE. Shifting percussion dullness o f the chest: a sign o f 30. Vives M , Poree! JM, De Vera MV, et al. A study of Light's
pleural effusion. South Med] 1 997;90 : 1 25 5 - 1 2 5 6 . criteria and possible modifi c ations for distinguishing exuda
10. Bernstein A , White F Z . Unusual physical findings i n pleural tive from transudative pleural effusions. Chest. 1 996; I 09:
effusion: intrathoracic manometric studies. Ann Intern Med 1 5 03- 1 507.
1 9 52;37:733-73 8 . 3 1 . Gazquez I, Poree! JM, Vives M, et al. Comparative analysis
11. Paddock F K . The diagnostic significance of serous fluids in of Light's criteria and other biochemical parameters for dis
disease. N Engl] Med. 1 940;223: 1 0 1 0- 1 0 1 5 . tinguishing transudates from exudates. Respir Med. 1 998;
12. Broaddus VC, Light RW. What i s the origin o f pleural transu 92:762-765 .
dates and exudates [Editorial] ? Chest. 1 992; 1 02:658. 3 2 . Vaz MAC, Teixeira LR, Vargas F S , e t al. Relationship between
13. Romero-Candeira S, Hernandez L, Romero-Brufao S, et al. Is pleural fluid and serum cholesterol levels. Chest. 200 1 ;
it meaningful to use biochemical parameters to discriminate 1 1 9 :204-2 1 0 .
between transudative and exudative pleural effusions? Chest. 33. Heffner J E , Brown LK, Barbieri CA. Diagnostic value o f tests
2002; 1 22: 1 524- 1 529. that discriminate between exudative and transudative pleural
14. Leuallen EC, Carr OT. Pleural effusion, a statistical study of effusions. Chest. 1 997; 1 1 1 :970-980.
436 patients. N Engl] Med. 1 95 5 ;252:79-83. 34. Romero-Candeira S, Fernandez C, Martin C, et al. Influence
15. Carr OT, Power MH. Clinical value of measurements of of diuretics on the concentration of proteins and other com
concentration of protein in pleural fluid. N Engl ] Med. ponents of pleural transudates in patients with heart failure.
1 9 5 8 ;259:926-927. Am ] Med. 200 1 ; 1 1 0:68 1-686.
16. Light RW, MacGregor Ml, Luchsinger PC, et al. Pleural effu 35. Bielsa S, Poree! JM, Castellote ], et al. Solving the Light's cri
sions: the diagnostic separation of transudates and exudates. teria misclassification rate of cardiac and hepatic transudates.
Ann intern Med. 1 972;77:5 07-5 1 3 . Respirology. 20 1 2; 1 7:72 1 -726.
17. Hamm H, Brohan U , Bohmer R , e t al. Cholesterol i n pleural 36. Heffner JE, Sahn SA, Brown LK. Multilevel likelihood
effusions: a diagnostic aid. Chest. 1 987;92:296-302. ratios for identifying exudative pleural effusion. Chest. 2002;
18. Valdes L, Pose A, Suarez J, et al. Cholesterol: a useful parameter 1 2 1 : 1 9 1 6-1 920.
for distinguishing between pleural exudates and transudates. 37. Heffner JE, Highland K, Brown LK. A meta-analysis deriva
Chest. 1 99 1 ;99: 1 097- 1 1 02. tion of continuous likelihood ratios for diagnosing pleural fluid
19. Costa M, Quiroga T, Cruz E. Measurement of pleural fluid exudates. Am ] Respir Grit Care Med. 2003; 1 67: 1 5 9 1- 1 599.
cholesterol and lactate dehydrogenase. A simple and accurate 3 8 . Paddock FK. The relationship between the specific gravity and
set ofindicators for separating exudates from transudates. Chest. the protein content in human serous effusions. Am J Med Sci.
1 995; I 08: 1 260- 1 263. 1 94 1 ;20 1 : 5 69-574.
20. Roth BJ, O'Meara TF, Cragun WH . The serum-effusion 39. Light RW Falsely high refractometric readings for the specific
albumin gradient in the evaluation of pleural effusions. Chest. gravity of pleural fluid. Chest. 1 979;76:300-30 1 .
l 990;98: 546-549. 40. Light RW, Ball WC. Lactate dehydrogenase isoenzymes in
21. Meisel S, Shamiss A, Thaler M, et al. Pleural fluid to serum pleural effusions. Am Rev Respir Dis. 1 973; 1 0 8 :660-664.
bilirubin concentration ratio for the separation of transudates 4 1 . Light RW, Erozan YS, Ball WC. Cells in pleural fluid: their value
from exudates. Chest. 1 990;9 8 : 1 4 1 - 1 44 . in differential diagnosis. Arch Intern Med 1 973; 1 32:854-860.
22. Yetkin 0, Tek I, Kaya A , et al. A simple laboratory mea 42. Light RW, Ball WC. Glucose and amylase in pleural effusions.
surement for discrimination of transudative and exudative JAMA. 1 973;225 :257-260.
pleural effusion: pleural viscosity. Respir Med. 2006; I 00: 43. Light RW, MacGregor Ml, Ball WC Jr, et al. Diagnostic signif
1 286- 1 290. icance of pleural fluid pH and Pco , . Chest. 1 973;64 : 5 9 1 -596.
23. Papageorgiou E, Kostikas K, Kiropoulos T, et al. Increased 44. Rolf LL, Travis OM. Pleural fluid-plasma bicarbonate gra
oxidative stress in exudative pleural effusions: a new marker dients in oxygen-toxic and normal rats. Am ] Physiol. 1 973;
for the differentiation between exudates and transudates? 224:857-86 1 .
Chest. 200 5 ; 1 28 :329 1 -3297. 4 5 . Pfister R, Schneider CA. Natriuretic peptides BNP and
24. Horvath LL, Gallup RA, Worley BO, et al. Soluble leuko NT-pro-BNP: established laboratory markers in clinical prac
cyte selectin in the analysis of pleural effusions. Chest. 200 I ; tice or j ust perspectives. Clin Chim Acta. 2004;349 :25-3 8 .
1 20:362-368. 46. Poree! JM. The use of probrain natriuretic peptide i n pleural
25. Alexandrakis MG, Kyriakou 0, Alexandraki R, et al. Pleural fluid for the diagnosis of pleural effusions resulting from heart
Interleukin- I beta in differentiating transudates and exudates: failure. Curr Opin Pulm Med. 2005; 1 1 :329-333.
1 20 PLE U RAL D I S EASES
47. Poree! JM, Vives M, Cao G, et al. Measurement of pro-brain late eosinophil accumulation in rats. Eur J Pharmacol. 1 998;
nacriurecic peptide in pleural fluid for the diagnosis of pleural 356:239-243.
effusions due co heart failure. Am J Med. 2004; 1 1 6:4 1 7-420. 70. Castro-Faria-Nero HC, Penido CM, Larangeira AP, et al.
48. Liao H, Na MJ, Dikensoy 0, et al. The diagnostic value of A role for lymphocytes and cycokines on the eosinophil
pleural fluid NT-pro-BNP levels in patients with cardiovascu migration induced by LPS . Mem Inst Oswaldo Cruz. 1 997;
lar diseases. Respirology. 2008 ; 1 3 : 5 3-57. 92 (suppl 2) : 1 97-200.
49. Poree! JM. Utilization of B-type natriurecic peptide and 71. Penido C, Castro-Faria-Nero HC, Larangeira AP, et al. The
NT-pro-BNP in the diagnosis of pleural effusions due co heart role of gamma-delta T lymphocytes in lipopolysaccharide
failure. Curr Opin Pulm Med. 20 1 1 ; 1 7:2 1 5-2 1 9 . induced eosinophil accumulation into the mouse pleural cavity.
5 0 . Poree! JM, Martinez-Alonso M, Cao G, et al. Biomarkers of ] Immunol. 1 997; 1 5 9 : 8 53-860.
heart failure in pleural fluid. Chest. 2009 ; 1 36:67 1 -677. 72. Pasquale CP, Martins MA, Bozza PT, et al. Bradykinin induces
5 1 . Long AC, O'Neal HR Jr, Peng S, et al. Comparison of pleural eosinophil accumulation in the rat pleural cavity. Int Arch
fluid N-terminal pro-brain natriuretic peptide and brain Allergy Appl Immunol. 1 9 9 1 ;95:244-247.
natriuretic-32 peptide levels. Chest. 20 1 0; 1 37: 1 369-1 374. 73. Nakamura Y, Ozaki T, Kamei T, et al. Factors chat stimulate
52. Janda S, Swiston J. Diagnostic accuracy of pleural fluid the proliferation and survival of eosinophils in eosinophilic
NT-pro-BNP for pleural effusions of cardiac origin: A system pleural effusion: relationship co granulocyte-macrophage
atic review and meta-analysis. BMC Pulm Med. 20 1 0; 1 0 : 5 8 . colony-stimulating factor, interleukin-5, and interleukin-3 .
5 3 . Han C H , Choi J E , Chung JH. Clinical utility of pleural fluid Am J Respir Cell Mo! Biol. 1 993;8:605-6 1 1 .
NT-pro brain natriuretic peptide (NT-pro-BNP) in patients 74. Schandene L , Namias B , Crusiaux A, et al. IL-5 i n post
with pleural effusions. Intern Med. 2008;47: 1 669- 1 674. traumatic eosinophilic pleural effusion. Clin Exp Immunol.
54. Kolditz M, Halank M, Schiemanck S, et al. High diagnostic 1 993;93: 1 1 5-1 1 9 .
accuracy of NT-pro-BNP for cardiac origin of pleural effu 75. Mohamed KH, Abdelhamid AI, Lee YC, e t al. Pleural fluid
sions. Eur Respir J 2006;28 : 1 44- 1 5 0 . levels of interleukin-5 and eosinophils are closely correlated.
5 5 . Rojas-Solano J R , Light RW, Brenes-Dietel A. Black pleural Chest. 2002; 1 22 : 576-5 80.
fluid. Arch Bronconeumol. 2009;4 5 : I 03- 1 04 . 76. Kalomenidis I, Mohamed KH, Lane KB, et al. Pleural fluid
56. Liao WC, Chen CH, Tu C Y. Black pleural effusion in mela levels of vascular cell adhesion molecule- I are elevated in
noma. CMAJ 2009; 1 82:E3 1 4 . eosinophilic pleural effusions. Chest. 2003; 1 24 : 1 5 9- 1 66.
57. Lyche KO , Jensen WA, Kirsch C M , et al. Pleuropulmo 77. Nakamura Y, Ozaki T, Yanagawa H, et al. Eosinophil colony
nary manifestations of hepatic amebiasis. West J Med. 1 990; stimulating factor induced by administration ofinterleukine-2
1 5 3 :275-278. into the pleural cavity of patients wirh malignant pleurisy. Am
5 8 . Dikensoy 0 , Stathopoulos GT, Zhu Z, et al. 0-dimer levels in ] Respir Cell Mo! Biol. 1 990;3:29 1 -300.
pleural effusions. Respir Med. 2006; 1 00: 1 337- 1 34 1 . 78. Bozza PT, Castro-Faria-Nero HC, Penido C, et al. IL-5 accounts
5 9 . Conner BO, Lee YCG, Branca P, et al. Variations in pleural fluid for rhe mouse pleural eosinophil accumulation triggered by anti
WBC count and differential counts with different sample con gen but not by LPS. Immunopharmacology. 1 994;27: 1 3 1 - 1 36.
tainers and different methods. Chest. 2003 ; 1 23 : 1 1 8 1 - 1 1 87. 79. Adelman M, Albelda SM, Gocclieb J, et al. Diagnostic utility
60. Aulesa C, Mainar I, Prieto M , et al. Use of the Advia 1 2 0 of pleural fluid eosinophilia. Am J Med. l 984;77: 9 1 5-920.
hematology analyzer i n the differential cycologic analysis o f 80. Askin FB, McCann BG, Kuhn C. Reactive eosinophilic pleu
biological fluids (cerebrospinal, peritoneal, pleural, pericar ritis. Arch Pathol Lab Med. 1 977; 1 0 1 : 1 87- 1 9 1 .
dia!, synovial, and others) . Lab Hematol. 2003;9:2 1 4-224. 81. Haro-Estarriol M , Alvarez-Castillo LA, Baldo-Padro X, et al.
6 1 . Spriggs AI, Boddingcon MM. The Cytology of Effesions, Influence of choracentesis and pleural biopsy on biochemical
2nd ed. New York, NY: Grune & Scraccon; l 968. parameters and cytology of pleural fluid. Arch Bronconeumol.
62. Broaddus VC, Hebert CA, Vitangcol RV, et al. lnrerleukin-8 2007;43:277-282.
is a major neutrophil chemotactic factor in pleural liquid of 82. Smit HJ, van den Heuvel MM, Barbieraco SB, et al. Analysis
patients wirh empyema. Am Rev Respir Dis. 1 992; 1 46: 825-830. of pleural fluid in idiopathic spontaneous pneumorhorax; cor
63. Antony VB, Godbey SW, Kunkel SL, et al. Recruitment of relation of eosinophil percentage with the duration of air in
inflammatory cells co rhe pleural space. Chemotactic cyco the pleural space. Respir Med. 1 999;93:262-267.
kines, IL-8, and monocyte chemotactic peptide- I in human 83. Kalomenidis I, Guo Y, Peebles RS, et al. Pneumothorax-asso
pleural fluids. ] Immunol. 1 993; 1 5 1 :72 1 6-7223. ciaced pleural eosinophilia in mice is interleukin-5 but not
64. Krenke R, Nasilowski J, Korczynski P, et al. Incidence and interleukin- 1 3 dependent. Chest. 2005 ; 1 28 :2978-2983.
etiology of eosinophilic pleural effusion. Eur Respir J 2009; 84. Maltais F, Laberge F, Cormier Y. Blood hypereosinophilia in
34: 1 1 1 1 - 1 1 1 7 . the course of post-traumatic pleural effusion. Chest. 1 990;
6 5 . Ferreiro L , San Jose E , Gonzalez-Barcala FJ, e t al. Eosinophilic 98 :348-3 5 1 .
pleural effusion: incidence, etiology and prognostic signifi 85. Romero Candeira S, Hernandez Blasco L , Soler MJ, e t al. Bio
cance. Arch Bronconeumol. 2 0 1 1 ;47;5 04-5 09. chemical and cycologic characteristics of pleural effusions sec
66. Naylor B, Novak PM. Charcoc-Leyden crystals in pleural fluids. ondary co pulmonary embolism. Chest. 2002; 1 2 1 :465-469.
Acta Cytol. 1 98 5 ;29:78 1-784. 86. Kalomenidis I, Stathopoulos GT, Barnette R, et al. Eotaxin-3
67. Wahl RW Curschmann's spirals in pleural and peritoneal and interleukin-5 pleural fluid levels are associated wirh pleu
fluids. Report of l 2 cases. Acta Cytol. 1 986;30 : 1 47- 1 5 1 . ral fluid eosinophilia in pose-coronary artery bypass grafting
6 8 . Klein A, Talvani A, Cara DC, et al. Stem cell factor plays a pleural effusions. Chest. 2005 ; 1 27:2094-2 1 00.
major role in the recruitment of eosinophils in allergic pleu 87. Kalomenidis I, Light RW. Eosinophilic pleural effusions. Curr
risy in mice via the production of leukotriene B 4 ' J Immunol. Opin Pulm Med. 2003;9:254-260.
2000; 1 64:427 1 -4276. 88. Philit F, Etienne-Mastroianni B, Parrot A, et al. Idiopathic
69. Perez S, Machado J, Cordeiro R, et al. Inhibition by the anti acute eosinophilic pneumonia: a study of 22 patients. Am J
micotic drug doxorubicin of platelet-activating-faccor-induced Respir Crit Care Med. 2002; 1 66: 1 235-1 239.
C H APT E R 7 / CLI N I CAL M A N I F E STAT I O N S A N D U S E F U L TESTS 121
89. Yacoubian HD. Thoracic problems associated with hydatid 1 1 0. Cailhier JF, Sawatzky DA, Kipari T, e t al. Resident pleural
cyst of the dome of the liver. Surgery. 1 976;79 :544-548. macrophages are key orchestrators of neutrophil recruit
90. Erzurum SE, Underwood GA, Hamilos DL, et al. Pleu ment in pleural inflammation. Am J Respir Grit Care Med.
ral effusion in Churg-Strauss syndrome. Chest. 1 989;95: 2006; 1 73 : 540-547.
1 3 57- 1 3 59. 1 1 1 . Gjomarkaj M, Pace E, Melis M, et al. Dendritic cells with
91. Okimoro N, Kurihara T, Honda Y, et al. Cause of basophilic a potent accessory activity are present in human exudative
pleural effusion. South Med]. 2003;96 :726-727. malignant pleural effusions. Eur Respir]. 1 997; 1 0 : 5 92-597.
92. Yam LT. Diagnostic significance of lymphocytes in pleural 1 1 2. Zinneman HH, Johnson JJ, Lyon RH . Proteins and muco
effusions. Ann Intern Med. 1 967;66:972-9 82. proteins in pleural effusions. Am Rev Tuberc Pulmon Dis.
93. Sadikot RT, Rogers JT, Cheng D-S, et al. Pleural fluid 1 957;76:247-2 5 5 .
characteristics of patients with sympromatic pleural effu 1 1 3 . Telvi L , Jaybert F, Eyquem A , e t al. Study ofimmunoglobulins
sion post coronary artery bypass surgery. Arch Intern Med. in pleura and pleural effusions. Thorax. l 979;34:389-392.
2000; 1 60 :2665-2668. 1 1 4. Yokogawa M, Kojima S, Araki K, et al. Immunoglobulin
94. Pettersson T, Klockars MD, Hellstrom P-E, et al. T and E: raised levels in sera and pleural exudates of patients with
B lymphocytes in pleural effusions. Chest. 1 978 ;73:49-5 1 . paragonimiasis. Am ] Trop Med Hyg. 1 976;2 5 : 5 8 1-586.
9 5 . Potrykus AM , Steinmann G , Stein E , et al. T- and B-cell 1 1 5 . Nash DR, Wallace RJ Jr. Immunoglobulin E and other
responses in patients with malignant pleural effusions. Br J immunoglobulins in patients with eosinophilic pleural effu
Cancer. 1 9 8 1 ;43:47 1 -477. sions. ] Lab Clin Med. 1 98 5 ; 1 06 : 5 1 2-5 1 6 .
96. Domagala W, Emeson EE, Kos LG. T and B lymphocyte 1 1 6. Vianna NJ. Non tuberculous bacterial empyema i n patients
enumeration in the diagnosis of lymphocyte-rich pleural with and without underlying diseases. JAMA. 1 97 1 ;
fluids. Acta Cytol. 1 98 1 ;25 : 1 08-1 1 0 . 2 1 5 :69-7 5 .
9 7 . Moisan T, Chandrasekhar AJ, Robinson J, e t al. Distribution 1 1 7. Light RW, Girard WM, Jenkinson SG, et a l . Parapneumonic
of lymphocyte subpopulations in patients with exudative effusions. Am ] Med. 1 9 80;69: 507-5 1 1 .
pleural effusions. Am Rev Respir Dis. 1 978; 1 1 7: 5 07-5 1 1 . 1 1 8 . Light RW, Rodriguez RM . Management of parapneumonic
98. Kockman S, Bernard J, Lavaud F, et al. T-lymphocyte sub effusions. Clin Chest Med. 1 998; 1 9 :373-3 82.
sets in pleural fluids: discrimination according ro traditional 1 1 9. Berger HW, Maher G . Decreased glucose concentration
and monoclonal antibody-defined markers. Eur J Respir Dis. in malignant pleural effusions. Am Rev Respir Dis. 1 97 1 ;
1 984;6 5 : 5 86-5 9 1 . 1 03 :427-429.
99. Lucivero G , Pierucci G , Bonomo L . Lymphocyte subsets 120. Rodriguez-Panadero F, Lopez Mejias J . Low glucose and
in peripheral blood and pleural fluid. Eur Respir ]. 1 988; pH levels in malignant pleural effusions. Am Rev Respir Dis.
1 :337-340. 1 989; 1 39:663-667.
1 00. Guzman J, Bross KJ, Wurtemberger G, et al. Immunocyrol 1 2 1 . Sahn SA, Good JT Jr. Pleural fluid pH in malignant effu
ogy in malignant pleural mesothelioma: expression of tumor sions. Ann Intern Med. 1 9 88; 1 08:345-349.
markers and distribution of lymphocyte subsets. Chest. 1 22. Sanchez-Armengol A, Rodriguez-Panadero F. Survival and
1 989;95: 5 90-5 9 5 . talc pleurodesis in metastatic pleural carcinoma, revisited.
1 0 1 . Dalbeth N, Gundle R , Davies RJO, e t a l . CD5 6b"gh' NK Report of 1 2 5 cases. Chest. 1 993; 1 04: 1 482- 1 4 8 5 .
cells are enriched at inflammarory sites and can engage with 1 2 3 . Rodriguez-Panadero F, Lopez-Mejias J. Survival time o f
monocytes in a reciprocal program of activation. J lmmunol. patients with pleural metastatic carcinoma predicted by
2004; 1 73 : 64 1 8-6426. glucose and pH studies. Chest. 1 9 89;95:320-324.
1 02. Okubo Y, Nakata M, Kuroiwa Y, et al. NK cells in carcino 1 24. Calnan WL, Winfield BJO, Crowley MF, et al. Diagnostic
marous and tuberculous pleurisy: phenotypic and functional value of the glucose content of serous pleural effusions.
analyses of NK cells in peripheral blood and pleural effu Br Med]. 1 9 5 1 ; 1 : 1 239-1 240.
sions. Chest. 1 9 87;92: 500-504. 1 2 5 . Barber LM, Mazzadi L, Deakins DO, et al. Glucose level in
1 03 . Dalbeth N, Lee YC. Lymphocytes in pleural disease. Curr pleural fluid as a diagnostic aid. Dis Chest. 1 9 57;3 1 :680-68 1 .
Opin Pulm Med. 2005; 1 1 :334-339. 1 26. Carr OT, Power MH. Pleural fluid glucose with special
1 04. Miserocchi G, Agostoni E. Contents of the pleural space. reference ro its concentration in rheumatoid pleurisy with
J Appl Physiol. 1 971 ;30:208-2 1 3. effusion. Dis Chest. 1 960;37:32 1 -324.
105. Hurwitz S, Leiman G, Shapiro C. Mesothelial cells in pleural 1 27. Lillington GA, Carr OT, Mayne JG. Rheumatoid pleurisy
fluid: TB or not TB? S A.fr Med]. 1 9 80;57:937-939. with effusion. Arch Intern Med. 1 97 1 ; 1 28:764-768.
1 06. Jones D, Lieb T, Narita M, et al. Mesothelial cells in tuber 128. Dodson WH, Hollingsworth JW Pleural effusion in rheu
culous pleural effusions of HIV-infected patients. Chest. matoid arthritis. N Engl] Med. 1 966;27 5 : 1 337- 1 342.
2000; 1 1 7:289-29 1 . 129. Halla JT, Schrohenloher RE, Volanakis JE. Immune
1 07. Antony VB, Sahn SA, Antony AC, et al. Bacillus Calmette complexes and other laboratory features of pleural effusions.
Guerin-stimulated neutrophils release chemotaxins for Ann Intern Med. 1 9 80;92:748-752.
monocytes in rabbit pleural space in vitro. J Clin Invest. 1 3 0. Good JT Jr, King TE, Antony VB, et al. Lupus pleuritis:
1 9 8 5;76: 1 5 1 4-1 52 1 . clinical features and pleural fluid characteristics with spe
1 0 8 . Gjomarkaj M , Pace E, Melis M , et al. Mononuclear cells in cial reference to pleural fluid antinuclear antibodies. Chest.
exudative malignant pleural effusions. Characterization of 1 983;84:7 1 4-7 1 8 .
pleural phagocytic cells. Chest. 1 994; 1 06: 1 042- 1 049. 1 3 1 . Branca P, Rodriguez RM , Rogers JT, e t al. Routine measure
1 09. Risberg B, Davidson B, Nielsen S, et al. Detection of ment of pleural fluid amylase is not indicated. Arch Intern
monocyte/macrophage cell populations in effusions: a com Med. 200 1 ; 1 6 1 :228-232 .
parative study using flow cytometric immunophenoryp 1 32. Lankisch PG, Groge M, Becher R. Pleural effusions: a new
ing and immunocyrochemistry. Diagn Cytopathol. 200 1 ; negative prognostic parameter for acute pancreatitis. Am J
2 5 : 2 1 4-2 1 9 . Gastroenterol. 1 994;89: 1 849- 1 8 5 1 .
1 22 PLE U RAL D I S EASES
1 33 . Kaye MD. Pleuropulmonary complications of pancreatitis. during the first hour following thoracentesis? Chest. 2000;
Thorax. 1 968;23:297-306. 1 1 7: 1 043- 1 048.
1 34. Rockey DC, Cello JP. Pancreaticopleural fistula. Report of 7 1 56. Goldstein LS, McCarthy K, Mehta AC, et al. Is direct
cases and review of the literature. Medicine. 1 990;69:332-344. collection of pleural fluid into a heparinized syringe impor
1 3 5 . Pottmeyer EW III, Frey CF, Matsuno S . Pancreaticopleural tant for determination of pleural pH? A brief report. Chest.
fistulas. Arch Surg. 1 987; 1 22:648-654. 1 997; 1 1 2:707-708.
1 36. Ende N. Studies of amylase activity and in pleural effusions 1 57. Jimenez Castro D, Diaz G , Perez-Rodriguez E, et al. Modi
and ascites. Cancer. 1 960; 1 3 :283-287. fication of pleural fluid pH by local anesthesia. Chest.
1 37. Kramer MR, Cepero RJ, Pitchenik AE. High amylase in 1 999; 1 1 6:399--402.
neoplasm-related pleural effusion. Ann Intern Med. 1 989; 1 58. Colice GL, Curtis A, Deslauriers J, et al. Medical and surgi
1 1 0: 5 67-5 69. cal treatment of parapneumonic effusions: an evidence-based
138. Abbott OA, Mansour KA, Logan WC, et al. Atraumatic guideline. Chest. 2000; 1 1 8 : 1 1 5 8-1 1 7 1 .
so-called "spontaneous" rupture of the esophagus. ] Thorac 1 5 9. Pine JR, Hollman JL. Elevated pleural fluid p H i n Proteus
Cardiovasc Surg. 1 970; 59:67-83 . mirabilis empyema. Chest. 1 983;84: 1 09-1 1 1 .
1 39. Sherr HP, Light RW, Merson MH, e t al. Origin of pleu 1 60. Poccs DE, Levin DC, Sahn SA. Pleural fluid p H in parapneu
ral fluid amylase in esophageal rupture. Ann Intern Med. monic effusions. Chest. 1 976;70 :328-33 1 .
1 972;76:985-9 86. 161. Maskell NA, Gleeson FV, Darby M , et al. Diagnostically
1 4 0 . Maulitz RM, Good JT Jr, Kaplan RL, et al. The pleuro significant variations in pleural fluid pH in loculated para
pulmonary consequences of esophageal rupture: an experi pneumonic effusions. Chest. 2004; 1 26:2022-2024.
mental model. Am Rev Respir Dis. 1 979; 1 20:363-367. 1 62. Dye RA, Laforet EG. Esophageal rupture: diagnosis by
1 4 1 . Wroblewski F, Wroblewski R. The clinical significance of pleural fluid pH. Chest. 1 974;66:454--456.
lactic dehydrogenase activity of serous effusions. Ann Intern 1 63 . Good JT Jr, Taryle DA, Sahn SA. The pathogenesis of the
Med. 1 9 5 8 ;48 : 8 1 3-822. low pleural fluid pH in esophageal rupture. Am Rev Respir
1 42. Raabo E, Rasmussen KN, Terkildsen TC. A study of the iso Dis. 1 983; 1 27:702-704.
enzymes of lactic dehydrogenase in pleural effusions. Scand] 1 64. Good JT Jr, Taryle DA, Maulitz RM , et al. The diagnostic
Respir Dis. 1 966;47: 1 5 0- 1 56. value of pleural fluid pH. Chest. 1 980;78 : 5 5-59.
1 43 . Lossos IS, Incrator 0, Berkman N, et al. Lactate dehydroge 165. Rodriguez-Panadero F, Lopez-Mejias L. Survival time of
nase isoenzyme analysis for the diagnosis of pleural effusion in patients with pleural metastatic carcinoma predicted by
haemato-oncological patients. Respir Med. 1 999;93:338-34 1 . glucose and pH studies. Chest. l 989;95:320-324.
1 44. Potts DE, Willcox MA, Good JT Jr, et al. The acidosis of 1 66. Johnson RJ, Johnson JR. Paragonimiasis in Indochinese ref
low-glucose pleural effusions. Am Rev Respir Dis. 1 978; ugees: roentgenographic findings with clinical correlations.
1 1 7:665-67 1 . Am Rev Respir Dis. 1 9 83; 1 2 8 : 5 34-5 3 8 .
1 4 5 . Chavalittamrong B , Angsusingha K, Tuchinda M , et al. Diag 1 67. Miller KS, Wooten S, Sahn SA. Urinothorax: a cause oflow pH
nostic significance of pH, lactic acid dehydrogenase, lactate transudative pleural effusions. Am ] Med. l 988;85 :448--449 .
and glucose in pleural fluid. Respiration. 1 979;38: 1 1 2- 1 20. 1 68 . Antonangelo L, Vargas FS, Acencio MM, et al. Effect of tem
1 46. Light RW, Luchsinger P. Metabolic activity of pleural fluid. perature and storage time on cellular analysis of fresh pleural
]Appl Physiol. 1 973;34:97- 1 0 1 . fluid samples. Cytopathology. 2 0 1 2;23: 1 03-1 07.
1 47 . Taryle DA, Good J T Jr, Sahn SA. Acid generation by pleural 1 69. Swiderek J, Morcos S, Donchireddy V, et al. Prospective
fluids: possible role in the determination of pleural fluid pH. study to determine the volume of pleural fluid required to
] Lab Clin Med. 1 979;93 : 1 04 1 - 1 046. diagnose malignancy. Chest. 20 1 0 ; 1 37:68-73.
148. Chandler TM, McCoskey EH, Byrd RP Jr, et al. Com 1 70. Marchevsky AM, Hauptman E, Gil J, et al. Computerized
parison of the use and accuracy of methods for determining interactive morphometry as an aid in the diagnosis of pleural
pleural fluid pH. South Med] 1 999;92:2 1 4-2 1 7 . effusions. Acta Cytol. 1 987;3 1 : 1 3 1 - 1 36.
1 49. Kohn G L , Hardie WD . Measuring pleural fluid pH: high 171. Stevens MW, Leong AS , Fazzalari NL, et al. Cytopathology
correlation of a handheld unit to a traditional tabletop blood of malignant mesothelioma: a stepwise logistic regression
gas analyzer. Chest. 2000; 1 1 8 : 1 626-1 629. analysis. Diagn Cytopathol. l 992;8:333-342.
1 5 0. Bowling M, Lenz P, Chatterjee A, et al. Perception versus 1 72. Jarvi OH, Kunnas RJ, Laitio MT, et al. The accuracy and
reality: the measuring of pleural fluid pH in the United significance of cytologic cancer diagnosis of pleural effusions.
States. Respiration. 20 1 2;83:3 1 6-322. Acta Cytol. 1 972; 1 6 : 1 52- 1 57.
1 5 1 . Cheng DS, Rodriguez RM , Rogers J, et al. Comparison of 1 73 . Grunze H. The comparative diagnostic accuracy, efficiency
pleural fluid pH values obtained using blood gas machine, pH and specificity of cytologic techniques used in the diagnosis
meter, and pH indicator scrip. Chest. 1 998 ; 1 1 4 : 1 368-1 372. of malignant neoplasm in serous effusions of the pleural and
1 52. Lesho EP, Roth BJ . Is pH paper an acceptable, low-cost alter pericardia! cavities. Acta Cytol. 1 964;8 : 1 50- 1 64.
native to the blood gas analyzer for determining pleural fluid 1 74. Bueno CE, Clemente G, Castro BC, et al. Cycologic and
pH? Chest. 1 997; 1 1 2 : 1 29 1 - 1 292. bacteriologic analysis of fluid and pleural biopsy specimens
1 53 . Bowling MR, Chatterjee A, Conforti J, et al. Perceptions vs. with Cope's needle. Arch Intern Med. 1 990; 1 5 0 : 1 1 90-1 1 94.
reality: measuring of pleural fluid pH in North Carolina. 175. Naylor B, Schmidt RW. The case for exfoliative cytology of
N C Med] 2009;70:9- 1 3 . serous effusions. Lancet. 1 964; 1 :7 1 1 -7 1 2 .
1 54. Rahman N M , Mishra EK, Davies H E , e t al. Clinically 1 76. Melamed M R . The cytological presentation o f malignant
important factors influencing the diagnostic measurement lymphomas and related diseases in effusions. Cancer. 1 963;
of pleural fluid pH and glucose. Am J Respir Crit Care Med. 1 6:4 1 3--43 1 .
2008 ; 1 78:483--490. 1 77. Dekker A, Bupp PA. Cytology o f serous effusions. An inves
1 5 5 . Sarodia BD, Goldstein LS, Laskowski DM, et al. Does tigation into the usefulness of cell blocks versus smears. Am J
pleural fluid pH change significantly at room temperature Clin Pathol. 1 978;70 : 8 5 5-860.
C H APT E R 7 I CLI N I CAL M A N I F E STAT I O N S A N D U S E F U L TESTS 1 23
1 7 8 . Filie AC, Copel C, Wilder AM, et al. Individual specimen triage 1 97. Coleman M, Henderson D W, Mukherjee T M . The ulrra
of effusion samples: an improvement in the standard of prac structural pathology of malignant pleural mesothelioma.
tice, or a waste of resources? Diagn Cytopathol. 2000;22:7- 1 0 . Pathol Annu. 1 9 89;24:303-3 5 3 .
1 79. Kielhorn E, Schofield K , Rimm D L . Use of magnetic enrich 1 98 . Jandik WR, Landas S K , Bray CK, et al. Scanning electron
ment for detection of carcinoma cells in fluid specimens. microscopic distinction of pleural mesotheliomas from ade
Cancer. 2002;94:205-2 1 1 . nocarcinomas. Mod Pathol. 1 993;6:76 1 -764.
1 8 0. Huang M-S, Tsai M-S, Hwang J-J, et al. Comparison of 1 99 . Yang GC. Long microvilli of mesothelioma are conspicu
nucleolar organiser regions and DNA flow cytometry in the ous in pleural effusions processed by Ultrafast Papanicolaou
evaluation of pleural effusion. Thorax. 1 994;49: 1 1 52-1 1 56. stain. Cancer. 2003;99 : 1 7-22.
1 8 1 . Ong KC, Indumathi V, Poh WT, et al. The diagnostic yield 200. Warnock ML, Stoloff A, Thor A. Differentiation of adenocar
of pleural fluid cytology in malignant pleural effusions. cinoma of the lung from mesothelioma. Periodic acid-Schiff,
Singapore Med]. 2000;4 1 : 1 9-23. monoclonal antibodies B72.3, and Leu M l . Am J Pathol.
1 82. Sujathan K, Kannan S, Pillai KR, et al. Significance ofAgNOR 1 988; 1 3 3 :30-38.
count in differentiating malignant cells from reactive meso 20 1 . Miedouge M, Rouzaud P, Salama G , et al. Evaluation of
thelial cells in serous effusions. Acta Cytol. 1 996;40:724-728 . seven tumour markers in pleural fluid for the diagnosis of
1 83 . Antonangelo L , Saldiva P H , Amaro Junior E, e t al. Utility o f malignant effusions. Br] Cancer. 1 999;8 1 : 1 059- 1 06 5 .
computerized morphometry combined with AgNOR stain 2 0 2 . Salama G, Miedouge M, Rouzaud P, e t al. Evaluation of
ing in distinguishing benign from malignant pleural effu pleural CYFRA 2 1 - 1 and carcinoembryonic antigen in the
sions. Anal Quant Cytol Histol. 1 994; 1 6:247-2 52. diagnosis of malignant pleural effusions. Br J Cancer. 1 998;
1 84. Mohanty SK, Dey P, Rana P. Manual and automated 77:472-476.
AgNOR count in differentiating reactive mesothelial from 203. San Jose ME, Alvarez D, Valdes L, et al. Utility of tumour
metastatic malignant cells in serous effusions. Anal Quant markers in the diagnosis of neoplastic pleural effusion. Clin
Cytol Histol. 2003;25 :273-276. Chim Acta. 1 997;265 : 1 93-205.
1 8 5 . Ordonez NG. What are the current best immunohistochem 204. Garcia-Pachon E, Padilla-Navas I, Dosda D, et al. Elevated
ical markers for the diagnosis of epithelioid mesothelioma? level of carcinoembryonic antigen in nonmalignant pleural
A review and update. Hum Pathol. 2007;3 8 : 1 - 1 6. effusions. Chest. 1 997; 1 1 1 :643-647.
1 86 . Barberis MC, Faleri M, Veronese S, et al. Calretinin. A selec 205 . Villena V, Lopez-Encuentra A, Echave-Sustaeta J, et al. Diag
tive marker of normal and neoplastic mesothelial cells in nostic value of CA 72-4, carcinoembryonic antigen, CA 1 5-3,
serous effusions. Acta Cytol. 1 997;4 1 : 1 757- 1 76 1 . and CA 1 9-9 assay in pleural fluid. Cancer. 1 996;78:736-740.
1 87. Granville LA, Younes M , Churg A, e t al. Comparison of 206. Poree! JM, Vives M, Esquerda A, et al. Use of a panel of
monoclonal versus polyclonal calretinin antibodies for tumor markers (carcinoembryonic antigen, cancer antigen
immunohistochemical diagnosis of malignant mesothe 1 25 , carbohydrate antigen 1 5-3, and cytokeratin 19 frag
lioma. Appl lmmunohistochem MolMorphol. 2005; 1 3 :75-79. ments) in pleural fluid for the differential diagnosis of benign
1 8 8 . Ordonez NG. D2-40 and podoplanin are highly specific and malignant effusions. Chest. 2004; 1 26: 1 757- 1 763.
and sensitive immunohistochemical markers of epithelioid 207. Trape J, Molina R, Sant F. Clinical evaluation of the simul
malignant mesothelioma. Hum Pathol. 2005;36:372-3 80. taneous determination of tumor markers in fluid and serum
1 8 9. Attanoos RL, Griffin A, Gibbs AR. The use of immunohis and their ratio in the differential diagnosis of serous effusions.
tochemisrry in distinguishing reactive from neoplastic meso Tumour Biol. 2004;25:276-28 1 .
thelium A novel use for desmin and comparative evaluation 208. Terracciano D , D i Carlo A, Papa P, et al. New approaches
with epithelial membrane antigen, p53, platelet-derived in the diagnostic procedure of malignant pleural effusions.
growth factor-receptor, P-glycoprotein and Bcl-2. Histopa Oneal Rep. 2004; 1 2 :79-83.
thology. 2003;43 :23 1-238. 209. McKenna JM, Chandrasekhar AJ, Henkin RE. Diagnostic
1 90. Longatto Filho A, Alves VA, Kanamura CT, et al. Identifi value of carcinoembryonic antigen in exudative pleural effu
cation of the primary site of metastatic adenocarcinoma in sions. Chest. 1 9 80;78 : 5 87-5 90.
serous effusions. Value of an immunocytochemical panel 2 1 0 . Ringers RA, Loewenstein MS, Feinerman AE, et al. Carci
added to the clinical arsenal. Acta Cytol. 2002;46:65 1 -6 5 8 . noembryonic antigen levels in benign and malignant pleural
1 9 1 . Afify A M , al-Khafaji BM. Diagnostic utility of thyroid tran effusions. Ann Intern Med. l 978;88 :63 1 -634.
scription factor- ! expression in adenocarcinomas presenting 2 1 1 . Tamura S, Nishigaki T, Moriwaki Y, et al. Tumor markers in
in serous fluids. Acta Cytol. 2002;46:675-678. pleural effusion diagnosis. Cancer. 1 9 8 8 ; 6 1 :298-302.
1 92. Ng WK, Chow JC, Ng PK. Thyroid transcription factor- ! 2 1 2 . Villena V, Lopez-Encuentra A, Echave-Sustaeta J, et al. Diag
is highly sensitive and specific in differentiating metastatic nostic value of CA 549 in pleural fluid. Comparison with
pulmonary from extrapulmonary adenocarcinoma in effu CEA, CA 1 5 .3 and CA 72.4. Lung Cancer. 2003;40 :289-294.
sion fluid cytology specimens. Cancer. 2002;96:43-48. 2 1 3 . Shimokata K, Totani Y, Nakanishi K, et al. Diagnostic value
1 93 . Chhieng DC, Ko EC, Yee HT, et al. Malignant pleural effu of cancer antigen 1 5-3 (CA1 5-3) detected by monoclonal
sions due to small-cell lung carcinoma: a cytologic and immu antibodies ( 1 1 5 D8 and DF3) in exudative pleural effusions.
nocytochemical study. Diagn Cytopathol. 200 1 ;2 5 : 3 5 6-360. Eur Respir]. 1 988; 1 :34 1-344.
1 94. Go RS, Klee GG, Richardson RL. Use of pleural fluid pros 2 1 4 . Pinto MM. CA- 1 5 . 3 assay in effusions: comparison with
tate specific antigen in the diagnosis of malignant effusion carcinoembryonic antigen and CA- 1 2 5 assay and cytologic
from metastatic prostate cancer. ] Urol. 2000; 1 64:459. diagnosis. Acta Cytol. 1 996;40:437-442 .
195. Guzman J, Bross KJ, Costabel U. Malignant lymphoma 2 1 5 . Niwa Y, Kishimoto H, Shimokata K. Carcinomatous and
in pleural effusions: an immunocytochemical cell surface tuberculous pleural effusion. Comparison of tumor markers.
analysis. Diagn Cytopathol. 1 9 9 1 ;7: 1 1 3- 1 1 8 . Chest. 1 9 8 5 ;87:35 1 -3 5 5 .
1 96 . Corson J M . Pathology o f diffuse malignant pleural mesothe 2 1 6. Lee YC, Knox BS, Garrett J E . Use o f cytokeratin frag
lioma. Semin Thorac Cardiovasc Surg. 1 997;9:347-3 5 5 . ments 1 9 . 1 and 1 9 . 2 1 (Cyfra 2 1 - 1 ) in the differentiation
1 24 PLE U RAL D I S EASES
of malignant and benign pleural effusions. Aust NZ j Med. A prospective study on I 02 patients as compared with cyro
1 999;29 :765-769. logic examination. Am ] Clin Pathol. 1 99 1 ; 9 5 : 6- 1 2 .
2 1 7. Lee YC, Chern JH, Lai SL, et al. Sialyl stage-specific embry 239. Rodriguez d e Castro F, Molero T, Acosta 0 , e t al. Value o f
onic antigen- I : a useful marker for differentiating the etiology DNA analysis i n addition c o cytological testing i n the diagno
of pleural effusion. Chest. 1 998; 1 1 4: 1 542- 1 545. sis of malignant pleural effusions. Thorax. I 994;49: 692-694 .
2 1 8. Ishikawa H, Sacoh H, Kamma H, et al. Elevated sialyl Lewis 240. Pinto MM. DNA analysis of malignant effusions. Compari
X-i antigen levels in pleural effusions in patients with carci son with cytologic diagnosis and carcinoembryonic antigen
nomacous pleuritis. Intern Med. 1 997;36:68 5-689. content. Anal Quant Cytol Risto!. 1 992; 1 4 :222-226.
2 1 9. Shimokara K, Niwa Y, Yamamoto M , er al. Pleural fluid 24 1 . Moriarty AT, Wiersema L, Snyder W, et al. lmmunophe
neuron-specific enolase. Chest. I 989;95:602-603. notyping of cytologic specimens by flow cytometry. Diagn
220. Hung TL, Chen FF, Liu JM, et al. Clinical evaluation of Cytopathol. I 993;9:252-25 8 .
HER-2/neu protein in malignant pleural effusion-associated 242. Lazcano 0 , Chen LM, Tsai C, e r al. Image analysis and flow
lung adenocarcinoma and as a tumor marker in pleural effu cycometric DNA studies of benign and malignant body cav
sion diagnosis. Clin Cancer Res. 2003;9:260 5-26 1 2 . ity fluids: reappraisal of the role of current methods in the
22 1 . Dikmen G, Dikmen E, Kara M, er al. Diagnostic implica differential diagnosis of reactive versus malignant conditions.
tions of telomerase activity in pleural effusions. Eur Respir]. Mod Pathol. 2000; 1 3 :78 8-796.
2003;22:422-426. 243 . Ikeda K, Tare G, Suzuki T, et al. Diagnostic usefulness of
222. Hess JL, Highsmith WE Jr. Telomerase detection in body EMA, !MP3, and GLUT- I for the immunocytochemical
fluids. Clin Chem. 2002;48 : 1 8-24. distinction of malignant cells from reactive mesothelial cells
223. Light RW Tumor markers in undiagnosed pleural effusions. in effusion cytology using cytospin preparations. Diagn Cyto
Chest. 2004; 1 26: 1 72 1 - 1 722. pathol. 20 1 1 ;39:39 5-40 1 .
224. Creaney J, Yeoman D, Naumoff L, er al. Soluble mesorhe 244. Dewald G , Dines DE, Weiland LH, et al. Usefulness of chro
lin in effusions-a useful cool for the diagnosis of malignant mosome examination in the diagnosis of malignant pleural
mesorhelioma. Thorax. 2007;62 : 5 69-576. effusions. N Engl] Med. 1 976;29 5 : 1 494-1 500.
225. Davies HE, Sadler RS, Bielsa S, et al. The clinical impact and 245 . Korsgaard R. Chromosome analysis of malignant human
reliability of pleural fluid mesothelin in undiagnosed pleural effusions in vivo. Scand j Respir Dis Suppl. 1 979; I 0 5 : 1 - 1 00 .
effusions. Am j Respir Crit Care Med. 2009; 1 80:437-444. 246. Fiegl M . The utility of fluorescence in-situ hybridization i n
226. Creaney J, Robinson BW Serum and pleural fluid biomarkers the diagnosis of malignant pleural effusion. Curr Opin Pulm
for mesorhelioma. Curr Opin Pulm Med. 2009; 1 5 : 366-370. Med. 2005; 1 1 :3 1 3-3 1 8 .
227. Zoppi JA, Pellicer EM, Sundblad AS. Diagnostic value of 247. Fiegl M, Massoner A, Steurer M, er al. Improving tumor
p53 protein in the study of serous effusions. Acta Cytol. cell detection in pleural effusions by interphase cytogenetics.
1 995;39:72 1 -724. Cytometry B Clin Cytom. 2003 ; 5 5 : 60-62.
228. Mayall F, Heryer A, Manga D, er al. p53 immunostaining is 248. Savic S, Franco N, Grilli B , er al. Fluorescence in situ hybrid
a highly specific and moderately sensitive marker of malig ization in the definitive diagnosis of malignant mesothelioma
nancy in serous fluid cytology. Cytopathology. 1 997;8:9- 1 2 . in effusion cytology. Chest. 20 1 0; 1 3 8 : 1 37- 1 44.
229. Tawfik MS, Coleman D V C-myc expression in exfoliated 249. Tyan YC, Wu HY, Lai WW, et al. Proteomic profiling of
cells in serous effusions. Cytopathology. 1 9 9 1 ;2: 83-92. human pleural effusion using two-dimensional nano liquid
230. Athanassiadou PP, Veneri SZ, Kyrkou KA, et al. Detection chromatography tandem mass spectrometry. j Proteome Res.
of c-Ha-ras oncogene expression in pleural and peritoneal 2005;4: 1 274- 1 286.
smear effusions by in situ hybridization. Cancer Detect Prev. 250. Tyan YC, Wu HY, Su WC, et al. Proteomic analysis of
1 993; 1 7 : 5 8 5-590. human pleural effusion. Proteomics. 2005;4:748-757.
23 1 . Rasmussen KN, Faber V Hyaluronic acid in 247 pleural flu 25 1 . Bard MP, Hegmans JP, Hemmes A, et al. Proteomic analysis
ids. Scandj Respir Dis. 1 967;48: 366-37 1 . of exosomes isolated from human malignant pleural effu
232. Nurminen M , Dejmek A, Marrensson G , er al. Clinical utility sions. Am ] Respir Cell Mo! Biol. 2004;3 1 : 1 14-1 2 1 .
of liquid-chromatographic analysis of effusions for hyaluro 252. Piras MA, Gakis C , Budroni M , e t al. Adenosine deaminase
nate content. Clin Chem. 1 994;40:777-780. activity in pleural effusions: an aid co differential diagnosis.
233. Hillerdal G, Lindqvisr U, Engstrom-Laurent A. Hyaluronan in Br Medj. 1 978;4: 1 7 5 1 - 1 752.
pleural effusions and in serum. Cancer. 1 9 9 1 ;67:24 1 0-24 1 4 . 253. Ocana IM, Martinez-Vazquez JM, Seguna RM, et al. Ade
234. Pettersson T, Froserh B, Riska H, et a l . Concentration o f nosine deaminase in pleural fluids. Chest. 1 983;84: 5 1 -53 .
hyaluronic acid i n pleural fluid a s a diagnostic aid for malig 254. Valdes L , San Jose E , Alvarez D, e t al. Diagnosis o f tuberculous
nant mesothelioma. Chest. 1 988;94: 1 037- 1 039. pleurisy using the biologic parameters adenosine deaminase,
235. Marrensson G, Thylen A, Lindquist U, et al. The sensitiv lysozyme, and interferon gamma. Chest. 1 993; 1 03:458-465.
ity of hyaluronan analysis of pleural fluid from patients 255. Liang QL, Shi HZ, Wang K, et al. Diagnostic accuracy of
with malignant mesothelioma and a comparison of different adenosine deaminase in tuberculous pleurisy: A meta-analysis.
methods. Cancer. 1 994;73 : 1 406- 1 4 1 0 . Respir Med. 200 8 ; 1 02:744-754.
236. Kawai T, Greenberg S D , Truong L O , e t al. Differences i n lec 256. Porcel JM, Esquerda A, Bielsa S . Diagnostic performance of
tin binding of malignant pleural mesorhelioma and adeno adenosine deaminase activity in pleural fluid: a single-center
carcinoma of the lung. Am ] Pathol. 1 9 88; 1 30:40 1 -4 1 0 . experience with over 2, 1 00 consecutive patients. Eurj Intern
237. Croonen AM , van der Valk P, Herman CJ, e t al. Cytology, Med. 20 1 0;2 1 :4 1 9-423 .
immunopathology and flow cycometry in the diagnosis of 25 7. Baba K , Hoosen AA , Langeland N, e t al. Adenosine deaminase
pleural and peritoneal effusion. Lab Invest. 1 98 8 ; 5 8 :725-732. activity is a sensitive marker for the diagnosis of tuberculous
238. Rijken A, Dekker A, Taylor S, et al. Diagnostic value of DNA pleuritis in patients with very low CD4 counts. PLoS One.
analysis in effusions by flow cyrometry and image analysis. 2008;3:e278 8 .
C H APT E R 7 I CLI N I CAL M A N I F E STAT I O N S A N D U S E F U L TESTS 1 25
2 5 8 . Chung JH, Kim YS , Kim SI, et al. The diagnostic value of 279 . Greco S , Girardi E , Masciangelo R , e t al. Adenosine deami
the adenosine deaminase activity in the pleural fluid of renal nase and interferon gamma measurements for the diagnosis of
transplant patients with tuberculous pleural effusion. Yonsei tuberculous pleurisy: a meta-analysis. Int J Tuberc Lung Dis.
Med]. 2004;45 :66 1 -664. 2003;7:777-786.
259. Ocana ], Ribera E, Martinez-Vazquez JM, et al. Adenosine 280. Chegou NN, Walz! G, Bolliger CT, et al. Evaluation ofadapted
deaminase activity in rheumatoid pleural effusion. Ann Rheum whole-blood interferon-gamma release assays for the diagnosis
Dis. 1 988 ;47:394-397. of pleural tuberculosis. Respiration. 2008;76: 1 3 1 - 1 3 8 .
260. Burgess LJ, Maritz FJ, Le Roux I, et al. Combined use of 28 1 . Zhou Q, Chen YQ, Qin S M , e t al. Diagnostic accuracy of
pleural adenosine deaminase with lymphocyre/neutrophil T-cell interferon-gamma release assays in ruberculous pleu
ratio. Increased specificity for the diagnosis of tuberculous risy: a meta-analysis. Respirology. 20 1 1 ; 1 6:473-480.
pleuritis. Chest. 1 996; 1 09 :4 1 4-4 1 9 . 282. Trajman A, Pai M, Dheda K, et al. Novel tests for diagnosing
26 1 . Ungerer J P, Grabler S M . Molecular forms o f adenosine ruberculous pleural effusion: what works and what does not?
deaminase in pleural effusions. Enzyme. 1 98 8;40:7- 1 3 . Eur Respir]. 2008;3 1 : 1 098-1 1 06.
262. Esteban C, Orib e M, Fernandez A , et al. Increased adenosine 283. Pai M, Flores LL, Hubbard A, et al. Nucleic acid amplifica
deaminase activity in Q fever pneumonia with pleural effu tion tests in the diagnosis of tuberclous pleuritis: a systematic
sion. Chest. 1 994; 1 0 5-648. review and meta-analysis. BMC Infect Dis. 2004;4:6.
263 . Dikensoy 0, Namiduru M, Hocaoglu S, et al. Increased pleu 284. Santos A, Cremades R, Rodriguez JC, et al. Comparison of
ral fluid adenosine deaminase in brucellosis is difficult to dif methods of DNA extraction for real-time PCR in a model of
ferentiate from tuberculosis. Respiration. 2002;69:5 5 6-5 5 9 . pleural tuberculosis. APMIS. 20 1 0; 1 1 8 : 60-6 5 .
2 6 4 . Dikensoy 0 , Fakili F, Elbek 0, e t al. High adenosine deami 2 8 5 . Yum H K , Choi S J . Detection o f mycobacterial DNA using
nase activity in the pleural effusion of a patient with Legion nested polymerase chain reaction of pleural biopsy speci
naires' disease. Respirology. 2008 ; 1 3 :473-474. mens: compared to pathologic fi n dings. Korean J Intern Med.
265. Lee YC, Rogers JT, Rodriguez RM, et al. Adenosine deami 2003; 1 8 : 89-93.
nase levels in nontuberculous lymphocytic pleural effusions. 286. Garcia-Pachon E, Soler MJ, Padilla-Navas I, et al. C-reactive
Chest. 200 1 ; 1 20:356-36 1 . protein in lymphocytic pleural effusions: a diagnostic aid in
266. Jimenez Castro D , Diaz Nuevo G , Perez-Rodriguez E, et al. tuberculous pleuritis. Respiration. 2005 ;72:486-489.
Diagnostic value of adenosine deaminase in nontuberculous 287. Chierakul N, Kanitsap A, Chaiprasert A, et al. A simple
lymphocyric pleural effusions. Eur Respir]. 2003;21 :220-224. C-reactive protein measurement for the differentiation
267. Shibagaki T, Hasegawa Y, Saito H, et al. Adenosine deami berween tuberculous and malignant pleural effusion. Respi
nase isozymes in tuberculous pleural effusion. J Lab Clin Med. rology. 2004;9:66-69.
1 996; 1 27:348-352. 288. Daniil ZD, Zintzaras E, Kiropoulos T, et al. Discrimination
268. Perez-Rodriguez E, Castro DJ . The use of ADA and ADA of exudative pleural effusions based on multiple biological
isoenzymes in the diagnosis of ruberculous pleuritis. Curr parameters. Eur Respir]. 2007;30:957-964.
Opin Pulm Med. 2000;6:259-266. 289. Asseo PP, Tracopoulos GD, Kotsovoulou-Fouskak V Lyso
269. Perez-Rodriguez E, Walton !J, Hernandez JJ, et al. ADA ! / zyme (muramidase) in pleural effusions and serum. Am J
ADAp ratio i n pleural tuberculosis: a n excellent diagnostic Clin Pathol. 1 9 82;78 :763-767.
parameter in pleural fl u id. Respir Med. 1 999;93 : 8 1 6-82 1 . 290. Verea Hernando HR, Masa Jimenez JF, Dominguez Juncal L,
270. Valdes L, Alvarez D , San Jose E, et al. Tuberculous pleurisy: a et al. Meaning and diagnostic value of determining the lyso
study of254 patients. Arch Intern Med 1 998; 1 5 8:20 1 7-202 1 . zyme level of pleural fluid. Chest. 1 9 87; 9 1 :342-345 .
27 1 . Miller KD, Barnette R, Light RW. Stability ofadenosine deam 29 1 . Wang CY, Hsiao YC, Jerng J S , e t al. Diagnostic value of
inase during transportation. Chest. 2004; 1 26: 1 933- 1 937. procalcitonin in pleural effusions. EurJ Clin Microbiol Infect
272. Antonangelo L, Vargas FS, Almeida LP, et al. Influence of Dis. 20 1 1 ;30:3 1 3-3 1 8 .
storage time and temperature on pleural fluid adenosine 292. Baig MME, Pettengell KE, Simgee AE, et al. Diagnosis of
deaminase determination. Respirology. 2006; 1 1 :48 8-492. ruberculosis by detection of mycobacterial antigens in pleu
273 . Aoki Y, Katoh 0 , Nakanishi Y, et al. A comparison study of ral effusions and ascites. S Afr Med]. 1 986;69: 1 0 1 - 1 02.
IFN-gamma, ADA, and CA 1 2 5 as the diagnostic parameters 293. Yew WW, Chan CY, Kwan SY, et al. Diagnosis of tubercu
in tuberculous pleuritis. Respir Med. 1 994;8 8 : 1 39- 1 43 . lous pleural effusion by the detection of ruberculostearic acid
274. Ribera E , Ocana I, Martinez-Vazquez J M , e t al. High level in pleural aspirates. Chest. 1 99 1 ; 1 00 : 1 26 1 - 1 263.
of interferon gamma in tuberculous pleural effusion. Chest. 294. Banchuin N, Pumprueg U, Pimolpan V, et al. Anti-PPD
1 9 88;93:308-3 1 1 . IgG responses in ruberculous pleurisy. J Med Assoc Thai.
275 . Barnes PF, Mistry SD, Cooper CL, et al. Compartmentaliza 1 987;70:32 1 -325.
tion of a CD4• T lymphocyte subpopulation in tuberculous 295. Dhand R, Gangul NK, Vaishnavi C, et al. False-positive
pleuritis. j Immunol. 1 989; 1 42: 1 1 1 4- 1 1 1 9 . reactions with enzyme-linked immunosorbent assay of Myco
276. Villena V, Lopez-Encuentra A , Echave-Sustaeta J, e t al. bacterium tuberculosis antigens in pleural fluid. ] Med Micro
Interferon-gamma in 388 immunocompromised and immu biol. 1 98 8 ;26:24 1 -243 .
nocompetent patients for diagnosing pleural tuberculosis. 296. Levy H, Wayne LG, Anderson BE, et al. Anti-mycobacterial
Eur Respir]. 1 996;9:2635-2639. antibody levels in pleural fluid reflect passive diffusion from
277. Villena V, Lopez-Encuentra A, Pozo F, et al. Interferon serum. Chest. 1 990;97: 1 1 44- 1 1 47.
gamma levels in pleural fluid for the diagnosis of tuberculo 297. Caminero JA, Rodriguez de Castro F, Carrillo T, et al. Diagnosis
sis. Am ] Med. 2003; 1 1 5 :365-370. of pleural tuberculosis by detection of specific IgG anti-antigen
278. Jiang J, Shi HZ, Liang QL, et al. Diagnostic value of 60 in serum and pleural fluid. Respiration. 1 993;60 : 5 8-62.
interferon-gamma in tuberculous pleurisy: a meta-analysis. 298. Van Vooren JP, Farber CM, De Bruyn J, et al. Antimycobac
Chest. 2007; 1 3 1 : 1 1 33-1 1 4 1 . terial antibodies in pleural effusions. Chest. 1 990;97:88-90.
1 26 PLE U RAL D I S EASES
299. Murate T, Mizoguchi K, Amano H, et al. Antipurified 320. Hughes RL, Mintzer RA, Hidvegi DF, et al. The manage
protein-derivative antibody in tuberculous pleural effusions. ment of chylothorax. Chest. 1 979;76:2 1 2-2 1 8 .
Chest. 1 990;97:670-673. 32 1 . Wolthuis A, Landewe RB, Theunissen PH, et al. Chylo
300. Yokoyama T, Rikimaru T, Kinoshita T, et al. Clinical urility of thorax or leakage of total parenteral nutrition? Eur Respir ].
lipoarabinomannan antibody in pleural fluid for the diagnosis 1 998; 1 2 : 1 233-1 235.
of tuberculous pleurisy. J Infect Chemother. 2005; 1 1 : 8 1-83. 322. Staats BA, Ellefson RD, Budahn LL, et al. The lipoprotein
30 1 . Kunter E, Cerrahoglu K, Ilvan A, et al. The value of pleural profile of chylous and nonchylous pleural effusions. Mayo
fluid anti-A60 IgM in BCG-vaccinated tuberculous pleurisy Clin Proc. 1 98 0 ; 5 5 : 700-704.
patients. Clin Microbial Infect. 2003;9:2 1 2-220. 323. Seriff NS, Cohen ML, Samuel P, et al. Chylothorax: diagno
302. Walker WC, Wright V Rheumaroid pleuritis. Ann Rheum sis by lipoprotein electrophoresis of serum and pleural fluid.
Dis. 1 967;26:467--474. Thorax. 1 977;32:98- 1 00.
303. Berger HW, Seckler SG. Pleural and pericardia! effusions in 324. Valdes L, San Jose ME, Pose A, et al. Usefulness of triglycer
rheumatoid disease. Ann Intern Med. 1 966;64: 1 2 9 1 - 1 297. ide levels in pleural fluid. Lung. 20 1 0; 1 8 8 :483-489.
304. Levine H, Szanto M, Brieble HG, et al. Rheumatoid fac 325. Romero S, Martin C, Hernandez L, et al. Chylothorax in
tor in non-rheumatoid pleural effusions. Ann Intern Med. cirrhosis of the liver: analysis of its frequency and clinical
1 968;69 :487--492. characteristics. Chest. 1 998 ; 1 1 4 : 1 54- 1 59.
305. Khare V, Baethge B, Lang S, et al. Antinuclear antibodies in 326. Coe JE, Aikawa JK. Cholesterol pleural effusion. Arch Intern
pleural fluid. Chest. 1 994; 1 06:866-87 1 . Med. 1 96 1 ; 1 0 8 :763-774.
306. Wang DY, Yang PC, Yu WL, et al. Serial antinuclear anti 327. Roy PH, Carr DT, Payne WS . The problem of chylothorax.
bodies titre in pleural and pericardia! fluid. Eur Respir ]. Mayo Clin Proc. 1 967;42:457--467.
2000; 1 5 : 1 1 06- 1 1 1 0. 328. Xiol X, Castellvi JM, Guardiola ], et al. Spontaneous bacterial
307. Leechawengwong M, Berger HW, Sukumaran M . Diag empyema in cirrhotic patients: a prospective study. Hepatol
nostic significance of antinuclear antibodies in pleural ogy. 1 996;23 : 7 1 9-723 .
effusion. Mt Sinai] Med. 1 979;46 : 1 37- 1 39 . 329. Menzies SM, Rahman NM, Wrightson JM, et al. Blood cul
308. Wang DY, Yang P C , Yu WL, e t a l . Comparison of differ ture bottle culture of pleural fluid in pleural infection. Thorax.
ent diagnostic methods for lupus pleuritis and pericarditis: 20 1 1 ;66:65 8-662.
a prospective three-year study. J Formos Med Assoc. 2000; 330. Everts RJ, Reller LB. Pleural space infections: microbiol
99:375-380. ogy and antimicrobial therapy. Semin Respir Infect. 1 999;
309. Chao TY, Huang SH, Chu CC. Lupus erythematosus cells 1 4: 1 8-30.
in pleural effusions: diagnostic of systemic lupus eryrhema 33 1 . Maartens G, Bateman ED. Tuberculous pleural effusions:
tosus? Acta Cytol. 1 997;4 1 : 1 23 1- 1 233. increased culture yield with bedside inoculation of pleural
3 1 0 . Nosanchuk JS, Naylor B . A unique cytologic picture in pleu fluid and poor diagnostic value of adenosine deaminase.
ral fluid from patients with rheumatoid arthritis. Am J Clin Thorax. 1 9 9 1 ;46:96-99.
Pathol. 1 968;50 :330-33 5 . 332. Barnes Tw, Olson EJ, Morgenthaler TI, et al. Low yield
3 1 1 . Chou C-W, Chang S-C. Pleuritis a s a presenting manifesta of microbiologic studies on pleural fluid specimens. Chest.
tion of rheumatoid arthritis: diagnostic clues in pleural fluid 2005 ; 1 27:9 1 6-92 1 .
cytology. Am ] Med Sci. 2002;323: 1 5 8-1 6 1 . 333. Ferrer A, Osset ] , Alegre ] , et al. Prospective clinical and
3 1 2. Hunder GG, McDuffie FC, Hepper NGG. Pleural fluid microbiological study of pleural effusions. EurJ Clin Micro
complement in systemic lupus eryrhematosus and rheuma biol lnfect Dis. 1 999; 1 8 :237-24 1 .
toid arthritis. Ann Intern Med. 1 972;76:357-362. 334. Maskell NA, Batt S , Hedley EL, et al. Th e bacteriology of
3 1 3 . Glovsky MM, Louie JS, Pitts WH Jr, et al. Reduction of pleural infection by genetic and standard methods and its
pleural fluid complement activity in patients with systemic mortality significance. Am ] Respir Crit Care Med. 2006;
lupus erythematosus and rheumatoid arthritis. Clin Immunol 1 74: 8 1 7-823.
Immunopathol. 1 976;6:3 l --4 l . 335. Maskell NA, Davies CW, Nunn AJ, et al. U.K. controlled
3 1 4. Poree! JM, Vives M , Gazquez I , et al. Usefulness of pleural trial of intrapleural streptokinase for pleural infection N Engl
complement activation products in differentiating tuber j Med. 2005;352: 865-874.
culosis and malignant effusions. Int J Tuberc Lung Dis. 336. Lampe RM, Chottipitayasunondh T, Sunakorn P. Detection
2000;4:76-82. of bacterial antigen in pleural fluid by counterimmunoelec
3 1 5 . Salomaa ER, Viander M, Saaresranta T, et al. Complement trophoresis. ] Pediatr. 1 976; 8 8 : 5 57-560.
components and their activation products in pleural fluid. 337. Coonrod JD, Wilson H D . Etiologic diagnosis ofintrapleural
Chest. 1 998; 1 1 4:723-730. empyema by counterimmunoelectrophoresis. Am Rev Respir
3 1 6. Vives M, Poree! JM, Gazquez I, et al. Pleural SC5b-9: a test Dis. 1 976; 1 1 3 :637-64 1 .
for identifying complicated parapneumonic effusions. Respi 338. Thadephalli H , Gangopadhyay PK. Rapid diagnosis of
ration. 2000;67:433--438 . anaerobic empyema by direct gas-liquid chromatography of
3 1 7. Andrews BS, Arora N S , Shadforth MF, e t al. Th e role of pleural fluid. Chest. 1 9 80;77:507-5 1 3 .
immune complexes in the pathogenesis of pleural effusions. 339. Luetscher J A Jr. Electrophoretic analysis o f the proteins of
Am Rev Respir Dis. 1 98 1 ; 1 24: 1 1 5- 1 20 . plasma and serous effusions. ] Clin Invest. 1 9 4 1 ;20:99- 1 06.
3 1 8 . Hunder GG, McDuffie FC, Huston KA , e t a l . Pleural fluid 340. Philip-Joet F, Alessi MC, Philip-Joet C, et al. Fibrinolyric
complement, complement conversion, and immune com and inflammatory processes in pleural effusions. Eur Respir].
plexes in immunologic and nonimmunologic diseases. J Lab 1 99 5 ; 8 : 1 3 52- 1 356.
Clin Med. 1 977;90 : 9 7 1 -980. 34 1 . Demirkazik A, Dinco! D, Hasturk S , et al. Diagnostic value
3 1 9. Bruneau R, Rubin P. The management of pleural effusions and of ferritin in the differential diagnosis of malignant effusions.
chylothorax in lymphoma. Radiology. 1 965; 8 5 : 1 085-1 092. Cancer Biochem Biophys. 1 998; 1 6:243-25 1 .
C H APT E R 7 I CLI N I CAL M A N I F E STAT I O N S A N D U S E F U L TESTS 1 27
342. Delpuech P, Desch G, Frucrus F. Fibroneccin is unsuitable 347. Bedrossian CWM, Stein DA, Miller WC, e t al. Levels o f
as a tumor marker in pleural effusions. Clin Chem. 1 9 89;35: angiotensin-converting enzyme i n pleural effusion. Arch
1 66- 1 68 . Pathol Lab Med. 1 98 1 ; 1 0 5 :345-346.
343. Arai H, Endo M, Yokosawa A , e t al. On acid glycosamino 348. Veran P, Moigneteau C, Lasausse G, et al. Les phospha
glycans (mucopolysaccharides) in pleural effusion. Am Rev tases desepanchements pleuraux de diverses natures. Leur
Respir Dis. 1 975 ; 1 1 1 :37-42. inten�c dans le cancer de la prostate. J Fr Med Chir Thorac.
344. Vladutiu AO, Brason FW, Adler RH. Differential diagnosis 1 965; 1 9:62 1 -643.
of pleural effusions: clinical usefulness of cell marker quanti 349. Migueres J, Jovger A, Abour P. Valuer theorique et pratique
tation. Chest. 1 9 8 1 ;79:297-30 1 . de certains dosages enzymatiques au cours des epanche
345. Brauer MJ, West M , Zimmerman HJ. Comparison of gly ments pleuraux (amylase, phosphatases, lacticodeshydroge
colytic and oxidative enzyme and transaminase values in nase) . A propos de 129 observations. J Fr Med Chir Thorac.
benign and malignant effusions with chose in serums. Cancer. l 969;23:443-4 5 8 .
1 963; 1 6 : 5 33-54 1 .
346. Feldstein AM, Samachson J , Spencer H . Levels of calcium,
phosphorus, alkaline phosphatase and protein in effusion
fluid and serum in man. Am ] Med. 1 963;35: 530-5 3 5 .
Approach to the Patient
Whenever a patient with an abnormal chest radio effusion that occurs after coronary artery bypass graft
graph is evaluated, the possibility of a pleural effusion (CABG) surgery.
should be considered. Increased densities on the chest
radiograph are frequently attributed to parenchymal S E PA RAT I O N O F E X U DAT E S F R O M
infiltrates when they actually represent pleural fluid. TRAN S U DAT E S
Most patients with pleural effusions have blunting of
the posterior costophrenic sulcus on the lateral chest I f free pleural fluid i s demonstrated o n the decubi
radiograph. If this angle is blunted, the patient should tus film, with ultrasound or with a CT scan, one
be evaluated with ultrasound, CT scan, or bilateral should consider performing a diagnostic thoracente
decubitus chest radiographs to ascertain whether free sis (Fig. 8 . 1 ) . It has been my experience that diag
pleural fluid is present (see Chapter 6) . This chap nostic thoracentesis is difficult if the thickness of the
ter provides a guide to the approach to a patient fluid on the decubitus radiograph, ultrasound, or
with an undiagnosed pleural effusion. The British the CT scan is less than 1 0 mm. If the thickness of
Thoracic Society has recently published guidelines the fluid is greater than 1 0 mm, however, consider
for the investigation of a unilateral pleural effusion in ation should be given to performing a diagnostic tho
adults ( 1 ) . The management of patients with pleural racentesis (see Chapter 28) . If the patient has obvious
effusions due to specific diseases is discussed in the congestive heart failure, I perform a diagnostic tho
chapters dealing with those diseases. racentesis if any of the following three conditions are
met: (a) the effusions are not bilateral and compa
rably sized, (b) the patient has pleuritic chest pain,
F R E Q U E N C I E S OF VA R I O U S
or (c) the patient is febrile. Otherwise, treatment of
D I AG N O S E S
the congestive heart failure is initiated. If the pleu
Pleural effusions can occur as complications of many ral effusions do not rapidly disappear, I then perform
different diseases (Table 8 . 1 ) . The vigor with which var a diagnostic thoracentesis several days later. It must
ious diagnoses are pursued depends on the likelihood be remembered, however, that the characteristics of
that the individual has that particular disease. Table 8.2 the pleural fluid may change from those of a transu
shows the approximate annual incidence for the most date to those of an exudate with diuresis. Romero et al.
common causes of pleural effusions. An epidemiologic (3) performed a thoracentesis on 1 5 patients with
study from the Czech Republic found that the four congestive heart failure before and every 48 hours
leading causes of pleural effusions in order of incidence after diuretic therapy was initiated. Before diuretics
were congestive heart failure, malignancy, pneumonia, were administered, only one effusion was misclassi
and pulmonary embolism (2) . Congestive heart fail fied as an exudate by Light's criteria, but at the time of
ure and cirrhosis cause almost all transudative pleural the third thoracentesis 10 effusions were misclassified
effusions, whereas malignant disease, pneumonia, and as exudates. Between the first and the third thoracen
pulmonary embolization are the three main causes of tesis, the mean protein level increased from 2.3 to
exudative pleural effusions. Two other frequent causes 3 . 5 g/dL and the mean lactate dehydrogenase (LDH)
of exudative pleural effusions are viral infections and the level increased from 1 76 to 262 IU/L (3) .
1 28
C HAPT E R 8 I APPROACH TO T H E PAT I E N T 1 29
One of the main purposes of the diagnostic tho surfaces can be ignored while the congestive heart
racentesis is to determine whether the patient has a failure, cirrhosis, or nephrosis, for example, is treated.
transudative or an exudative pleural effusion. This Alternately, if any of the three criteria in Figure 8 . 1
distinction is made by analysis of the levels of protein is met, the patient probably has an exudative pleu
and LDH in the pleural fluid and in the serum (4) . If ral effusion. The exudative nature indicates that the
none of the criteria in Figure 8 . 1 is met, the patient has pleural effusion resulted from local disease where the
a transudative pleural effusion. Therefore, the pleural fluid originated, and further investigation should
1 30 PLE U RAL D I S EASES
D I F F E R E N T I AT I N G A M O N G VA R I O U S
E X U DAT I V E P L E U RA L E F F U S I O N S
@
Blunting of costrophrenic angle? >------.. P l e u ral effusion u n l i kely
Observe
P robable exudate
Patient has C H F
o r ci rrhosis?
Treat C H F ,
ci rrhosis, or
n e p h rosis
Transudate
Exudate Serum-p l e u ral
fluid p rotei n
g radient >3 . 1
Appearance of p l e u ral f l u i d
G l ucose of p l e u ral f l u i d
NT-pro B N P
Cytology and d iffe rential c e l l Treat C H F
> 1 500
c o u n t of pleural f l u i d
P l e u ral f l u i d marker for TB
Appearance of p l e u ral f l u i d
Bl oody ?
Go to chem ical
Li kely diagnosis Blood i n ess not analysis
tumor, p u l monary sign ificant
emblous, or trauma
C hylothorax or
pseudochylothorax
Pseudochylothorax
Lipoprotei n analysis
Chylomicrons 1----.i
usually occur when the pleural effusion has been The patient with an exudative pleural effusion
present for many years. Cholesterol crystals may be with predominantly polymorphonuclear leukocytes
found in the sediment, and high levels of triglyc and without parenchymal infiltrates most likely has
erides are not usually present in the pleural fluid. pulmonary embolus, viral infection, gastrointestinal
In contrast, chylothoraces are more acute, do not disease, asbestos pleural effusion, malignant pleural
contain cholesterol crystals, and are characterized disease, or acute tuberculous pleuritis. Accordingly,
by high levels of triglycerides. The management of a the patient should undergo a CT angiogram scan or
patient with a chylothorax or a pseudochylothorax a lung scan for evaluation of pulmonary embolus.
is discussed in Chapter 26. A gastrointestinal etiology of the pleural effusion
can be evaluated with an abdominal CT scan or
ultrasound. A careful history should be taken for
Routine Measurements on Exudative
asbestos exposure. The marker for tuberculosis (ade
Pleural Fluids
nosine deaminase [ADA] or interferon-gamma) will
When a patient has an undiagnosed exudative effusion, indicate whether the patient has tuberculosis, and
there are several tests that should be routinely obtained, the cytology will provide the first evaluation for
namely, a pleural fluid cell count and differential, a pleural malignancy.
pleural fluid glucose and LDH level, examination of The patient with an exudative pleural effusion
the pleural fluid for malignant cells, and a pleural fluid with predominantly mononuclear cells in the pleu
marker for tuberculosis (1 O). A good starting point for ral fluid has a chronic process involving the pleural
the diagnostic assessment of an unknown exudate is space. Malignant disease, pulmonary embolization,
the pleural fluid cytology. An algorithm starting with pleural effusions following CABG, and tuberculosis
the pleural cytology is presented in Figure 8.3. are the four most common causes of this picture.
In previous editions of this book, I recommended Again, the cytology on the pleural fluid is the first
that a pleural fluid amylase also be obtained. How step in evaluating the possibility of malignancy. The
ever, the pleural fluid amylase only helps occasion history will reveal whether the patient underwent
ally in making a diagnosis and, therefore, it should CABG surgery the previous year. An elevated level of
not be obtained on a routine basis ( 1 1 ) . It should be ADA or interferon-gamma in the pleural fluid essen
obtained if acute pancreatitis, esophageal rupture, or tially establishes the diagnosis of tuberculous pleuritis
chronic pancreatic pleural effusion is suspected ( 1 1 ) . (see Chapter 1 3) . If none of the tests mentioned in
the preceding text is positive, the possibility of pul
Pleural Fluid Differential Cell Count
monary embolus should be assessed with a CT angio
In patients with exudative pleural effusions, the cell gram or a lung scan.
count and the differential provide clues about the
Pleural Fluid Glucose
etiology of the pleural effusion. When neutrophils
predominate in the pleural fluid, an acute process The routine measurement of the pleural fluid glu
is affecting the pleural surfaces, and the chest radio cose level is recommended because the presence of a
graph should be evaluated for parenchymal infil reduced (<60 mg/dL) pleural fluid glucose reduces
trates. The presence of an infiltrate indicates that the spectrum of diagnostic possibilities. Most patients
the patient probably has a parapneumonic effusion with a reduced pleural fluid glucose level ( <60 mg/dL)
although pulmonary embolus and bronchogenic car have one of four conditions: parapneumonic effusion,
cinoma should also be considered. The diagnosis of a malignant pleural effusion, tuberculous pleuritis, or
parapneumonic effusion is likely if purulent sputum rheumatoid pleural effusion ( 1 2) . Other rare causes of
is present. When purulent sputum or peripheral leu a low glucose pleural effusion include paragonimiasis,
kocytosis is not seen, the patient should undergo a hemothorax, Churg-Strauss syndrome, urinothorax,
CT angiogram scan to rule out pulmonary embolus. and occasionally lupus pleuritis. Most patients with a
In the event of a negative scan result, a bronchoscopy reduced pleural fluid glucose level also have a reduced
with transbronchial biopsy should be performed to pleural fluid pH and an increased pleural fluid LDH
determine the cause of the parenchymal infiltrate. If level. Laboratory errors in the performance of one of
after all these studies the diagnosis is still not clear, these three tests should be suspected when these rela
video-assisted thoracoscopy (see Chapter 30) should tionships are not maintained.
be performed if the infiltrate is worsening or the effu Patients with either parapneumonic effusions
sion is increasing in size. or tuberculous pleuritis may have an acute illness
1 34 PLE U RAL D I S EASES
P l e u ral f l u i d cytology
TB
ADA or
i nte rfe ron-y
Abdom inal CT
P u ru l ent sputum Negative CT angiogram Negative
scan or
u ltrasound
Patient
Positive
i m p rovi ng
Parapneumonic P u l monary
CT angiogram GI d i sease
effusion embolism
Observe
Consider thoracoscopy
Consider u n usual cases of p l e u ral effusion
Drug reactions
D ress l e r's syndrome
Asbestos p l e u ral effusion
Fungus and parasitic d isease
Collagen vascular d i sease
characterized by fever, cough, and pleuritic chest most patients with parapneumonic effusions have
pain and a low pleural fluid glucose level. Patients predominantly neutrophils in their pleural fluid,
with parapneumonic effusions usually have radiologi whereas most patients with tuberculous pleuritis have
cally evident parenchymal infiltrates, whereas most predominantly lymphocytes.
of those with tuberculous pleuritis have no infil Patients with subacute or chronic symptoms and
trates. The differential cell count on the pleural fluid a low pleural fluid glucose level may have malignant
is also useful in making the differentiation because pleural disease, rheumatoid disease, tuberculosis, or
C HAPT E R 8 I APPROACH TO T H E PAT I E N T 135
even a chronic bacterial infection. The diagnosis of Pleural Fluid Adenosine Deaminase Level
rheumatoid pleuritis (see Chapter 2 1 ) is usually easy.
The diagnosis of tuberculosis is virtually established
The differentiation among tuberculosis, malignant
if the pleural fluid ADA level is more than 40 U/L
disease, and chronic bacterial infection may be more
and the patient has predominantly lymphocytes in
difficult. The pleural fluid cytology is usually posi
the pleural fluid (see Chapter 7) . The higher the pleu
tive for malignant cells in patients with a malignant
ral fluid ADA level, the more likely the patient is to
pleural effusion and a low pleural fluid glucose level.
have tuberculous pleuritis. In one study, the ADA
The pleural fluid marker for tuberculosis should be
was more than 47 U/L in 253 of 254 patients with
positive with tuberculous pleuritis, and neutrophils
tuberculous pleuritis ( 1 3) . In a second report, only
should predominate in the pleural fluid if a chronic
5 of 1 73 patients (3%) with pleural effusions due to
bacterial infection is present.
other etiologies, including 46 with malignancy and
Pleural Fluid lactate Dehydrogenase 30 with pneumonia, had ADA levels that exceeded
45 U/L ( 1 4) . The two other disease entities that
Although the level of LDH in the pleural fluid is not tend to have a high pleural fluid ADA level are empy
particularly useful in the differentiation of the vari ema and rheumatoid pleuritis ( 1 4), and both these
ous exudative pleural effusions, a pleural fluid LDH conditions are easily distinguished from pleural
should be obtained every time a thoracentesis is tuberculosis by the clinical picture.
performed on a patient with an undiagnosed pleural
effusion. The pleural fluid LDH is a reliable indica Pleural Fluid Interferon-Gamma Levels
tor of the degree of pleural inflammation. If, with
repeated thoracentesis, the pleural fluid LDH level Pleural fluid interferon-gamma levels are also elevated
increases, the degree of inflammation in the pleural with tuberculous pleuritis (see Chapter 7) . Pleural fluid
space is becoming progressively worse and one should interferon-gamma levels are very efficient at differenti
be aggressive in pursuing a diagnosis. Alternatively, if ating tuberculous from nontuberculous pleural effu
the pleural fluid LDH level decreases with repeated sion. Using a cutoff level of 3.7 U/mL, Villena et al.
thoracentesis, the degree of inflammation in the pleu ( 1 5) demonstrated that this test resulted in a sensitivity
ral space is becoming progressively less and one need of 0.98 and a specificity of 0.98 in a series of 595
not be as aggressive in the approach to the patient. pleural effusions including 82 tuberculous effusions.
These results were better than the same group reported
Pleural Fluid Cytology with ADA levels ( 1 6) .
If a patient has malignancy, cytologic examination
of the pleural fluid is a fast, efficient, and minimally O PT I O N S W H E N N O D I A G N O S I S
invasive means by which to establish the diagnosis. I S O BTA I N E D A FT E R I N I T I A L
The percentage of malignant pleural effusions that T H O RAC E N T E S I S
are diagnosed with cytology has been reported to be
anywhere between 40% and 87%. There are several When n o diagnosis has been obtained after an initial
factors that influence the diagnostic yield with cytol thoracentesis that includes a pleural fluid marker for
ogy. Almost all adenocarcinomas are diagnosed with tuberculosis and cytology, the options are as follows
cytology, but the yield is less with squamous cell car in the subsequent text. The first thing that we recom
cinoma, Hodgkin's disease, and sarcomas. Obviously, mend is a CT angiogram. With the CT angiogram,
the yield will also depend on the skill of the cytolo the possibility of pulmonary embolus can be evalu
gist. It also depends on the extent of the tumor-the ated and the presence of pulmonary infiltrates, pleural
greater the tumor burden in the pleural space, the masses, or mediastinal lymphadenopathy can also be
more likely the cytology is to be positive. evaluated ( 1 7) . If the CT angiogram scan does not
demonstrate a pulmonary embolus, then there are five
Pleural Fluid Markers for Tuberculosis options available to the physician, namely, observa
tion, needle biopsy of the pleura, bronchoscopy, tho
Over the last 50 years, the diagnosis of tuberculous racoscopy, or thoracotomy with open biopsy.
pleuritis was usually established with needle biopsy
of the pleura. However, it is now possible to make
O bservation
the diagnosis of pleural tuberculosis by measuring
the level of ADA or the level of interferon-gamma This is probably the best option if the patient is
in the pleural fluid. improving and there are no parenchymal infiltrates.
1 36 PLE U RAL D I S EASES
reported that thoracotomy failed to provide a spe pleural effusion particularly if the effusion is septated
cific diagnosis for 7 of 21 patients with an undiag or hyperechoic.
nosed pleural effusion. The group at the Mayo Clinic Patients in the ICU who are receiving mechani
reviewed their experience with open pleural biopsy cal ventilation and have large pleural effusions appear
for undiagnosed pleural effusion between 1 962 and to benefit if the fluid is drained. Kupfer et al. (3 1 )
1 972 and reported that during this time no diagnosis compared the duration o f mechanical ventilation for
was established in 5 1 such patients (27) . In 3 1 of the 80 patients who received chest tubes and 88 who did
patients (6 1 %), there was no recurrence of the pleural not receive chest tubes and reported that the duration
effusion, and no cause ever became apparent. How of mechanical ventilation was significantly shorter
ever, 1 3 of the patients (25 . 5%) eventually proved to (3.8 days) in the patients that received chest tubes
have malignant disease (6 patients with lymphoma, than in the patients that did not receive chest tubes
4 patients with mesothelioma, and 3 with other (6. 5 days) . All the patients had transudative effu
malignancies) . sions and the decision as to whether to insert a chest
tube was made by the patient's private physician
(3 1 ) . When patients on mechanical ventilation are
S P E C I A L S IT U AT I O N S subjected to therapeutic thoracentesis their oxygen
There are certain special situations in which a differ ation status improves (32,33) . Walden et al. (32) per
ent approach is sometimes indicated. Such situations formed 1 5 thoracentesis with the removal of a mean
include patients with pleural effusions in the inten of 1 ,872 ml in 1 0 patient on mechanical ventilation.
sive care unit (ICU) , patients with massive effusions Post thoracentesis, the mean Pa0 2 increased from 82
and those with bilateral pleural effusions. to 1 1 5 mm Hg and the P:F ratio increased from 1 69
to 237 (32) . These improvements were maintained
for 48 hours (32) . Liang and coworkers (34) have
Pleural Effusions in the Intensive Care U nit demonstrated that drainage of large pleural effusion
in the ICU via pigtail catheters is effective regardless
There is a high prevalence of pleural effusion in
of the etiology of the effusion.
patients in the ICU. Mattison et al. performed ultra
sound examinations of the chest and found that
pleural effusions were present in 62% of 1 00 patients
Massive Pleural Effusions
(28) . In this series, 92% of the effusions were small
and most were thought to be transudates (28) . In a On occasion a patient is found to have a pleural
second study, Fartoukh et al. screened 1 ,35 1 patients effusion that occupies the entire hemithorax and
in the medical intensive care unit (MICU) and found such an effusion is said to be massive. As discussed
that 1 1 3 patients (8 .4%) had a pleural effusion in Chapter 6, it is important to assess the position
detectable by physical examination and obscuring of the mediastinum in such patients before proceed
one third of the lung field (29) . Thoracentesis on 82 ing with the workup. Poree! et al. (3 5) reviewed chest
of these patients revealed transudates in 20 (24.4%) , radiographs from 766 patients with pleural effusion
parapneumonic effusion or empyema in 35 (42.7%) seen at an academic medical center in Lleida, Spain,
and noninfectious exudates in 27 (32.9%) (29) . In during a 1 0-year period. Ninety-three of the patients
a third study, Tu et al. (30) performed thoracente ( 1 2%) had a massive pleural effusion and 70 addi
sis on 94 patients in the MICU who had ultrasono tional patients (9%) had effusions occupying more
graphic evidence of pleural effusions and who had a than two thirds of the hemithorax. This distribution
temperature greater than 38°C for more than 8 hours. of the diagnoses in the patients with massive and
They reported that 62% of the patients had infectious large effusions was similar. Of the 93 patients with
exudates which included 36 parapneumonic effu massive effusions, 5 5 (59%) had malignancy, 2 1
sions, 1 5 empyemas, 1 5 urosepsis, 2 liver abscesses, (23%) had parapneumonic effusions, 9 ( 1 0%) had
1 deep neck infection, and 1 wound infection. The tuberculous effusions, 4 had other exudates, and 4
sonographic pattern of the effusion was hyperechoic had transudative effusions. All the massive parapneu
and/or septated in all 1 5 patients with empyema (30) . monic effusions were either empyemas or compli
The studies in the preceding text proves that there is a cated parapneumonic effusions (35). Although only
high incidence of empyema in the patient with effu one patient in Porcel's series had a hepatic hydrotho
sion in the MICU. A thoracentesis is recommended rax, it has been my experience that hepatic hydro
for patients in the ICU with more than a minimal thorax is one of the more common causes of massive
1 38 PLE U RAL D I S EASES
2 1 . de Groot M, Walther G. Thoracoscopy in undiagnosed pleural 3 1 . Kupfer Y, Senevirame C, Chawla K, et al. Chest rube drain
effusions. S Afr Med]. 1 998;88 :706-7 1 1 . age of transudative pleural effusions hastens liberation from
22. Diacon AH, Van de Wal BW, Wyser C , et al. Diagnostic tools mechanical ventilation. Chest. 20 1 0 ; 1 39 : 5 1 9-523.
in tuberculous pleurisy: a direct comparative study. Eur Respir]. 32. Walden AP, Garrard CS, Salmon J. Sustained effects of thora
2003;22:5 89-5 9 1 . cocentesis on oxygenation in mechanically ventilated patients.
2 3 . Daniel TM. Diagnostic thoracoscopy fo r pleural disease. Ann Respirology. 20 1 0; 1 5 :986-992.
Thorac Surg. l 993;56:639-640. 33. Chen WL, Chung CL, Hsiao SH, et al. Pleural space elastance
24. Light RW. Closed needle biopsy of the pleura is a valuable and changes in oxygenation after therapeutic thoracentesis in
diagnostic procedure. Con closed needle biopsy. J Bronchol. ventilated patients with heart failure and rransudative pleural
1 999;5 :332-336. effusions. Respirology. 20 1 0; 1 5 : 1 00 1 - 1 008.
25. Prakash URS, Reiman HM. Comparison of needle biopsy 34. Liang SJ, Tu CY, Chen HJ, et al. Application of ultrasound
with cyrologic analysis for the evaluation of pleural effusion: guided pigtail catheter for drainage of pleural effusions in the
analysis of 4 1 4 cases. Mayo Clin Proc. 1 9 8 5 ;60: 1 5 8- 1 64. ICU. Intensive Care Med. 2008 ;35:350-354.
26. Douglass BE, Carr DT, Bernatz PE. Diagnostic thoracoromy 3 5 . Poree! JM, Vives M . Etiology and pleural Buid characteristics
in the study of "idiopathic" pleural effusion. Am Rev Tuberc. of large and massive effusions. Chest. 2003; 1 24:978-983.
1 9 56;74:9 54-9 57. 36. Jimenez D, Diaz G, Gil D, et al. Etiology and prognostic sig
27. Ryan CJ, Rodgers RF, Unni KK, et al. The outcome of patients nificance of massive pleural effusions. Respir Med. 2005;99:
with pleural effusion of indeterminate cause at thoracotomy. 1 1 83-1 1 87.
Mayo Clin Proc. 1 9 8 1 ; 56: 1 45-149. 37. Rabin CB, Blackman NS. Bilateral pleural effusion. Its sig
28. Mattison LE, Coppage L, Alderman DF, et al. Pleural effusion nificance in association with a heart of normal size. J Mt Sinai
in the medical ICU. Prevalence, causes and clinical implica Hosp N Y. l 957;24:45-63.
tions. Chest. 1 997; 1 1 1 : 1 0 1 8- 1 023. 38. Kalomenidis I, Rodriguez M, Barnette R, et al. Patient with
29. Fartoukh M, Azoulay E, Galliot R, et al. Clinically docu bilateral pleural effusion: are the findings the same in each
mented pleural effusions in medical ICU patients. How useful Buid? Chest. 2003; 1 24 : 1 67- 1 76.
is routine thoracentesis? Chest. 2002; 1 2 1 : 1 78- 1 84. 39. Poree! JM, Civit MC, Bielsa S, et al. Contarini's syndrome:
30. Tu CY, Hsu WH, Hsia TC, et al. Pleural effusions in febrile Bilateral pleural effusion, each side from different causes.
medical ICU patients: chest ultrasound study. Chest. 2004; J Hosp Med. 2 0 1 2;7: 1 64-1 6 5 .
1 26: 1 274- 1 280.
Transudati ve Pl eural E ffusions
Transudative pleural effusions occur when the systemic undergone significant modifications. In the past, it
factors influencing the formation and absorption of was believed that the pleural Buid that accumulated
pleural Buids are altered so that pleural Buid accumu in patients with CHF was due to increased pressure
lates. In this chapter, the various causes of transudative in the capillaries in the visceral or the parietal pleura.
pleural effusions are discussed. These increased pressures were thought to result in an
increased entry of Buid into the pleural space from
CO N G E ST I V E H E A RT FA I L U R E the parietal pleura and a decreased removal of Buid
through the visceral pleura, according to Starling's
Congestive heart failure (CHF) is probably the most equation.
common cause of pleural effusion. The reason for the The current theories on pleural Buid formation and
low incidence of pleural effusions secondary to heart reabsorption give us a different entry pathway and a
failure in most studies is that researchers interested in
different exit pathway for pleural Buid in patients
pleural effusions usually do not see most patients with with CHE It appears that most of the Buid that
pleural effusions of this origin. In an epidemiologic
enters the pleural space in patients with CHF comes
study from the Czech Republic, CHF was the most
from the alveolar capillaries rather than the pleural
common cause of pleural effusion ( 1 ) . The incidence
capillaries (4) . When the pressure in the pulmonary
of pleural effusions in patients with CHF is high.
capillaries is elevated, increased amounts of Buid enter
Kataoka and Takada (2) studied 60 patients admitted
the interstitial spaces of the lung. The increased Buid
to a Japanese hospital for an exacerbation of stable
in the interstitial spaces results in an increased inter
CHF with a computed tomography (CT) scan, ultra
sound, and a chest radiograph. They reported that by stitial pressure in the subpleural interstitial spaces (5) .
CT scan, 50 patients (83%) had a right-sided pleural The Buid then moves from the pulmonary intersti
effusion, whereas 46 patients (77%) had a left-sided tial spaces across the visceral pleura into the pleural
pleural effusion. Approximately one third of the effu space. There appears to be relatively little resistance
sions had volumes that exceeded 700 mL (2) . to Buid movement from the pulmonary interstitial
Race et al. (3) reviewed the autopsies at the Mayo spaces across the visceral pleura (4) . When pulmonary
Clinic between 1 948 and 1 9 5 3 of 402 patients who edema is produced in sheep with volume overloading,
had CHF during life. The researchers found that 290 approximately 25% of the pulmonary edema Buid
of the patients (72%) had pleural effusions with vol exits the lung through the visceral pleura (6) .
umes greater than 250 mL. Of these, 88% had bilateral Currently, it is believed that almost all Buid exits
pleural effusions, whereas 8% and 4% had unilateral the pleural space through the lymphatics in the pari
right-sided and left-sided effusions, respectively (3) . etal pleura rather than by passively diffusing across
the visceral pleura (see Chapter 2) . Pleural Buid accu
mulates in patients with CHF when the rate of entry
Pathophysiology
of Buid into the pleural space exceeds the capacity of
In recent years, concepts of pleural Buid formation the lymphatics in the parietal pleura to remove the
and reabsorption in patients with heart failure have Buid. In normal sheep, the capacity of the lymphatics
1 40
C H APTE R 9 I TRA N S U DATIVE PLE U RAL E F F U S I O N S 141
to remove fluid is approximately 0.28 mL/kg/hr (7) . the entry o f fluid into the pleural space overwhelms
If there is elevated pressure in the systemic veins, the the capacity of the lymphatics in the parietal pleura to
lymphatic clearance is decreased and the rate of fluid remove the fluid. Small amounts of fluid may
formation from the capillaries in the parietal pleura is enter the pleural space from the capillaries in either
increased ( 8). pleural surface. Elevation of the systemic venous
In the clinical situation, it appears that the accu pressure also leads to the accumulation of pleural
mulation of pleural fluid in patients with CHF is fluid because of increased fluid formation from the
related more to left ventricular failure than to right capillaries in the parietal pleura and decreased lym
ventricular failure. Wiener-Kronish et al. (9) prospec phatic clearance from the pleural space.
tively evaluated 37 patients with CHF secondary to
ischemic heart disease or to cardiomyopathy who were
Clinical Manifestations
admitted to a coronary care unit. Nineteen patients
had a pleural effusion. The mean wedge pressure in Pleural effusions due to CHF are usually associated
the patients with an effusion (24. 1 ± 1 .3 mm Hg) was with other manifestations of that disease. The patient
significantly higher than in those without an effusion often has a history of increasing dyspnea on exer
( 1 7.2 ± 1 . 5 mm Hg) . There was also a greater likeli tion, increasing peripheral edema, and orthopnea or
hood of finding pleural effusions if severe rather than paroxysmal nocturnal dyspnea. The dyspnea is fre
mild pulmonary edema was found roentgenographi quently out of proportion to the size of the effusion.
cally. In a subsequent study ( I O) , these same research Physical examination usually reveals signs of both
ers were unable to demonstrate any pleural effusions right-sided heart failure with distended neck veins
in 27 patients with chronic pulmonary hypertension and peripheral edema and left-sided heart failure
or chronically elevated right atrial pressures. with rales and an 53 ventricular gallop as well as signs
Pleural effusions also occur in some patients with of the pleural effusions.
right ventricular failure ( 1 1 - 1 3 ) . Tang et al ( 1 1 ) The chest radiograph almost always reveals
reviewed the records o f 1 47 patients with idiopathic cardiomegaly and usually bilateral pleural effusions.
or familial hypertension and reported that 2 1 patients CHF is by far the most common cause of bilateral
( 1 4%) had pleural effusion for which there was no pleural effusions, but if cardiomegaly is not present,
explanation. The patients with pleural effusion had an alternate explanation should be sought. In one
significantly higher mean right atrial pressures than series of 76 patients with bilateral pleural effusions
did those without effusion ( 1 1 ) . The majority of the but a normal-sized heart, only 3 (4%) were due to
effusions were trace to small and right sided or bilat CHF ( 1 4) . Although in the past it was thought that
eral ( 1 1 ) . When the fluid was analyzed, it usually was pleural effusions due to CHF were commonly unilat
a transudate ( 1 1 ) . In another study by the same group, eral on the right or at least were much larger on the
Luo et al. ( 1 2) reviewed the records of 89 patients right side, this does not appear to be the case. In the
with pulmonary arterial hypertension due to con autopsy series of Race et al. (3) , 88% of the patients
nective tissue disease and reported that 32.6% of the studied had bilateral pleural effusions. Moreover, the
patients had a pleural effusion without an alterna mean volume of pleural fluid in the right pleural
tive diagnosis. The patients with pleural effusion had space ( 1 ,084 mL) was only slightly greater than the
significantly higher right atrial pressures, BNP, and mean volume of pleural fluid in the left pleural space
lower cardiac outputs ( 1 2) . Patients with scleroderma (9 1 3 mL) . In this series, 35 patients had unilateral
had a higher incidence than did patients with sys pleural effusions, and of these 35 patients, 1 6 (46%)
temic lupus erythematosus ( 1 2) . The pleural effusions had either pulmonary embolism or pneumonia (3) .
were predominantly small in size and bilateral. It A recent letter ( 1 5) to the editor summarized the
therefore appears that many patients with right heart sidedness of 444 effusions due to heart failure in five
failure have pleural effusions ( 1 3) . separate studies. Of the effusions, 69% were bilateral,
In summary, it appears that pleural fluid accumu 2 1 % were unilateral on the right, and 9% were uni
lates in patients with CHF when they have either left lateral on the left. Of the bilateral pleural effusions,
or right ventricular failure. The high pressures in the most were similarly sized but either side can be larger.
pulmonary capillaries lead to increased amounts of In patients with pleural effusions secondary to
fluid in the interstitial spaces. The fluid in the intersti CHF, mediastinal lymphadenopathy is common. Erly
tial spaces enters the pleural space through the highly et al. ( 1 6) reported that the prevalence of mediasti
permeable visceral pleura. Fluid accumulates when nal lymph nodes with a diameter greater than 1 cm
1 42 PLE U RAL D I S EASES
on the short axis was 81 % in 36 patients with an ejec Th e pleural fluid from approximately 2 0 % t o 2 5 %
tion fraction less than 35%. o f patients with CHF will b e classified as exudates by
Light's criteria (22) . Most patients who are misclas
sified are receiving diuretics (22) . If the pleural fluid
Diagnosis
meets exudative criteria but the effusion is thought to
The diagnosis of pleural effusions secondary to CHF be due to CHF, the serum to pleural fluid protein gra
comes readily to mind every time a patient is seen dient should be examined. If this gradient is greater
with CHE One must be careful to avoid the trap of than 3 . 1 g/dL, the pleural effusion in all probability is
ascribing the pleural effusion to CHF when it has due to the CHF and additional diagnostic studies are
another cause. In the series of Race et al. (3) , more not indicated ( 1 7) . In patients with CHF on diuretic
than 25% of the patients with CHF and pleural effu therapy, the protein gradient does not decrease much
sions had either pulmonary emboli or pneumonia at with diuresis ( 1 7) . If the protein and LDH criteria
autopsy. Certainly, if the patient is febrile, has pleural are not met and the protein gradient is less than 3 . 1
effusions that are greatly disparate in size, has a uni g/dL, the pleural effusion is probably not due to the
lateral pleural effusion, pleuritic chest pain, or does heart failure. Rather, the patient has an exudative
not have cardiomegaly, a diagnostic thoracentesis pleural effusion, and further diagnostic tests such as
should be performed. pleural fluid cytologic study and a CT angiogram of
If the patient has cardiomegaly and bilateral pleu the chest should be obtained. In previous editions of
ral effusions, is afebrile, and does not have pleuritic this book, it was recommended that in borderline
chest pain, we initiate treatment of the CHF and pleural effusions the albumin gradient between the
observe the patient to determine whether the pleural serum and pleural fluid be measured. If this gradi
fluid is reabsorbed. If the effusions do not disappear ent was more than 1 .2 g/dL, then the effusion could
within a few days, we then perform a diagnostic tho be ascribed to the CHE Currently, the protein gradi
racentesis. One problem with this approach is that ent of 3 . 1 g/dL is preferred to the albumin gradient
with diuresis, the characteristics of the pleural fluid because the protein gradient is already available when
may change from those of a transudate to those of Light's criteria are measured. Moreover, the protein
an exudate. Romero et al. ( 1 7) performed a thora gradient appears to be as effective as the albumin gra
centesis on 2 1 patients with CHF before and every dient in making the distinction ( 1 7) .
48 hours after diuretic therapy was initiated. Before Another test that should b e considered fo r estab
diuretics were administered, only one effusion was lishing the diagnosis of CHF is measurement of the
misclassified as an exudate by Light's criteria, but serum or pleural fluid N-terminal pro-brain natri
at the time of the third thoracentesis 1 0 effusions uretic peptide (NT-proBNP) . When the ventricles are
were misclassified as exudates. Between the first subjected to increased pressure or volume, the biologi
and the third thoracentesis, the mean protein level cally active BNP and the larger aminoterminal part
increased from 2.3 to 3 . 5 gm/dL and the mean lac NT-pro-BNP are released in equimolar amounts in
tate dehydrogenase (LDH) level increased from 1 76 the circulation (23) . The thresholds for the diagnosis
to 262 IU/L ( 1 7) . Other authors ( 1 8 , 1 9) have also of heart failure via the BNP recommended by the
reported that the characteristics of the pleural fluid manufacturers are 1 00 pg/ mL for the Triage BNP assay
changed with diuresis. and 1 2 5 pg/mL for the Elecsys pro-BNP assay (23) .
The pleural fluid from a patient with CHF is typi However, in clinical practice, levels below 1 00 pg/mL
cally a transudate with a ratio of pleural fluid to serum are thought to make CHF unlikely, whereas levels
protein below 0.5, a ratio of pleural fluid to serum above 500 pg/mL are considered diagnostic of CHF
LDH under 0.6, and an absolute pleural fluid LDH (24) . It should be noted that serum BNP levels are
level below two-thirds of the upper limit of normal for increased with acute pulmonary embolism when
serum (Light's criteria) (20) . If the foregoing criteria there is right ventricular dysfunction (25) .
are satisfied in a patient with CHF, the patient has a The pleural fluid levels of NT-proBNP are elevated
transudative pleural effusion that can be ascribed to the in patients with heart failure. Poree! et al. (26) mea
CHF, and no further diagnostic studies are indicated. sured NT-proBNP levels in 1 1 7 pleural fluid samples
Such transudative pleural effusions may be blood including 44 with heart failure, 25 with malignancy, 20
tinged, and the pleural fluid differential cell count may with tuberculous pleurisy, and 1 0 with hepatic hydro
reveal predominantly polymorphonuclear leukocytes, thorax. They reported that the median level of NT
small lymphocytes, or other mononuclear cells (2 1 ) . proBNP was 6,93 1 pg/mL in the patients with CHF,
C H APTE R 9 I TRA N S U DATIVE PLE U RAL E F F U S I O N S 1 43
which was significantly higher than that of 5 5 1 pg/ mL In one study of 20 patients with heart failure who
in the patients with hepatic hydrothorax or 292 pg/ met Light's criteria for exudates, 1 8 had NT-pro-BNP
mL in the 63 patients with exudative effusions. A cut levels above 1 ,300, 16 had NT-pro-BNP levels above
off level of 1 ,500 pg/mL in the pleural fluid provided 1 ,500, but only 1 0 had serum pleural fluid protein
a sensitivity of 9 1 % and a specificity of 93% in the gradient greater than 3. 1 g/dL (3 1 ) .
diagnosis of heart failure. Tomcsanyi et al. (27) com
pared the pleural fluid and serum NT-pro-BNP levels
Treatment
in 14 patients with CHF and 1 4 patients with pleural
effusions of other etiologies. They reported that the The preferred treatment of pleural effusion second
median NT-pro-BNP levels in the patients with heart ary to heart failure is to treat the heart failure with
failure and other diseases were 6,295 and 276, respec digitalis, diuretics, and afterload reduction. When the
tively, in pleural fluid and 5, 7 1 3 and 23 1 , respectively, heart failure is successfully managed, the pleural effu
in serum (27) . There was no overlap between the two sion disappears. Such treatment effectively manages
groups. Interestingly, in the latter study, the correla the pleural effusion in most patients with heart failure.
tion between the pleural fluid BNP levels and the For example, in a recent study of patients undergoing
serum levels was very high (If = 0.95) (27) . Kolditz orthotopic heart failure, preoperatively only 1 9 of 60
et al. (28) measured the serum and pleural fluid NT (32%) had a pleural effusion and only 1 patient had a
pro-BNP levels in 93 patients including 25 with CHE pleural effusion that occupied more than 25% of the
They confirmed the results of the study by Tomcsanyi hemithorax (33) .
et al. (27) in that the levels of serum and pleural fluid Occasionally, large pleural effusions cause patients
NT-pro-BNP again were closely correlated (R2 = to be very dyspneic. The removal of 500 to 1 ,000 mL
0.90). They reported that an NT-pro-BNP cutoff level of pleural fluid from such patients may rapidly relieve
of 4,000 pg/mL had a sensitivity of 92% and a speci the dyspnea. If the effusions cannot be controlled
ficity of 9 3% in making the diagnosis of CHE More with standard therapy, interventions to control the
over, in this study nine patients with heart failure met pleural effusions should be considered. One option is
Light's exudative criteria and all of them had pleural pleurodesis with a sclerosing agent (34) and currently,
and serum NT-pro-BNP levels greater than 4,000 pg/ doxycycline 500 mg through a chest tube is recom
mL (28) . From the latter two studies, it appears that mended (see Chapter 1 0) . Bleomycin is not recom
there is no need to measure both the pleural fluid and mended in this situation because it is not an effective
the serum NT-pro-BNP levels. agent in the rabbit model with normal pleura (35).
We compared the pleural fluid NT-pro-BNP levels An alternative approach is to insert an indwelling
in 1 0 patients each with pleural effusions due to pleural catheter or to use a pleuroperitoneal shunt.
CHF, pulmonary embolism, post-coronary artery Herlihy et al. (36) inserted indwelling catheters
bypass surgery, and malignancy. All the patients with (PleurX, CareFusion Corporation, San Diego, CA)
CHF had NT-pro-BNP levels above 1 ,500 pg/mL, into five patients with pleural effusions refractory to
whereas none of the other patients had such high the usual therapy. Initially, the catheter was effective
levels (29) . In view of the data given in the preced in controlling the effusion in all patients (36) . How
ing text, it appears that a pleural fluid NT-pro-BNP ever, after 5 and 1 5 months, two patient developed an
level above 1 ,500 pg/mL is almost diagnostic that the empyema and after 4 months another patient devel
patient has CHE oped a loculated pleural effusion (36) . We have used
It should be emphasized that the serum or pleural this device in several patients with excellent results.
fluid BNP and NT-pro-BNP cannot be used inter With the PleurX, the patient connects the catheter
changeably in the diagnosis of pleural effusions due to to vacuum bottles every other day to drain the
CHF (30) . The NT-pro-BNP levels are about 1 0 times fluid (see Chapter 1 0) . The pleuroperitoneal shunt
higher than the BNP levels. There is not a close corre (CareFusion Corporation, San Diego, CA) consists of
lation between the BNP levels and the NT-pro-BNP two catheters connected with a valved pump cham
levels ( r = 0.78) (3 1 ) . The diagnostic usefulness of the ber. The two one-way valves in the pump chamber
NT-pro-BNP in making the diagnosis of heart failure are positioned such that fluid can only flow from the
is superior to that of the BNP (3 1 ,32) . The pleural pleural space to the peritoneal cavity through the
fluid NT-pro-BNP is also superior to the BNP and pump chamber. Because the pleural pressure is almost
the protein gradient in identifying patients with heart always more negative than the peritoneal pressure,
failure who meet Light's criteria for exudates (3 1 ) . the pumping chamber must be used to move fluid
1 44 PLE U RAL D I S EASES
from the pleural cavity to the peritoneal cavity. Little vessels. This conclusion appears to have been incor
et al. (37) reported that two patients with refractory rect. Datta et al. (42) injected radiolabeled human
pleural effusions secondary to CHF were managed serum albumin into the peritoneal cavity of a patient
successfully with the pleuroperitoneal shunt. with ascites and a large pleural effusion. They were
able to demonstrate that when the labeled protein
was picked up by the lymphatic system in the dia
H E PAT I C H Y D R O T H O RAX
phragm, it flowed into normal mediastinal lymphatic
Pleural effusions occur occasionally as a complica channels and from them into the subclavian vein. It
tion of hepatic cirrhosis. Pleural effusions usually did not enter the pleural space.
occur only when ascitic fluid is present. Lieberman Studies by Lieberman et al. ( 38 ) suggest that fluid
et al. (38) reviewed 330 patients with cirrhosis and passes directly from the peritoneal to the pleural cav
ascites and found that 1 8 (5 .5%) had pleural effu ity through pores in the diaphragm. These researchers
sions. Johnston and Loo (39) found that 6.0% of introduced 500 to 1 ,000 mL of air into the perito
200 patients with cirrhosis had pleural effusions. In the neal cavity of five patients with cirrhosis, ascites, and
second series, none of the 54 patients having cirrhosis pleural effusions. In all five patients, a pneumothorax
without ascites had a pleural effusion (39) . In some developed 1 to 48 hours after the induction of the
patients, the ascites is not clinically evident, but it can pneumoperitoneum. Thoracoscopic examination was
almost always be demonstrated with ultrasonography performed in three other patients after the induc
(40) . The pleural effusion in patients with cirrhosis tion of the pneumoperitoneum, and in one of these
and ascites is usually right sided (67%) , but occasion patients, air bubbles were seen coming through an
ally it is left sided ( 1 6%) or bilateral ( 1 6%) (38,39) . otherwise undetectable diaphragmatic defect (38 ) .
At postmortem examination, diaphragmatic defects
were demonstrated in two of the patients (38) .
Pathophysiology
Huang et al. (43) prospectively studied 1 1 patients
Patients with cirrhosis frequently have decreased with hepatic hydrothorax who underwent thoracos
plasma oncotic pressure (39), and from Figure 2. 1 , copy. These researchers classified the morphologic
one might hypothesize that the pleural effusions arise findings into four morphologic types: type I, no obvi
because of it. Indeed, in the experimental animal, the ous defect (one patient) ; type II, blebs lying on the
induction of decreased plasma oncotic pressure leads diaphragm (four patients); type III, broken defects
to the accumulation of pleural fluid (4 1 ) . However, (fenestrations) in the diaphragm (eight patients) ;
this mechanism does not appear to be the predomi and type IV, multiple gaps in the diaphragm (one
nant cause of pleural effusions in patients with cir patient) . The movement of fluid from the peritoneal
rhosis and ascites. Rather, the pleural effusions appear cavity through the diaphragm can also be demon
to be produced by movement of the ascitic fluid from strated by contrast enhanced ultrasound (44) .
the peritoneal cavity into the pleural cavity. From the foregoing studies, it is evident that
Johnston and Loo (39) demonstrated that after the pleural fluid in these patients originates from
the intraperitoneal injection of India ink, cells in the ascitic fluid. It is probable that the fluid passes
the pleural fluid contained many carbon particles, directly into the pleural space through defects in
whereas cells in the peripheral blood contained none. the diaphragm. In the patient with tense ascites and
In addition, after the intravenous injection of radiola increased intraabdominal pressure, the diaphragm
beled albumin, the albumin first appeared in the peri may be stretched, causing microscopic defects. The
toneal fluid and then in the pleural fluid. Following increased hydrostatic pressure in the ascitic fluid
the intraperitoneal injection of radiolabeled albumin, results in a one-way transfer of fluid from the peri
the concentration of the labeled protein was greater toneal to the pleural cavity. My experience with the
in the pleural fluid than in the plasma; after intra placement of chest tubes in such patients leads me to
pleural injection, the labeled protein appeared in the believe that the direct movement of fluid is the domi
plasma before it appeared in the peritoneal fluid (39) . nant mechanism. To control the symptoms of large
Because no air entered the pleural space following hydrothoraces in several patients with cirrhosis and
the intraperitoneal injection of carbon dioxide in one ascites, I have performed tube thoracostomy, followed
patient, these researchers concluded that the pleural by the injection of a sclerosing agent. In each instance,
effusion arose from the transfer of ascitic fluid from the placement of the chest tube was followed by rapid
the peritoneal to the pleural space by the lymphatic (within minutes) diminution in the amount of ascites.
C H APTE R 9 I TRA N S U DATIVE PLE U RAL E F F U S I O N S 1 45
does not look like pus (53) , I prefer to use the term Certain patients are refractory to salt restriction
spontaneous bacterial pleuritis to emphasize its simi and diuretics and remain symptomatic from the pres
larity with spontaneous bacterial peritonitis. Patients ence of the large pleural effusion. In such patients,
are diagnosed with spontaneous bacterial pleuritis the treatment of choice is liver transplantation (58) .
if they have cirrhosis along with a positive bacterial The presence of the hepatic hydrothorax does not
culture of the pleural fluid and a neutrophil count appear to adversely affect the results with liver trans
greater than 250 cells/mm3 or a negative culture and a plantation (58) . Interestingly, approximately 1 0% of
neutrophil count greater than 500 cells/mm3. Patients living donors for liver transplant will develop a pleu
with pneumonia are excluded. In a subsequent study, ral effusion (59).
Xiol et al. (53) prospectively studied 24 episodes of The next best treatment is to implant a transjugular
spontaneous bacterial pleuritis in 16 of 1 20 patients intrahepatic portal systemic shunt (TIPS) . A recent
( 1 3%) admitted with a diagnosis of hepatic hydro randomized controlled study demonstrated that TIPS
thorax. In 1 4 of the 24 episodes, there was con was superior to serial thoracentesis in patients with
comitant spontaneous bacterial peritonitis (53) . The refractory ascites (60) . In this study, 60 patients were
pleural fluid cultures were positive in 1 8 patients randomly assigned to TIPS or serial thoracentesis
and included 8 patients with Escherichia coli, 4 with and there was a significant survival advantage with the
streptococcus species, 3 with enterococcus species, TIPS procedure (60) . A large uncontrolled series on
and 2 with Streptococcus pneumoniae. In the largest the use ofTIPS for hepatic hydrothorax was reported
series, spontaneous bacterial pleuritis occurred in 8 1 by Gordon et al. (6 1 ) . In this series, 24 patients were
(20%) o f 390 patients with hepatic hydro thorax (54) . treated and 1 4 of the patients had complete relief of
Spontaneous peritonitis was concomitantly present symptoms after shunt placement and did not require
in 54% (54) . The pleural fluid with spontaneous further thoracentesis (6 1 ) . Five additional patients
bacterial pleuritis with the exception of the increased required fewer thoracenteses, but the remaining five
number of neutrophils is similar to that with hepatic patients developed worsening liver function and died
hydrothorax (54) . The pleural fluid/serum protein within 45 days (6 1 ) .
level is usually less than 0.50, the LDH ration is less I f neither TIPS nor liver transplantation i s feasible,
than 0.6 and the pH is above 7.30 (54) . the best alternative treatment is probably videotho
Patients with spontaneous bacterial pleuritis tend racoscopy with closure of the diaphragmatic defects
to have worse liver disease. The mortality rate in and pleurodesis (60,62) . In one report, 1 8 patients
the large series was 38% (54) . The diagnosis can be were subjected to 2 1 thoracoscopies with talc insuffia
made at the bedside with the use of reagent strips for tion (3 patients were subjected to a second procedure
leukocyte esterase designed for testing urine (5 5). after the first failed) (63) . Diaphragmatic defects were
The treatment of choice is an antibiotic to which the detected and closed in 5 of the 1 8 patients (28%) .
cultured bacteria are susceptible. Tube thoracostomy The procedure was effective in 1 0 of 2 1 patients
is usually not indicated (54,56) . (48%) . The median hospital stay was 1 5 days. The
precarious medical condition of patients with hepa
tothorax is reflected in the 30% mortality in the
Treatment
3 months following the surgery (63) . Mouroux et al.
The management of pleural effusions associated with (64) performed this procedure in eight patients using
cirrhosis and ascites should be directed toward treat talc insuffiation. Diaphragmatic defects were found
ment of the ascites because the hydrothorax is an and closed in six of the patients, and none had a
extension of the peritoneal fluid. The patient should recurrent pleural effusion. No defects were found in
be put on a low-salt diet, and diuretics should be the remaining two patients, but after talc insuffiation,
administered. The best diuretic therapy appears to be the effusions occupied only the lower one third of the
the combination of furosemide and spironolactone. hemithorax. Ferrante et al. (65) could find no dia
The initial starting dose is 40 mg of furosemide and phragmatic defects in 1 5 patients with hepatic hydro
1 00 mg of spironolactone. This combination appears thorax subjected to thoracoscopy and talc insuffiation.
to have the optimal ratio for the two diuretics. The In this series, the hepatic hydrothorax was controlled
doses can be increased up to 1 60 mg of furosemide by the talc insuffiation in 8 of the 1 5 (53%) patients
and 400 mg of spironolactone daily (57) . Serial ther (65) . I do not recommend talc insuffiation for the
apeutic thoracenteses are not indicated because the reasons discussed in Chapter 1 0, but would instead
pleural fluid rapidly reaccumulates. recommend pleural abrasion. One small series of five
C H APTE R 9 I TRA N S U DATIVE PLE U RAL E F F U S I O N S 1 47
diaphragm and clip the areas that are leaking. I would, by administering diuretics in conjunction with a
therefore, definitely recommend that this be done. low-sodium diet, angiotensin-converting enzyme
I would also recommend parietal pleurectomy rather inhibitors, and using nonsteroidal antiinBammatory
than talc to effect the pleurodesis due to the dangers of agents cautiously (8 1 ) . Serial therapeutic thoracente
talc inducing the acute respiratory distress syndrome. ses should not be performed because they only further
Some patients on peritoneal dialysis develop exu deplete the protein stores. In selected individuals
dative pleural effusions. Kwan et al. (77) retrospec who are symptomatic from the pleural effusion, one
tively reviewed 82 chronic peritoneal dialysis patients should consider a pleurodesis with a sclerosing agent.
that had pleural effusion and found that 22 had unex
plained exudative effusions. When the dialysis therapy Pulmonary Veno-Occlusive Disease
in these patients was intensified, the exudative effusion
tended to resolve (77) . The researchers concluded that Pulmonary veno-occlusive disease is a rare disease that
these exudative effusions were uremic pleural effusions is characterized pathologically by evidence of repeated
(see Chapter 23) . pulmonary venous thrombosis. The characteristic his
tologic feature of pulmonary veno-occlusive disease
is obstruction of pulmonary venules and veins by
O T H E R CAU S E S O F TRA N S U DAT I V E intimal fibrosis; intravascular fibrous septa are nearly
P L E U RA L E F F U S I O N S always present (82) . The etiology of this disease is
unknown and may be related to more than one source
Nephrotic Syndrome
or mechanism (82) .
Pleural effusion is common in patients with the Clinically, patients with pulmonary veno-occlusive
nephrotic syndrome. In one study of 52 patients, disease typically have slowly progressive dyspnea and
2 1 % had pleural effusions (78) . With the nephrotic orthopnea punctuated by attacks of pulmonary edema.
syndrome, the pleural effusions are usually bilateral Physical examination reveals signs of pulmonary
and are frequently infrapulmonary in location (78) . hypertension. Small pleural effusions are present in a
The mechanism responsible for the transudative pleu significant fraction of patients. In one study (83), CT
ral effusion associated with the nephrotic syndrome scans were obtained on eight patients and five of the
is probably the combination of decreased plasma patients had bilateral pleural effusions. The pleural
oncotic pressure and increased hydrostatic pressure. effusion occupied less than 5% of the hemithorax in
The increased hydrostatic pressure is due to salt reten four patients, but in the fifth patient the effusions occu
tion producing hypervolemia (79) . pied one third the volume of the thoracic cavity. In a
The diagnosis of pleural effusion secondary to the second study, however, pleural effusions were present in
nephrotic syndrome is not difficult in the typical clini only 4 of 1 5 patients (27%) (84) .
cal situation. A diagnostic thoracentesis should be per When a pleural effusion is seen in a patient with
formed to ascertain that the pleural Buid is indeed a pulmonary hypertension, the diagnosis of pulmo
transudate. One should always consider the possibility nary veno-occlusive disease should be considered
of pulmonary emboli in patients with the nephrotic syn although patients with pulmonary hypertension and
drome and pleural effusion. In one series of 36 patients right heart failure may also have pleural effusions
with the nephrotic syndrome, 22% had pulmonary ( 1 3) . The characteristics of the pleural Buid have not
emboli (80) . In addition, the nephrotic syndrome may been delineated, but it is likely that it is transudative
be due to or complicated by renal vein thrombosis, and (85) . The pleural effusions are probably related to the
in such instances, the incidence of pulmonary emboli increased interstitial Buid that results from obstruc
is high (80) . A lung scan or a CT angiogram should be tion of the pulmonary veins (85).
obtained in all patients with the nephrotic syndrome The diagnosis of pulmonary veno-occlusive disease
and pleural effusion. If the lung scan or CT angio can be definitively made only by surgical lung biopsy.
gram is equivocal, evidence of deep venous thrombosis The prognosis with this disease is grim, with most
should be sought with venograms, impedance plethys patients dying within 2 years of diagnosis. The only
mograms, or a pulmonary arteriogram. treatment that significantly improves the prognosis
The treatment of the pleural effusion associated of patients with pulmonary veno-occlusive disease is
with the nephrotic syndrome should be aimed at lung transplantation (82) . Patients with pulmonary
decreasing the protein loss in the urine to increase veno-occlusive disease who are treated with agents for
the plasma protein and to decrease the increased pulmonary hypertension frequently develop pulmo
extracellular volume. This is best accomplished nary edema (86) .
C H APTE R 9 I TRA N S U DATIVE PLE U RAL E F F U S I O N S 1 49
Glomerulonephritis Hypoalbuminemia
Patients with acute glomerulonephritis frequently From Starling's equation (Equation 2. 1 ) , hypoalbu
have pleural effusions. In a series of 76 children, 42 minemia would decrease the oncotic pressure of the
(55%) had pleural effusions (93) . The pleural effusions blood and increase the rate of pleural fluid formation.
are transudative and are probably due to increased It is unclear how frequently hypoalbuminemia causes
intravascular pressures because most patients have a pleural effusion. In patients with cirrhosis and
cardiomegaly or peripheral edema in addition to the hypoalbuminemia, the pleural effusions are related to
pleural effusions. the transdiaphragmatic transfer of ascitic fluid rather
1 50 PLE U RAL D I S EASES
22. Burgess LJ, Maritz FJ, Taljaard JJ. Comparative analysis of the 4 3 . Huang PM, Chang YL, Yang CY, et al. The morphology
biochemical parameters used to distinguish between pleural of diaphragmatic defects in hepatic hydrothorax: tho
transudates and exudates. Chest. 1 995; 1 07: 1 604- 1 609. racoscopic finding. J Thorac Cardiovasc Surg. 2005 ; 1 3 0 :
23. Pfister R, Schneider CA. Natriuretic peptides BNP and 14 1-145.
NT-pro-BNP: established laborarory markers in clinical prac 44. Foschi F G , Piscaglia F, Pompili M, e t al. Real-time contrast
tice or j ust perspectives. Clin Chim Acta. 2004;349:25-3 8 . enhanced ultrasound-a new simple tool for detection of
2 4 . Poree! JM. The use of probrain natriuretic peptide in pleural peritoneal-pleural communications in hepatic hydrothorax.
fluid for the diagnosis of pleural effusions resulting from heart Ultraschall Med. 2008;29 : 5 3 8-442 .
failure. Curr Opin Pulm Med. 200 5 ; 1 1 :329-333. 4 5 . Light RW, Jenkinson SG, Minh V, et al. Observations on pleu
2 5 . Pruszczyk P. N-terminal pro-brain natriuretic peptide as an ral pressures as fluid is withdrawn during thoracentesis. Am
indicator of right ventricular dysfunction. J Card Fail. 2005; Rev Respir Dis. 1 980; 1 2 1 :799-804.
1 1 (suppl l ) :S65-S69. 46. Lazaridis KN, Frank JW, Krowka MJ, et al. Hepatic hydro
26. Poree! JM, Vives M, Cao G , et al. Measurement of pro thorax: pathogenesis, diagnosis, and management. Am ] Med.
brain natriuretic peptide in pleural fluid for the diagnosis of 1 999; 1 07:262-267.
pleural effusions due to heart failure. Am J Med. 2004; 1 1 6: 47. Xiol X, Castelloce J, Corces-Beuc R, et al. Usefulness and com
4 1 7-420. plications of thoracentesis in cirrhotic patients. Am J Med.
27. Tomcsfoyi J, Nagy E, Soml6i M, et al. NT-brain natriuretic 2001 ; 1 1 1 :67-69.
peptide levels in pleural fluid distinguish between pleural tran 48. Bielsa S, Poree! JM, Castellote J, et al. Solving the Light's cri
sudates and exudates. Eur] Heart Fail. 2004;6:753-756. teria misclassification rate of cardiac and hepatic transudates.
28. Kolditz M, Halank M, Schiemanck S, et al. High diagnostic Respirology. 20 1 2 ; 1 7:72 1 -725 .
accuracy of NT-proBNP for cardiac origin of pleural effusions. 4 9 . Kakizaki S, Kacakai K , Yoshinaga T, e t al. Hepatic hydrochorax
Eur Respir]. 2006;28 : 1 44- 1 5 0 . in the absence of ascites. Liver. 1 998; 1 8 : 2 1 6-220.
29. Liao H, N a M J , Dikensoy 0, e t a l . The diagnostic value o f 50. Daly JJ, Potts JM, Gordon L, et al. Scintigraphic diagnosis
pleural fluid NT-pro-BNP levels in patients with cardiovascu of peritoneo-pleural communication in the absence of ascices.
lar diseases. Respirology. 2008; 1 3 : 5 3-57. Clin Nucl Med. 1 994; 1 9 :892-894.
30. Poree! JM. Utilization of B-type natriuretic peptide and 5 1 . Umino J, Tanaka E, Ichijoh T, et al. Hepatic hydrothorax in
NT-pro-BNP in the diagnosis of pleural effusions due to heart the absence of ascices diagnosed by intraperitoneal spraying of
failure. Curr Opin Pulm Med. 20 1 1 ; 1 7:2 1 5-2 1 9 . indocyanine green. Intern Med. 2004;43:283-288.
3 1 . Poree! J M , Martinez-Alonso M, Cao G, e t al. Biomarkers of 52. Xiol X, Castelloce ], Baliellas C, et al. Spontaneous bacterial
heart failure in pleural fluid. Chest. 2009 ; 1 36:67 1 -677. empyema in cirrhotic patients: analysis of eleven cases. Hepa
32. Long AC, O'Neal HR Jr, Peng S , et al. Comparison of pleural tology. 1 990; 1 1 :365-370.
fluid N-terminal pro-brain natriuretic peptide and brain 53. Xiol X, Cascellvi JM, Guardiola J, et al. Spontaneous bacterial
natriurecic-32 peptide levels. Chest. 20 1 0 ; 1 37 : 1 369-1 374. empyema in cirrhotic patients: a prospective srudy. Hepatology.
33. Misra H, Dikensoy 0, Rodriguez RM, et al. Prevalence 1 996;23 : 7 1 9-723.
of pleural effusions post orthotopic heart transplantation. 54. Chen CH, Shih CM, Chou JW, et al. Outcome predictors of
Respirology. 2007; 1 2 : 8 87-890. cirrhotic patients with spontaneous bacterial empyema. Liver
34. Glazer M, Berkman N, Lafair JS, et al. Successful talc slurry Int. 20 1 1 ; 3 1 : 4 1 7-424.
pleurodesis in patients with nonmalignant pleural effusion. 55. Cascelloce J, Lopez C, Gornals J, et al. Use of reagent strips for
Chest. 2000; 1 1 7: 1 404- 1 409. the rapid diagnosis of spontaneous bacterial empyema. J Clin
3 5 . Vargas FS, Wang N-S, Lee HM, et al. Effectiveness of bleomy Gastroenterol. 2005;39:278-28 1 .
cin in comparison to tetracycline as pleural sclerosing agent in 56. Tu CY, Chen CH. Spontaneous bacterial empyema. Curr Opin
rabbits. Chest. 1 993; 1 04: 1 5 82- 1 5 84. Pulm Med 20 1 2; 1 8 :3 5 5-3 5 8 .
36. Herlihy JP, Loyalka P, Gnananandh J, et al. PleurX catheter for 5 7 . Runyon BA. Care of patients with ascites. N Engl J Med.
the management of refractory pleural effusions in congestive 1 994;330:337-342.
heart failure. Tex Heart Inst]. 2009;36:38-43 . 5 8 . Xiol X, Tremosa G, Castellote J, et al. Liver transplanta
37. Little AG, Kodowaki MH, Ferguson MK, et al. Pleuroperito tion in patients with hepatic hydrochorax. Transpl Int.
neal shunting. Alternative therapy for pleural effusions. Ann 2005 ; 1 8 :672-675 .
Surg. 1 9 88;208 :443-450. 59. Kwon KH , Kim yw, Kim S I , e t al. Postoperative liver regen
3 8 . Lieberman FL, Hidemura R, Peters RL, et al. Pathogenesis eration and complication in live liver donor after partial hepa
and treatment of hydrothorax complicating cirrhosis with tectomy for living donor liver transplantation. Yonsei Med ].
ascites. Ann Intern Med. 1 966;64:34 1 -3 5 1 . 2003;44: I 069- 1 077.
39. Johnston RF, Loo RV. Hepatic hydrothorax: srudies to deter 60. Rossie M , Ochs A, Gulberg V, et al. A comparison of
mine the source of the fluid and report of thirteen cases. Ann paracentesis and transj ugular intrahepatic portosys
Intern Med. 1 964;6 1 : 385-40 1 . temic shunting in patients with ascites. N Engl J Med.
40. Rubinstein D , Mcinnes IE, Dudley FJ . Hepatic hydrochorax 2000;342: 1 7 0 1 - 1 707.
in the absence of clinical ascites: diagnosis and management. 6 1 . Gordon FD, Anastopoulos HT, Crenshaw W, et al. The
Gastroenterology. 1 9 8 5 ; 8 8 : 1 8 8-1 9 1 . successful treatment of symptomatic, refractory hepatic
4 1 . Mellins RB, Levine OR, Fishman AP. Effect o f systemic and hydrochorax with cransjugular intrahepacic porcosyscemic
pulmonary venous hypertension on pleural and pericardia! shunt. Hepatology. 1 997;25 : 1 366- 1 369.
fluid accumulation. J Appl Physiol. 1 970;29 : 5 64-569. 62. Huang PM, Kuo SW, Lee JM. Thoracoscopic diaphragmatic
42. Datta N, Mishkin FS, Vasinrapee P, et al. Radionuclide dem repair for refractory hepatic hydrothorax: application of pleu
onstration of peritoneal-pleural communication as a cause for ral flap and mesh onlay reinforcement. Thorac Cardiovasc Surg.
pleural fluid. JAMA. 1 984;252:2 I 0 . 2006;54 :47-5 0.
1 52 PLE U RAL D I S EASES
63. Milanez de Campos JR, Filho LOA, Werebe EC, et al. Tho 8 5 . Light RW. Effusions from vascular causes. In: Light RW,
racoscopy and talc poudrage in the management of hepatic Lee YCG, eds. Textbook of Pleural Diseases. London, Eng
hydrothorax. Chest. 2000; 1 1 8 : 1 3- 1 7. land: Arnold Publishers; 2003:289-296.
64. Mouroux J, Perrin C, Venissac N, et al. Management of pleu 86. Montani D, Achouh L, Dorfmuller P, et al. Pulmonary
ral effusion of cirrhotic origin. Chest. 1 996; 1 09 : 1 093- 1 096. veno-occlusive disease: Clinical, functional, radiologic, and
65. Ferrante D, Arguedas MR, Cerfolio RJ, et al. Video-assisted hemodynamic characteristics and outcome of 24 cases con
thoracoscopic surgery with talc pleurodesis in the manage firmed by histology. Medicine (Baltimore). 2008;87:220-233.
ment of symptomatic hepatic hydrothorax. Am J Gastroenterol. 87. Gallelli A, Pelaia G , Calderazzo M, et al. Transudative tight
2002;97 : 3 1 72-3 1 7 5 . pleural effusion due to compression of the brachiocephalic
66. Northup P G , Harmon RC, Pruett TL, e t al. Mechanical vein caused by an intrathoracic goiter. Monaldi Arch Chest
pleurodesis aided by peritoneal drainage: procedure for hepatic Dis. 2000; 5 5 : 2 1 0-2 1 1 .
hydrothorax. Ann Thorac Surg. 2009;87:245-250. 8 8 . Poree! JM. Unilateral pleural effusion secondary to brachio
67. Orman ES, Lok AS . Outcomes of patients with chest rube cephalic venous thrombosis: a rare complication of central
insertion for hepatic hydrothorax. Hepatol Int. 2009;3: 5 82-5 86. vein catheterization. Respiration. 2002;69:569.
6 8 . Ikard RW, Sawyers JL. Persistent hepatic hydrothorax after 89. Muthuswamy P, Alausa M, Reilly B . Clinical problem
petitoneojugular shunt. Arch Surg. 1 980; 1 1 5 : 1 1 25-1 1 27. solving. The effusion that would not go away. N Engl J Med.
69. Artemiou 0, Marta GM, Klepetko W, et al. Pleurovenous 200 1 ;345 :756-759.
shunting in the treatment of nonmalignant pleural effusion. 90. Belie JA, Milan D . Pleural effusion secondary to ureteral
Ann Thorac Surg. 2003;76:23 1 -233. obstruction. Urology. 1 979; 14:27-29.
70. Widrich WC, Johnson WC, Robbins AH, et al. Esophagogas 9 1 . Garcia-Pachon E, Padilla-Navas I . Urinothorax: case report
tric variceal hemorrhage: its treatment by percutaneous rranshe and review of the literature with emphasis on biochemical
patic coronary vein occlusion. Arch Surg. 1 978; 1 1 3 : 1 33 1 - 1 338. diagnosis. Respiration. 2004;7 1 : 5 33-536.
7 1 . Nomoto Y, Suga T, Nakajima K, et al. Acute hydrothorax in 92. Garcia-Pachon E, Romero S . Utinothorax: a new approach.
continuous ambulatory peritoneal dialysis-a collaborative Curr Opin Pulm Med. 2006; 1 2 : 2 5 9-263.
study of 1 6 1 centers. Am ] Nephrol. 1 9 89;9:363-367. 93. Kirkpatrick JA Jr, Fleisher DS. The roentgen appearance of
72. Rudnick MR, Coyle JF, Beck LH, et al. Acute massive hydro the chest in acute glomerulonephtitis in children. J Pediatr.
thorax complicating peritoneal dialysis, report of2 cases and a l 964;64:492-498 .
review of the literature. Clin Nephrol. 1 979; 1 2 : 3 8-44 . 9 4 . Gottehrer A , Roa J, Stanford G G , e t al. Hypothyroidism and
73. Tang S, Chui WH, Tang AW, et al. Video-assisted thoraco pleural effusions. Chest. 1 990;9 8 : 1 1 30-1 1 32 .
scopic talc pleurodesis is effective for maintenance of peri 9 5 . Smolar E N , Rubin J E , Avramides A , et a l . Cardiac tampon
toneal dialysis in acute hydrothorax complicating peritoneal ade in primary myxedema and review of the literature. Am ]
dialysis. Nephrol Dial Transplant. 2003; 1 8 : 804-808. Med Sci. 1 976;272:345-352.
74. Chow CC, Sung JY, Cheung CK, et al. Massive hydrothorax in 96. Beach C, Manthey DE. Tension hydrothorax due to ven
continuous ambulatory peritoneal dialysis: diagnosis, manage rriculopleural shunting. ] Emerg Med. 1 998; 1 6: 33-36.
ment and review of the literature. NZMed] 1 988;27:475-477. 97. Muramatsu H, Koike K. Pleural effusions appearing in
7 5 . Chow KM , Szeto CC, Li PK. Management options for the rehabilitation ward after ventriculoperitoneal shunts: a
hydrothorax complicating peritoneal dialysis. Semin Dial. report of two adult cases and a review of the literature. Brain
2003; 1 6:3 89-394. Inj. 2004; 1 8 : 835-844.
76. Halstead JC, Lim E, Ritchie AJ. Acute hydro thorax in CAPD . 98. Monla-Hassan J, Eichenhorn M, Spickler E, et al. Duro
Early thoracoscopic (VATS) intervention allows return to pleural fistula manifested as a large pleural rransudate: an
peritoneal dialysis. Nephron. 2002;92 :72 5-727. unusual complication of rransthoracic diskectomy. Chest.
77. Kwan BC, Chow KM , Pang WF, et al. Unexplained exudative 1 998; 1 1 4 : 1 786- 1 789.
pleural effusion in chronic peritoneal dialysis patients. Perit 99. Gupta SM, Frias ], Garg A, et al. Aberrant cerebrospinal
Dial Int. 2 0 1 0;30: 524-5 27. fluid pathway. Detection by scintigraphy. Clin Nucl Med.
78. Cavina C, Vichi G . Radiological aspects of pleural effusions 1 9 86; 1 1 : 5 93-594.
in medical nephropathy in children. Ann Radio! Diagn. 1 95 8 ; 1 00. Huggins JT, Sahn SA. Duropleural fistula diagnosed by
3 1 : 1 63-202. bera2-transferrin. Respiration. 2003;70:423-425 .
79. Kinasewitz GT. Transudative effusions. Eur Respir J 1 997; 1 0 1 . Eid AA , Keddissi J I , Kinasewitz GT. Hypoalbuminemia a s a
1 0 : 7 1 4-7 1 8 . cause of pleural effusions. Chest. 1 999; 1 1 5 : 1 066- 1 069.
8 0 . Llach F, Arieff AI, Massry S G . Renal vein thrombosis and 1 02. Mattison LE, Coppage L, Alderman DF, et al. Pleural effu
nephrotic syndrome: a prospective study of 36 adult patients. sion in the medical ICU. Prevalence, causes and clinical
Ann Intern Med. 1 975;83:8-14. implications. Chest. 1 997; 1 1 1 : 1 0 1 8- 1 023.
8 1 . Palmer BF. Southwestern internal medicine conference: 1 03 . Lowell JR. Pleural Ejfosions. A Comprehensive Review. Balti
nephrotic edema-pathogenesis and treatment. Am J Med Sci. more, MD: University Park Press; 1 977.
1 993;306:53-67. 1 04. Neusradr JE, Levy RC. Hemorrhagic pleural effusion in
82. Mandel J, Mark EJ, Hales CA. Pulmonary veno-occlusive dis Meigs' syndrome. JAMA. 1 968;204: 1 79- 1 80.
ease. Am ] Respir Crit Care Med. 2000; 1 62: 1 964- 1 973. 105. Solomon S, Farber SJ, Caruso LJ. Fibromyomata of the
8 3 . Swensen SJ, Tashjian JH, Myers JL, et al. Pulmonary veno uterus with hemorhorax. Meigs' syndrome? Arch Intern Med.
occlusive disease: CT findings in eight patients. A]R Am J 1 97 1 ; 1 27: 307-309.
Roentgenol. 1 996; 1 67:937-940. 1 06. Romero Candeira S, Hernandez Blasco L, Soler MJ, et al.
84. Resten A, Maitre S, Capron F, et al. [Pulmonary hypertension: Biochemical and cyrologic characteristics of pleural effu
CT findings in pulmonary veno-occlusive disease] [Article in sions secondary to pulmonary embolism. Chest. 2002; 1 2 1 :
French] . ] Radio!. 2003;84 ( 1 1 Pt 1 ) : 1 739- 1 74 5 . 465-469.
Pl eural E ffusi ons Rel ated to M etastati c
M ali gnanci es
Tu m o r n % N %
L u n g carci n o m a 275 43 32 24
B reast carc i n o m a 1 57 25 35 26
Lym p h o m a a n d l e u ke m i a 52 8 34 26
Ova r i a n carci n o m a 27 4 9 7
S a rcoma ( i n cl u d i n g m e l a n o m a ) 13 2 5 4
Ute r i n e a n d cervica l ca rci noma 6 1 3 2
Stomach carc i n o m a 18 3 1 1
C o l o n ca rci n o m a 9 0 0
Pa ncreatic carci n o m a 7 0 0
B l a d d e r ca rci n o m a 7 0 0
O t h e r ca rci noma 23 4 6 4
Pri m a ry u n known 40 6 8 6
Tota l 634 1 33
was similar whether the pleural fluid cytology was exploratory thoracotomy if the patient has no other
positive or negative ( 1 3) . It has been suggested that contraindication to curative resection.
patients with lung cancer and pleural effusion be clas On occasion, a significant pleural effusion will be
sified as M l a which would make them a stage IV ( 1 4) . recognized only at the time of thoracotomy. Ruffini
Another study also demonstrated that with multivari et al. ( 1 7) reported that 52 of 1 ,279 patients (4%) oper
ate analysis, the presence of a pleural effusion at the ated upon between 1 993 and 1 999 had a pleural effu
time of diagnosis adversely affected prognosis (6) . sion with a volume more than 1 00 mL. The median
How should the patient with bronchogenic car survival for 1 6 patients who were inoperable was
cinoma and an ipsilateral cytology-negative pleural 6 months, whereas the median survival of 8 patients
effusion be evaluated? Rodriguez-Panadero ( 1 5) per who had positive cytology but who underwent resection
formed thoracoscopy on 2 1 patients with lung cancer was 9 months. However, the 3- and 5-year survivals of
and an ipsilateral cytology-negative pleural effusion. the 2 1 patients with negative cytology who underwent
At thoracoscopy, only five patients were believed to resection were 68% and 54%, respectively ( 1 7) .
be potentially resectable, but when these five were When a patient with lung cancer i s found t o have
subjected to thoracotomy, their tumors were found to a pleural effusion and is not symptomatic from the
be unresectable because of mediastinal invasion ( 1 5) . effusion, should the effusion be treated? The answer
I n another older study from the Mayo Clinic, 5 of to this question is probably no. Tremblay et al. ( 1 8)
73 patients with bronchogenic carcinoma and ipsi studied 1 4 such patients and reported that no patient
lateral cytology-negative pleural effusions had long required any specific pleural intervention for the
term survival after the lung cancer was resected ( 1 6) . pleural effusion.
In view o f the two studies mentioned in the preceding Epidermal growth factor receptor (EGFR) muta
text, it is recommended that patients with broncho tions are strong determinants of tumor response to
genic carcinoma and an ipsilateral cytology-negative EGFR tyrosine kinase inhibitors in non-small-cell
pleural effusion undergo thoracoscopy. If the thora lung cancer. The EGFR mutation status can be deter
coscopy is negative, a computed tomography (CT) mined by analyzing the DNA in malignant cells in
scan of the chest should be obtained to evaluate the the pleural fluid ( 1 9,20) .
mediastinal lymph nodes. If the CT scan demon
strates lymph node enlargement, a mediastinoscopy
Breast Carcinoma
should be performed. If the CT scan demonstrates
no lymph node enlargement and the thoracoscopy The second leading cause of malignant pleural effusion
is negative, consideration should be given to an is metastatic breast carcinoma. Fracchia et al. (2 1 )
C H APTE R 1 0 I PLE U RAL E F F U S I O N S R E LAT E D TO M ETASTATIC MALI G N A N C I E S 1 55
reviewed 60 1 patients with disseminated breast car 20% to 70% have evidence of mediastinal disease and
cinoma and found that 48% had pleural effusions. 90% have evidence of disease elsewhere. For non
The effusions were large enough to warrant therapeu Hodgkin's lymphoma, large-cell lymphomas more
tic intervention in 48% of the patients. Goldsmith frequently have associated pleural disease than do
et al. (22) reviewed the autopsies of 365 patients small-cell lymphomas (33) . The presence of a pleural
who had died of disseminated breast carcinoma and effusion at the time of presentation does not adversely
reported that 46% had pleural effusions. Pleural effu affect complete remission or survival rates with non
sions were more common with lymphangitic spread Hodgkin's lymphoma (35). The cytology on the pleu
(63%) than without lymphangitic spread (4 1 %) ral fluid is positive in almost all cases (34,35). During
(22) . In this series, the pleural effusions were on the the course of the disease, as many as 40% will have
same side as the primary breast carcinoma in 58% a pleural effusion (36) . Approximately 20% of the
of these patients, on the opposite side in 26%, and effusions present with non-Hodgkin's lymphomas
on both sides in 1 6% (22) . In a second series, the are chylothoraces (34) .
effusion was ipsilateral in 70%, contralateral in 20%, Vieta and Craver (3 1 ) also reported that 1 2% of
and bilateral in 1 0% (23) . Ipsilateral effusions are less 1 5 8 patients with lymphatic leukemia and 4% of
common if radiotherapy was part of the initial treat 52 patients with myelogenous leukemia had pleural
ment (24) . With breast carcinoma, the mean interval effusions. Parietal pleural involvement, however, was
berween the development of the primary tumor and uncommon at autopsy with leukemia. The most com
the appearance of the pleural effusion is approxi mon cause of pleural effusion in patients with chronic
mately 2 and 5 years (24,25), but this interval can myeloid leukemia and acute lymphobastic leukemia
be as long as 20 years (26) . In patients with pleural is as a side-effect from the ryrosine-kinase inhibitor
effusions secondary to breast carcinoma, determina dasatinib (see Chapter 22) (37) .
tion of the steroid receptors in the effusion is useful Angioimmunoblastic T-cell lymphoma was for
in planning therapy. merly called angioimmunoblastic lymphadenopathy
One paper suggested that when a patient with (AILD) and is characterized by the acute onset of
breast carcinoma is found to have a pleural effusion, it constitutional symptoms, generalized lymphade
is preferable to treat the patient with systemic therapy nopathy, hepatosplenomegaly, anemia, and poly
plus pleurodesis rather than systemic therapy alone clonal hypergammaglobulinemia. Pathologically,
(27) . In this article, the mean pleural progression this disorder is characterized by extensive infiltration
free interval was 8.5 months in the 1 02 patients rhat of lymph nodes with atypical lymphocytes, prolif
underwent pleurodesis compared to 4. 1 months in the eration of arborizing small vessels, and the deposi
78 patients rhat did not undergo pleurodesis originally tion of amorphous acidophilic material ( 1 9) . Most
(27) . There was no difference in overall survival. cases of AILD contain monoclonal T-cell popula
tions as well as clonal cytogenetic abnormalities
(5 1 ) . AILD is accepted as a lymphoma and a distinct
Lymphomas
clinicopathologic entity in the current World Health
Lymphomas, including Hodgkin's disease, are the Organization (WHO) classification (5 1 ) . A charac
third leading cause of malignant pleural effusions teristic feature of angioimmunoblastic AITL, seen
(28) . The incidence of pleural effusion with Hodg in more than 9 5 % of all patients, is the presence
kin's disease at presentation has varied from 7% to of increased numbers of Epstein-Barr virus (EBV)
2 1 % (29,30) . During the course of the disease the infected cells compared with both normal lymph
incidence of pleural effusion is approximately 1 6% nodes and peripheral T-cell lymphomas (5 1 ) . The
(3 1 ) . Patients with Hodgkin's disease who have EBY-infected cells are mostly B cells and the EBV is
pleural effusions almost invariably have intratho not thought to have a primary role in the pathogen
racic lymph node involvement, frequently without esis of AITL (5 1 ) .
microscopic pleural involvement (32) . Most patients Approximately 40% o f patients with AITL have
with Hodgkin's disease and pleural effusion have the pleural effusions (38) . The pleural fluid is an exu
nodular sclerosis rype (33) . Approximately only 3% date with a preponderance of mononuclear cells
of the effusions present with Hodgkin's disease are (39) . Other findings on the chest radiograph include
chylothoraces. interstitial infiltrates and mediastinal or hilar ade
The reported incidence of pleural effusion at pre nopathy, each in 1 5 % to 20% of patients (40) . The
sentation in non-Hodgkin's lymphoma has varied diagnosis is made by biopsy examination of an
from 6% to 50% (29,30,34) . With this neoplasm, enlarged lymph node.
1 56 PLE U RAL D I S EASES
The incidence of malignant pleural effusion in in the series of Leckie and Tothill (45), a patient with
patients with multiple myeloma is approximately bronchogenic carcinoma had the second highest
1 % (4 1 ) . The effusions develop at an average of amount of protein entering the pleural space of the
1 2 months after the diagnosis of multiple myeloma. 40 patients studied. The mechanism by which pleu
Most patients will have associated pleural or chest ral metastases increase the permeability of the pleura
wall plasmacytomas or pulmonary parenchymal is not definitely known. However, we postulate that
lesions on CT scan (4 1 ) . it is due to the production of vascular endothelial
growth factor (VEGF) by the tumor (42) . Indeed, the
PAT H O P H Y S I O LO G I C F E AT U R E S
median level ofVEGF in pleural effusions secondary
to malignancy is much higher than that in patients
There are several different mechanisms that can be with effusions secondary to inflammatory disease
responsible for the development of a pleural effusion (46,47) . The pleural fluid VEGF levels are also higher
in patients with malignancy (Table 1 0.2) . Although in hemorrhagic malignant effusions than in nonhem
it is frequently written that lymphatic obstruction is orrhagic malignant pleural effusions (46) . VEGF is
the primary pathophysiologic abnormality respon one of the most potent agents known for increasing
sible for the pleural effusion with malignancy, this vascular permeability (48) . Yano et al. have developed
appears not to be true (42) . The basis for the con an animal model of a malignant pleural effusion by
tention that lymphatic obstruction is responsible is injecting human adenocarcinoma cells into the pleu
the observation that at postmortem studies, the pres ral space of nude mice (49) . The formation of pleural
ence of pleural effusions is correlated with metasta fluid in this model is markedly reduced if the animals
ses to the lymph nodes (43). However, the normal are given an inhibitor of the VEGF receptor (49) or
rate of pleural fluid formation is thought to be only if the cells are transfected with an antisense VEGF-
1 5 mL/ day. Therefore, if there was complete blockage 1 65 gene (50) (see discussion on malignant pleural
of the lymphatics, the rate of pleural fluid accumula effusions in Chapter 4) .
tion should only be 1 5 mL/ day. Certainly, the rate of It is likely that lymphatic blockade and the result
pleural fluid accumulation frequently exceeds this in ing decreased clearance of fluid from the pleural
patients with pleural malignancy (44) . Moreover, if space contributes to the accumulation of pleural fluid
the fluid accumulation was solely due to lymphatic although, for the reasons outlined earlier, this is not
obstruction, one would expect the fluid to be a tran the predominant mechanism in most cases. Leckie
sudate, but it is almost always an exudate. and Tothill (45) reported that the mean amount of
We believe that the most likely explanation for the protein leaving the pleural space in patients with
pleural effusion with metastatic disease to the pleura malignant pleural effusions was less than that leaving
is increased permeability of the pleura (42) . Indeed, the pleural space in patients with tuberculosis, pul
monary embolism, or congestive heart failure. This
decreased lymphatic drainage can occur through two
TAB L E 1 0. 2 • M e c h a n i s m s by W h i c h
separate mechanisms. First, because the fluid leaves
M a l i g n a nt D isease Leads to P l e u ra l
the pleural space through stomas in the lymphatic
Effus i o n s vessels in the parietal pleura (5 1), metastases to the pari
etal pleura that obstruct these stomas can decrease
D i rect Res u lt
fluid clearance. Second, the lymphatic vessels of the
P l e u r a l m etastases with i n creased permea b i l ity
P l e u r a l m etastases with obstruction of p l e u ra l
parietal pleura drain mainly through the mediasti
lym p h atic vesse ls
nal lymph nodes. Therefore, neoplastic involvement
M e d i asti n a l lym p h node i nvo lvement with of the mediastinal lymph nodes can decrease the
decreased p l e u ra l lym p h atic d r a i n a g e lymphatic clearance of the pleural space.
Thoracic d u ct i nterruption (chylothorax) Malignant tumors can also produce pleural
B ronc h i a l obstruction (decreased p l e u ra l p ressu res) effusions by obstructing the thoracic duct, in which
Perica rd i a ! i nvolvement case the resulting pleural effusion is a chylothorax.
I n d i rect Result In fact, most chylothoraces that are not traumatic
Hypo p rote i n e m i a
in origin are secondary to neoplastic involvement of
Posto bstru ctive p n e u m o n itis
the thoracic duct. Lymphomas are responsible for
P u l m o n a ry e m bo l i s m
Postrad iation thera py
75% of chylothoraces secondary to malignant disease
(see Chapter 26) .
C H APTE R 1 0 I PLE U RAL E F F U S I O N S R E LAT E D TO M ETASTATIC MALI G N A N C I E S 1 57
Another mechanism by which malignant tumors 24 cases (53). Parietal pleural metastases result from
produce pleural effusion is through bronchial direct extension from the visceral pleura (43,53) .
obstruction. When a neoplasm obstructs the main In patients with nonbronchogenic carcinoma, the
stem bronchus or a lobar bronchus, the lung distal visceral pleura is also almost always involved. Involve
to the obstruction becomes atelectatic. Therefore, the ment of the parietal pleura again appears to result
remaining lung must overexpand or the ipsilateral from direct extension from the visceral pleura (43,53) .
hemithorax must contract to compensate for the loss The origin of these metastases is controversial. Meyer
of volume of the atelectatic lung. These events result attributed them to tertiary spread from secondary
in a more negative pleural pressure, and it is easy to hepatic tumors (43) . In his series of 23 patients with
see from Figure 2 . 1 that such a negative pleural pres pleural metastases, 1 9 (83%) had hepatic metastases.
sure causes pleural fluid to accumulate. My associ In the series by Rodriguez-Panadero et al., however,
ates and I studied a patient with obstruction of the hepatic metastases could be demonstrated only in
bronchus intermedius in whom the pleural pressure 7 1 % of the patients and they attributed the visceral
dropped from - 1 2 to -48 cm H 2 0 as 200 mL pleural pleural metastases to blood-borne metastases from
fluid was removed (52) . the primary (53) . The latter explanation appears more
Pericardia! involvement is frequent with meta plausible to me. The presence of pleural metastases
static malignant diseases. When a pericardia! effusion does indicate systemic dissemination of the disease
is caused by such involvement and hydrostatic pres and renders the patient incurable with surgery alone.
sures become elevated in the systemic and pulmonary Not all patients with pleural metastases have pleu
circulation, transudative pleural effusions may result. ral effusions. In Meyer's series, only 60% of patients
It is also likely that some of the malignant pericardia! with pleural metastases had pleural effusion (43). In
fluid is cleared through the pleural space, which can Rodriguez-Panadero's study, only 30 of 5 5 patients
lead to an exudative effusion (see Chapter 1 9) . (5 5%) with metastatic disease to the pleura had pleu
Not all pleural effusions in patients with malignant ral effusion (53) . Meyer found that the presence of
disease are related to intrathoracic involvement by the a pleural effusion was more closely related to neo
neoplasm. Pulmonary infection distal to a partially or plastic invasion of the mediastinal lymph nodes than
totally occluded bronchus may produce a parapneu to the extent of pleural involvement by nodular
monic effusion (see Chapter 1 2) . The incidence of metastases (43) .
pulmonary embolization is higher in patients with
malignant disease, and emboli frequently cause exuda
C L I N I CA L M A N I F E STAT I O N S
tive pleural effusions (see Chapter 1 7). Patients with
intrathoracic neoplasms frequently receive radiother The most common symptom reported by patients
apy for their tumors, and this treatment can also result with malignant pleural effusions is dyspnea, which
in pleural effusions (see Chapter 23), as can some types occurs in more than 50% (7) . Symptoms attributable
of chemotherapy (see Chapter 22) . Many patients with to the tumor itself are also frequent. In one series,
malignant disease are malnourished and have hypo weight loss occurred in 32%, malaise in 2 1 %, and
proteinemia, and this disorder can, on rare occasions, anorexia in 1 4% of patients (7) . When patients with
lead to the formation of transudative pleural effusions malignant pleural effusions are compared to those
(see Chapter 9). with benign pleural effusions, patients with malig
nant pleural effusions are more likely to have dull
A U TO P SY STU D I E S
chest pain (34% vs. 1 1 %), whereas patients with
benign disease are more likely to have pleuritic chest
The most detailed autopsy series o n pleural involve pain (5 1 % vs. 24%) (54) . Temperature elevations are
ment in malignant disease are those of Meyer (43) significantly more common in patients with benign
and Rodriguez-Panadero et al. (53) . It appears that disease (73%) than in patients with malignant disease
pleural metastases with bronchogenic carcinoma (37%) (54) .
are usually due to pulmonary arterial emboli to the
ipsilateral pleura. Virtually, all patients with meta
Chest Radiographs
static pleural involvement from lung carcinoma have
involvement of the visceral pleura (43,53) . In the The size of a malignant pleural effusion varies from a
series of Rodriguez-Panadero et al., pulmonary vas few milliliters to several liters, with the fluid occupying
cular invasion by the tumor was found in 1 9 of the the entire hemithorax and shifting the mediastinum
1 58 PLE U RAL D I S EASES
to the contralateral side. Malignant disease is the 43%; chest wall involvement, 1 2%; lymphangitic car
most common cause of a massive pleural effusion and cinoma, 7%; and suspicious lung masses, nodules, or
accounted for 3 1 of 46 ( 67%) of effusions occupying infiltrates, 53%.
an entire hemithorax, in one series (55) . In a more
recent series, malignancy was responsible for 5 5 % of
Pleural Fluid
1 63 effusions that occupied more than two thirds of
the hemithorax (56) . The pleural fluid from a malignant pleural effusion
Almost all patients with pleural effusions second is almost always an exudate (62,63) . Ashchi et al.
ary to bronchogenic carcinoma have radiographically (64) reviewed the medical records of 1 7 1 patients
demonstrable pulmonary abnormalities besides the with malignant pleural effusion and found that in
effusion. At times, a therapeutic thoracentesis must 8 of the cases, the fluid was transudative. There were
be performed before the pulmonary abnormality is alternative explanations for the transudative effusions
evident. Although almost all patients with pleural in seven of the eight cases (64) . In another study, 97
effusions secondary to lymphoma have mediastinal of 98 patients with malignant pleural effusions had
lymph node involvement at autopsy, this involve exudates (63) . The ratio of the pleural fluid to the
ment is not always evident in chest radiographs (3 1 ) . serum protein level is less than 0.5 in approximately
I n a series o f 2 2 patients with chylothorax due to 20% of malignant pleural effusions (7,62) . However,
lymphoma, only 5 patients (23%) had hilar or medi in these 20%, the ratio of the pleural fluid to the
astinal adenopathy demonstrable on routine chest serum lactate dehydrogenase (LDH) or the absolute
radiographs (57) . In another series, however, 7 1 % pleural fluid LDH almost always meet exudative cri
of 2 1 patients with pleural effusions secondary to teria (62) . Most pleural effusions that meet exudative
lymphoma had visible mediastinal lymph node criteria by the LDH level but not by the protein level
involvement on chest radiographs (58) . In a third are malignant pleural effusions (62) .
series of 1 9 patients with non-Hodgkin's lymphoma, The presence of grossly bloody pleural fluid (red
only 4 patients had mediastinal lymphadenopa blood cell [RBC] count > 1 00,000/mm3) suggests
thy (59) . The chest radiographs of patients with pleu malignant pleural disease. In our series of 22 such
ral effusions due to malignant tumors other than lung effusions, 1 2 (5 5%) were due to malignant dis
carcinoma or lymphoma often reveal only a pleural ease ( 1 ) . Approximately 30% to 50% of malignant
effusion. In a series of 1 05 patients with pleural pleural effusions, however, have red blood cell counts
effusion due to breast carcinoma (26) , only 9% had less than 1 0,000/mm3 and do not appear bloody
radiographically evident pulmonary metastases. ( 1 ,65) . The pleural fluid white blood cell (WBC)
In patients with undiagnosed pleural effusions, the count with malignant pleural effusion is variable,
chest CT scan is useful in indicating whether the effu with the usual count between 1 ,000 and 1 0,000/mm3
sion has a benign or malignant etiology. Yilmaz et al. ( 1 ) . The predominant cells in the pleural fluid differ
(60) reviewed the CT scans in 1 46 patients with pleu ential white cell count of these effusions are lympho
ral effusions including 59 that were malignant and 87 cytes in approximately 45%, other mononuclear cells
that were benign. They reported that the following in approximately 40%, and polymorphonuclear leu
four findings were suggestive of malignancy: (a) pleu kocytes in approximately 1 5 % ( 1 ) . In the past, it was
ral nodularity, (b) pleural rind, (c) mediastinal pleural stated that pleural fluid eosinophilia (> 1 0%) made
involvement, and (d) pleural thickening greater than pleural malignancy unlikely. However, in a review of
1 cm (60) . In a second study, pleural surfaces assessed 392 cases of eosinophilic pleural effusions that were
by CT scan were abnormal in 27 of 32 patients with not associated with pleural air and/or blood, malig
malignancy but in none of the 8 patients with benign nancy was the final diagnosis in 1 7% (66) . A more
disease (6 1 ) . It should be noted that both the series recent paper from a single institution reported that
mentioned in the preceding text had a large per malignancy was responsible for 34.8% of 1 3 5 eosino
centage of mesotheliomas, which are more likely to philic pleural effusions (67) .
have abnormalities of the pleural surfaces. Concur The pleural fluid glucose level is reduced to below
rent abnormalities are frequently present in patients 60 mg/dL in approximately 1 5 % to 20% of malig
with documented malignant effusions. In one study nant pleural effusions (68-70) . A low pleural fluid
of 86 patients, the following incidences of concurrent glucose level in association with a malignant pleural
abnormalities were reported: pericardial effusion, 3%; effusion indicates that the patient has a high tumor
pericardial thickening, 1 4%; mediastinal adenopathy, burden in the pleural space. Rodriguez-Panadero
C H APTE R 1 0 I PLE U RAL E F F U S I O N S R E LAT E D TO M ETASTATIC MALI G N A N C I E S 1 59
and Lopez-Mejias (70) performed thoracoscopy on which cytologic study of the pleural fluid establishes
77 patients with a malignant pleural effusion and the diagnosis of a malignant pleural effusion ranges
found that the extent of the tumor was significantly from 40% to 87% (77-8 1 ) . The reasons for this vari
greater in those with a low pleural fluid glucose level. ability in the diagnostic yield with cytologic study are
Cytology and pleural biopsy are more likely to be discussed in Chapter 7. When three separate pleural
positive in patients with low-glucose pleural effu fluid specimens from a patient with malignant pleu
sions (70) . Because of the large tumor burden, patients ral disease are submitted to an experienced cytologist,
with a low pleural fluid glucose level have a worse one should expect a positive diagnosis in approxi
prognosis (7 1 ) . It appears that the low-glucose levels mately 80% of patients. The incidence of positive
with malignant pleural effusion are due to impaired results depends on the primary tumor. Most cases of
glucose transfer from blood to pleural fluid (72) . metastatic adenocarcinoma can be diagnosed by pleu
Increased glucose utilization by the pleural tumor ral fluid cytology. Positive results are uncommon with
also probably plays a role in producing the low pleural squamous cell carcinoma (SCC) because the pleural
fluid glucose. effusions are usually due to bronchial obstruction or
Approximately one third of patients with malig lymphatic blockade ( 1 ,3,82) . With lymphoma, the
nant pleural effusions have a pleural fluid pH below cytologic test is positive in approximately 25% of
7.3 (70,73,74) . Patients with a low pleural fluid pH patients with Hodgkin's disease and in 50% to 60%
level also tend to have a low pleural glucose level of patients with non-Hodgkin's lymphoma. By cyto
(70,74) . As one might anticipate, they have a greater logic examination, the neoplasm can usually be clas
tumor burden, are more likely to have positive pleural sified into a histologic type such as adenocarcinoma,
fluid cytology and pleural biopsy, and have a shorter but the primary site of the tumor cannot usually be
survival than individuals with malignant pleural effu identified (82) .
sions and a pH level above 7.3 (70,74) . The patho
genesis of the low pH level with malignant pleural lmmunohistochemical Tests
effusions appears to be due to the combination of
acid production by the pleural fluid or the pleura and The use of monoclonal antibodies directed against
a block to the movement of carbon dioxide out of the various antigens is useful in distinguishing malignant
pleural space (72) . from benign pleural effusions (83) . Metastatic adeno
Approximately 1 0 % of patients with malignant carcinomas tend to stain positive with carcinoem
pleural effusions have an elevated pleural fluid amy bryonic antigen (CEA) , MOC-3 1 , B72.3, Ber-EP4,
lase level (69) . Usually, the primary tumor is not in BG-8, and TTF- 1 (84) , whereas malignant mesothe
the pancreas in these patients (69,75) . Analysis of the lial cells and benign mesothelial cells stain positive
amylase isoenzymes has demonstrated that the amy with calretinin, keratin5/6, podoplanin, and WT l
lase in malignant effusions is the salivary isoenzyme (83) (see Chapter 7) . TTF-1 has high specificity for
rather than the pancreatic isoenzyme (76) , and there lung carcinoma (83) . When attempts are made to dif
fore amylase isoenzyme analysis can be used to differ ferentiate metastatic carcinoma from mesothelioma,
entiate pancreatic effusions from malignant effusions. stains should be made with two of the antibodies that
stain positive for adenocarcinomas and two that stain
positive for mesothelioma (84) . In the situation where
D I AG N O S I S
the cytology of the pleural fluid is positive, but the
Th e diagnosis o f a malignant pleural effusion is estab primary is unknown, immunohistochemical tests can
lished by demonstrating malignant cells in the pleu help identify the site of the primary (see Chapter 7) .
ral fluid or in the pleura itself In most cases, this is The technology with these monoclonal antibodies
done by cytologic examination of the pleural fluid or continues to improve, and it is recommended that
biopsy of the pleura. all laboratories that deal with significant numbers of
pleural fluids develop the capability to perform these
Cytologic Examination
immunohistochemical tests.
in establishing the diagnosis of malignant pleural biopsy in establishing the diagnosis of pleural malig
effusions. Tumor markers evaluated have included nancy. Pleural biopsy has a lower diagnostic yield
CEA; carbohydrate antigens (CA) 1 5-3, 1 9-9, 549 than pleural Buid cytologic examination because, in
and 72-4 (85); neuron-specific enolase (85); s e c approximately 50% of patients with malignant pleural
antigen; cytokeratin 19 fragments (CYFRA2 1 - l ) disease, the costal parietal pleura is not involved (93) .
(85,86) ; and sialyl state-specific mouse embryonic In one large series of 28 1 patients with malignant
antigen (SSEA- 1 ) (87) . Although the mean levels pleural disease, the cytology was positive for malig
of tumor markers in the pleural Buid are signifi nant cells in 1 62 patients (58%), whereas the pleural
cantly higher in malignant effusions than in benign biopsy was positive in 1 23 (44%) (80) . The diagnosis
effusions, there is almost always some overlap (88) . of malignancy was established by pleural biopsy alone
For tumor markers to be useful in the diagnosis of in only 20 of the 28 1 patients (7. 1 %) (80) . Another
malignant effusions, they have to be 1 00% specific. way to look at this series is to consider the 1 1 8 patients
One does not want to misdiagnose a benign pleural with malignancy but negative cytology. Pleural biopsy
effusion as a malignant effusion and inform the established the diagnosis of malignancy in only 20 of
patient that they only have several months to live. these 1 1 8 patients ( 1 7%) .
In view of the reasons mentioned in the preceding What is the role of blind needle biopsy of the
text, the use of tumor markers for the diagnosis of pleura in establishing the diagnosis of malignant pleu
pleural malignancy is not recommended (89) . One ral effusion in the twenty-first century? Because thora
possible use of pleural Buid tumor markers is to use coscopy is very effective at establishing this diagnosis
high levels to select patients for more invasive studies and because the needle biopsy is diagnostic in less
(88,89). However, other criteria are probably better at than 20% of patients with malignancy and negative
predicting that a patient has malignancy. Ferrer et al. cytology, I rarely perform needle biopsy of the pleura.
(65) found that the following four characteristics Certainly, if the cytology is negative and thoracoscopy
were predictive of malignancy in 93 patients under is unavailable or an outpatient procedure is desired,
going thoracoscopy: a symptomatic period of more consideration can be given to performing needle
than 1 month, the absence of fever, the presence of biopsy of the pleura.
serosanguineous pleural Buid, and a chest CT scan An alternative means by which specimens of the
suggestive of pleural malignancy. All 30 patients pleural biopsy can be obtained is with CT-guided
who had all four criteria had malignancy, whereas all cutting-needle biopsy. The CT-guided biopsy is made
20 patients with one or no criteria had a nonmalig where the pleura is the most thickened. In one study,
nant disease (65 ) . 50 patients with unilateral pleural effusion and a
clinical suspicion of malignancy were randomized to
Other Noninvasive Tests
pleural biopsy through an Abram's needle biopsy or a
CT-guided cutting-needle biopsy (94) . The Abram's
Other noninvasive tests for the diagnosis of malig needle biopsy was not necessarily aimed where the
nant pleural effusions are discussed in Chapter 7. pleural thickening was greatest. The diagnosis of
These include measurement of oncogenes, hyaluronic malignancy was made in 13 of l 5 patients (87%) with
acid, lectin binding, chromosomal analysis, Buores the CT-guided biopsy but in only 8 of 1 7 patients
cence in situ hybridization (FISH) , Bow cytometry (47%) with the Abram's needle (94) . Although these
and proteomics. In general, the only test that is rec results look very promising, it should be noted that
ommended is Bow cytometry when a pleural lym 20 of the 27 patients with malignancy in this study
phoma is suspected. Demonstration of the clonality had mesothelioma, which is notoriously difficult to
of the lymphocytes in the pleural Buid establishes the diagnose with Abram's needle biopsy of the pleura.
diagnosis of lymphoma (90) .
O bservation, Thoracoscopy or an
Pleural Biopsy Open Thoracotomy
Needle biopsy of the pleura can establish the diagno In many patients with exudative pleural effusions, no
sis of a malignant pleural effusion. The percentage of diagnosis is apparent after the initial diagnostic thora
positive pleural biopsies in patients with malignant centesis including pleural Buid cytology and a pleural
pleural disease ranges from 39% to 75% (80,9 1 ,92) . Buid marker for tuberculosis. In such instances, a
In general, pleural Buid cytology is superior to pleural CT scan of the chest is recommended. If pulmonary
C H APTE R 1 0 I PLE U RAL E F F U S I O N S R E LAT E D TO M ETASTATIC MALI G N A N C I E S 161
emboli are demonstrated with the CT scan, the eti As discussed early i n this chapter and i n Chapter 7,
ology of the effusion is established. If parenchymal immunohistochemical tests are useful in making this
abnormalities are present, a bronchoscopy should be differentiation. In addition, histochemical tests using
performed. If the CT scan suggests mesothelioma, periodic acid-Schiff stain (PAS) or Alcian blue and
then a CT-guided needle biopsy or thoracoscopy electron microscopy are very useful in differentiat
should be performed to establish this diagnosis. ing metastatic adenocarcinoma from mesothelioma
If the CT scan shows nothing other than the pleu (see Chapter 1 1 ) .
ral effusion, the approach to the patient should be
governed by the clinical picture. If there is nothing
Lipid Analysis
in the patient's history to suggest carcinoma and if
the patient's symptoms are improving, then it is pro The possibility of a chylothorax should be considered in
bably best to observe the patient for several weeks every patient with malignant disease and a pleural effu
because only a small percentage of these patients sion. If a chylothorax is present, the mediastinal lymph
have malignant pleural disease (95). Alternatively, nodes are probably involved, and the treatment of
if the symptoms of the patient are worsening or if choice is radiation to the mediastinum or chemother
there is something in the clinical picture that sug apy. The supernatant of the pleural fluid from patients
gests malignancy, the patient should undergo thora with malignant pleural effusions should be examined.
coscopy. Ferrer et al. (65) reported that the following If the supernatant is turbid, a chylothorax should be
four characteristics were predictive of malignancy suspected, and the triglyceride level in the pleural fluid
in 93 patients undergoing thoracentesis including should be determined. If the pleural fluid triglyceride
50 with malignancy: (a) duration of symptoms more level exceeds 1 1 0 mg/dL, the patient probably has a
than 1 month, (b) absence of fever, (c) blood-tinged chylothorax; if the level is below 50 mg/dL, the patient
pleural fluid, and (d) chest radiograph suggestive of does not have a chylothorax (99) . If the level is between
malignancy (pulmonary or pleural masses, pulmo 50 and 1 1 0 mg/dL, lipoprotein electrophoresis should
nary atelectasis, or lymphadenopathy) . Twenty-eight be performed (99) (see Chapter 26) .
patients had all four criteria and all had malignancy.
Twenty-one patients had at most one criterion and
Other Diagnostic Tests
none had malignancy (65) . Thoracoscopy will estab
lish the diagnosis of malignancy in approximately Numerous articles have recommended various diag
90% of patients with malignancy (96,97) . If facili nostic tests such as Row cytometry, chromosomal
ties for thoracoscopy are not available, an alternative analysis of pleural fluid cells, or LDH isoenzymes in
approach is to perform a thoracotomy with open the diagnosis of malignant pleural effusions. These
biopsy of the pleura or to perform a needle biopsy various tests are discussed in Chapter 7. In general,
of the pleura. If thoracoscopy or thoracotomy is per they are not recommended. Flow cytometry with
formed, a procedure such as a pleural abrasion should immunophenotyping is useful in making the diagno
be performed to prevent recurrence of the pleural sis of lymphoma (59) .
effusion.
Unknown Primary
Mesothelioma
Most patients who are diagnosed with a malignant
The possibility of a malignant mesothelioma (see pleural effusion are already known to have a malig
Chapter 1 1 ) should be considered whenever a patient's nancy. However, if the patient presents with a malig
pleural fluid cytologic study or pleural biopsy suggests nant pleural effusion, where is its likely origin? One
metastatic adenocarcinoma, because the epithelial study ( 1 00) reviewed 42 consecutive patients referred
form of malignant mesothelioma is frequently mis to the Royal Marsden Hospital with a malignant
interpreted as adenocarcinoma on cytologic examina pleural effusion and no known primary. The patients
tion or pleural biopsy (98) . If no primary tumor is included 27 men and 1 5 women. Despite CT scans
evident, a CT scan of the thorax should be obtained. of chest and abdomen and mammography and pelvic
If the CT scan suggests mesothelioma, one should ultrasound in 1 0 patients, a primary was determined
consider thoracoscopy or exploratory thoracotomy in only 1 5 ( I O men and 5 women) , and the primary
for staging and possible radical pleuropneumonec was in the lung in all cases ( 1 00). The median survival
tomy (see Chapter 1 1 ) . in this group was 1 2 months from diagnosis ( 1 00) .
1 62 PLE U RAL D I S EASES
In view of the findings mentioned in the pre number of pleural adhesions at thoracoscopy ( 1 03)
ceding text, it is recommended that patients with a and the higher the VEGF level in pleural effusions
malignant pleural effusion and an unknown primary due to lung cancer ( 1 04) , the poorer the prognosis.
tumor have a CT scan of the chest, abdomen, and All of the poor prognostic factors mentioned in the
pelvis. If pulmonary parenchymal abnormalities are preceding text probably reflect a greater tumor bur
discovered, then a bronchoscopy is indicated with den in the pleural space (7 1 , 1 02) .
special attention to the area of abnormality. If there
are no parenchymal abnormalities, then bronchos T R E AT M E N T
copy will probably be nondiagnostic ( 1 00) . Masses
in the abdomen should be evaluated. If the patient The presence of a malignant pleural effusion indi
has symptoms referable to a specific organ, that organ cates that the malignancy cannot be cured by surgery.
should be evaluated. If the patient is a woman, mam Management of the effusion centers on palliation of
mography and a careful pelvic examination should be symptoms because no available treatments prolong
performed. If the foregoing sequence of tests does not survival (Fig. 1 0 . 1 ) .
identify the site of the primary tumor, it is recom Th e initial step i s t o identify the location o f the
mended that further tests not be undertaken ( 1 00) . primary lesion. Frequently, the location of the pri
mary is already known when the pleural effusion is
first identified. If the primary is unknown, then the
PROG N O S I S
procedures outlined in the preceding paragraph on
Th e prognosis o f patients with malignant pleural effu unknown primary should be followed.
sions is not good. In a one report, the median survival The main reason to identify the primary tumor
of 4 1 7 patients with malignant pleural effusions was is to decide whether systemic chemotherapy is indi
only 4 months (7 1 ) . This 4-month median survival cated. Systemic chemotherapy is effective at least in
is on the optimistic side because all 4 1 7 patients were some patients with small-cell carcinoma, breast car
judged to be fit enough to undergo pleurodesis (7 1 ) . cinoma non-small-cell lung cancer and lymphoma.
Th e most important factor influencing the life expec If the patient has a chylothorax, radiation should be
tancy in patients with malignant pleural effusion is administered to the mediastinum.
the source of the tumor. In the study mentioned in the If the patient has a tumor that is not responsive to
preceding text, the median survivals were 3 months chemotherapy or fails to respond to chemotherapy, a
for 1 46 patients with lung cancer, 2.3 months for 1 8 procedure should be considered to remove the pleural
patients with gastrointestinal primaries, 5 months for fluid. The two primary modes of treatment to con
60 patients with breast carcinoma and 5 1 patients trol the accumulation of pleural fluid are the inser
with unknown primary, and 6 months for 29 patients tion of an indwelling pleural catheter or the creation
with mesothelioma (7 1 ) . In a more recent report of a pleurodesis. Only patients who are dyspneic and
( 1 0 1 ) of 284 patients with malignant pleural effu whose dyspnea improves after a therapeutic thora
sion, the overall median survival was 5 .4 months fol centesis should be considered candidates for fluid
lowing diagnosis. Again survival varied significantly removal. If the patient is not symptomatic from the
depending on the primary tuber being 1 7.4 months pleural effusion, no treatment is recommended. Most
for mesothelioma, 1 3.2 months for breast cancer, patients who have small effusions that do not produce
7 months for lymphoma and 2.6 months for lung symptoms never become symptomatic ( 1 05). The
cancer ( 1 0 1 ) . A second factor that is very impor dyspnea of patients who do not improve after a thera
tant in determining the prognosis of patients is their peutic thoracentesis should be treated with opiates
Karnofsky Performance Scale (KPS) score. Burrows and/or oxygen. Before a pleurodesis is attempted, the
et al. ( 1 02) reported that the median survival of position of the mediastinum on the chest radiograph
patients with a KPS score less than 30 was 34 days, should be noted. If the mediastinum is shifted toward
whereas the median survival of patients with a KPS the side of the effusion, a bronchoscopy should be
score greater than 70 was 395 days. done before pleurodesis is attempted because it is
Other factors associated with a poor prognosis are likely that the patient has an obstructed bronchus.
a pleural fluid pH level below 7.20, a pleural fluid The presence of an obstructed bronchus is a contrain
glucose level below 60 mg/dL, or a pleural fluid LDH dication to pleurodesis.
level more than twice the upper limit of normal for If the patient is dyspneic and if the dyspnea is
serum (7 1 , 1 0 1 , 1 02) . In addition, the greater the relieved by a therapeutic thoracentesis, the procedure
C H APTE R 1 0 I PLE U RAL E F F U S I O N S R E LAT E D TO M ETASTATIC MALI G N A N C I E S 1 63
�
Is it responsive to chemotherapy? Chemotherapy
Cell t 1e "'-- Ji' \t and/or rdiothe rapy
'-:4. No Yes
Small cell Non-small Response
t
.J, / \
cell
r
Adm i n iste r c emotherapy
Che � othe rapy No Yes
� Response .J, t
Response t .j, Chylothorax Conti nue
.j, t Yes No .j, .j, chemothe rapy
Yes No � \ t �
.J, � Cont i n u e chemother P ! J> No Yes
.J,
Conti n u e Consider i ndwe l l i n g I Med iastinal
chemotherapy catheter or p l e u rodesis radiation
t
Is the patient symptomatic
from the effusion?
.j, .j,
No Yes
J. .j,
Observe Is the mediasti n u m sh ifted
toward the ipsilateral side?
/1
Ji' t
J/Yes �o
B ronchoscopy Therapeutic
/ 1
th racentesis
I
Yes No
.J, Treat symptomatically
Laser therapy or rad iotherapy with opiates/oxygen
.
to obstruction Outpatient treatment
Ji' """
Yes No
t .j,
P l e u rX Chest tube
.j,
Lung expands
Ji' """
No Yes
,j, J.
P l e u roperitoneal P l e u rodesis
shunt
of choice for outpatients who can receive home health be treated with either an indwelling pleural catheter
care or who have strong family support for control or a pleuroperitoneal shunt.
ling the pleural fluid is the insertion of an indwelling
catheter with periodic drainage of the pleural fluid
Systemic Chemotherapy
through vacuum bottles. If patients do not meet these
criteria, pleurodesis with a sclerosing agent through a The presence of a malignant pleural effusion usually
chest tube is the procedure of choice. If the lung does indicates disseminated tumor, at least for nonbron
not expand after the chest tube is inserted, pleurode chogenic carcinoma (43) . Therefore, the only hope
sis should not be attempted ( 1 06) . These patients can for cure or prolonged palliation is with systemic
1 64 PLE U RAL D I S EASES
chemotherapy. Fentiman et al. (26) reported that pleu malignant pleural effusion indicates that the tumor is
ral effusions were controlled in 7 of 22 patients (32%) disseminated and that local chemotherapy is unlikely
with metastatic breast carcinoma who were given to be useful. However, if intrapleural chemotherapy
systemic chemotherapy, whereas Jones et al. ( 1 07) decreased the number of tumor cells in the pleu
reported positive responses in 6 of 8 patients (75%) ral space, the rate of pleural Buid formation might
given systemic chemotherapy. There are no recent decrease.
reports to my knowledge concerning the outcomes of In one recent study, excellent results were obtained
malignant pleural effusions with breast cancer treated in 27 patients with NSCLC who received a combi
with systemic chemotherapy. Livingston et al. ( 1 08) nation of radiotherapy, intrapleural chemotherapy,
reported that 19 of 53 patients (36%) with small-cell and systemic chemotherapy ( 1 1 5) . The regimen con
lung carcinoma had complete disappearance of their sisted of cisplatin 60 mg/m2 intrapleurally on day 1 ;
pleural effusions with chemotherapy. The median sur gemcitabine 1 ,000 mg/m2 intravenously on days 1 ,
vival of patients with small-cell lung carcinoma with 8 , and 1 5 and every 4 weeks fo r three times. This was
limited disease and a pleural effusion is 1 3.9 months, followed by radiotherapy (7,020 cGy/39 fr) and then
compared with a median survival of 1 8.3 months if no postradiation chemotherapy (docetaxel 60 mg/m 2
pleural effusion is present ( 1 09) . In a second study of every 3 weeks for three to six times) . Only two
62 patients with small-cell lung carcinoma, the pleural patients experienced recurrence of the pleural effu
effusion disappeared after first-line chemotherapy in sion ( 1 1 5) . The overall response rate was 5 5 % with
34 patients (55%) (8) . In addition, pleural effusions in 7% complete and 48% partial remissions ( 1 1 5) . The
lymphomas frequently respond to chemotherapy (35). median overall survival was 1 8 months, which is
There is one report in which the administration approximately six times longer than one expects with
of cisplatin, ifosfamide, and irinotecan to 34 patients malignant pleural effusions secondary to NSCLC. It
with NSCLC and pleural effusion resulted in the is unclear how much one cycle of intrapleural chemo
complete disappearance of the effusion in 1 3 patients therapy contributed to the overall results.
(38%) , a partial resolution in 7 patients (2 1 %), and a In a second study, the therapeutic efficacy of
mean survival of 362 days. If these results can be con Staphylococcus aureus superantigen (SSAg) , a power
firmed, this regimen may prove useful in treating and ful T-cell stimulant, was evaluated in patients with
extending the life of patients with adenocarcinomas malignant pleural effusion from NSCLC ( 1 1 6) . In a
of the lung. In recent years, it has been shown that small study of 1 4 patients with a median pretreat
the anti-VEGF antibody bevacizumab, combined ment KPS score of 40 received the pleural instillation
with standard first-line chemotherapy, provides a sta of SSAg, 1 00 to 400 mg, once or twice weekly until
tistically and clinically significant survival advantage the pleural effusions resolved ( 1 1 6) . The effusions
in patients with NSCLC ( 1 1 0) . It is important not were completely controlled in 1 1 patients and the
to attempt pleurodesis in patients who are receiving median survival was 7 months ( 1 1 6) . If these results
an anti-VEGF drug because angiogenesis is necessary are confirmed in additional studies, this regimen is
for pleurodesis and angiogenesis is inhibited by anti attractive because it not only controls the effusion,
VEGF drugs ( 1 1 1 ) . but also increases survival.
I n patients undergoing systemic chemotherapy There is one case report in which the intrapleu
with methotrexate, pleural effusions should be aspi ral administration of increasing doses of rituximab, a
rated before chemotherapy is given because the anti monoclonal antibody directed against the CD20 anti
neoplastic drugs may accumulate in the pleural space gen on the surface of B lymphocytes, was effective in
and lead to increased systemic toxicity ( 1 1 2, 1 1 3) . controlling a pleural effusion in a patient with a CD 20+
However, i f the patients are receiving chemotherapy non-Hodgkin's lymphoma (1 1 7). Interferon-gamma
with pemetrexed, there is no reason to aspirate the ( 1 1 8), tumor necrosis factor ( 1 1 9), interleukin-2 ( 1 1 9),
pleural Buid ( 1 1 4) . cisplatin ( 1 20), and the monoclonal antibody catumax
omab ( 1 2 1 ) have all been tried in small numbers of
lntrapleural Chemotherapy
patients with results that are not particularly impressive.
by the neoplastic process. Therefore, it is logical to In the initial multicenter study using this catheter
administer radiotherapy to the mediastinum in such to treat malignant pleural effusions, the efficacy and
patients who have tumor rypes that are resistant to safery of the catheter were compared with those of
primary chemotherapy. In one series, mediastinal doxycycline pleurodesis through tube thoracostomy.
radiation resulted in adequate control of the chylo All patients were initially hospitalized. The median
thorax for the remainder of the patients' lives in 68% hospitalization for the 94 patients who received
of those with lymphomas and in 50% of those with the PleurX catheter was 1 day, whereas the median
metastatic carcinoma ( 1 22) . Other means to manage hospitalization for the 4 1 patients who had doxycy
chylothorax are discussed in Chapter 26. cline pleurodesis was 6.5 days. Of the 94 patients who
received the PleurX catheter, the effusion was initially
Indwelling P leural Catheter (PleurX)
controlled by the catheter in 9 1 (97%) . The presence
of the catheter leads to a spontaneous pleurodesis in
The PleurX catheter (Fig. 1 0.2) or other indwelling approximately 50% of patients with a median time
catheters such as the Rocket (UK) allow patients to pleurodesis of 28 days ( 1 25). Patients who experi
with recurrent pleural effusions to have their pleural enced a pleurodesis tended to have a gradual decrease
fluid drained repeatedly without having to return to in the amount of pleural fluid formed daily (44) .
the hospital. The PleurX catheter is a 1 5 .5 F silicone Several of the patients had the catheter in place for
rubber catheter, 66 cm in length, with fenestrations more than 1 year.
along the proximal 24 cm (CareFusion Corporation, Since the original study, several studies have docu
San Diego, CA) . It is inserted into the pleural space mented the feasibiliry of implanting the PleurX cathe
using the Seldinger technique under local anesthesia ter as an outpatient. Putnam et al. ( 1 24) retrospectively
by pulmonologists, interventional radiologists, or compared their experience with 1 00 patients (includ
surgeons. The catheter is maintained in place with a ing 60 outpatients) who were treated with the PleurX
chest wall tunnel 5 to 8 cm in length ( 1 23, 1 24) . This catheter and 68 patients treated with tube thoracos
particular catheter has a special valve on its distal end tomy using either doxycycline or talc as a pleurodesing
that is designed to enhance the safery of the product. agent. The median survival was 3.4 months and did not
The valve prevents fluid or air from passing in either differ significantly between treatment groups ( 1 24) .
direction through the catheter unless the catheter is The early mean charges were approximately US$8,000
accessed with the matched drainage line. The pleural in the inpatient chest tube group and US$3,400 in
fluid is drained intermittently by inserting the access the outpatient pleural catheter group ( 1 26) . The cost
tip of the drainage line into the valve of the catheter differential is decreased over longer periods as the suc
and then draining the fluid through an external tube tion bottles are expensive. This cost can be markedly
into vacuum bottles. decreased if the pleural fluid is drained repeatedly into
an accordion drain ( 1 27) . Musani et al. ( 1 28) reported
their results with the outpatient insertion of the cath
eters in 24 patients. They reported that all patients
experienced symptomatic relief and spontaneous
pleurodesis occurred in 58%.
There have been two studies with at least 250
pleural catheter insertions ( 1 25, 1 29) . In the first large
series regarding the utiliry of the indwelling cath
eter, Tremblay and Michaud ( 1 25) retrospectively
reviewed their experience at a single center with 250
tunneled pleural catheter insertions in 223 patients
for malignant pleural effusion. Home nursing support
was organized for all patients to assist with catheter
care. They reported that the tunneled catheter inser
tion resulted in complete symptom control in 38.8%,
partial control in 50% and no control in 3.6%. In
addition, there were 4% failed insertions and the
symptoms were not evaluated in 3.6% ( 1 25). In most
FIGURE 1 0.2 • P l e u rX catheter.
patients, the catheter was inserted on an outpatient
1 66 PLE U RAL D I S EASES
When managing such a patient, the first question to an endobronchial stent should be considered for relief
answer is whether the patient is symptomatic from of the bronchial obstruction. If no obstructing lesion
the effusion. The only symptom likely to be relieved is found, the lung is probably encased by the tumor,
with pleurodesis is dyspnea. If the patient does not and pleurodesis is likely to fail. In this situation, if
have symptoms attributable to the pleural effusion, the patient has symptomatic relief after a therapeu
it does not make sense to insert a chest tube and to tic thoracentesis, insertion of a PleurX catheter or a
attempt a pleurodesis just to make the chest radio pleuroperitoneal shunt should be considered. In rare
graph look better. In a similar vein, if the patient is situations a pleurectomy (see the discussion on pleu
moribund from disseminated tumor, he or she should rectomy in this chapter) can be considered.
not be tortured for the last few days of his or her life If the mediastinum is in midline position or has
with chest tubes. If a patient's quality of life is dimin shifted to the contralateral side (Fig. 1 0.4), then a ther
ished by dyspnea and he or she has a life expectancy apeutic thoracentesis (see Chapter 28) should be per
of more than a few weeks, however, one should con formed. The purpose of this procedure is to determine
sider pleurodesis. This local therapy probably does whether it relieves the dyspnea of the patient. Only
not improve the duration of the patient's life, but it those patients who experience significant symptom
can improve the quality of life. atic improvement from this thoracentesis should be
Before a pleurodesis is attempted, the position considered to be candidates for chemical pleurodesis.
of the mediastinum on the chest radiograph should The exercise tolerance does not increase in a substan
be evaluated, because its position tells much about tial fraction of patients with a malignant pleural effu
the pleural pressure on the side of the effusion. If the sion after a therapeutic thoracentesis ( 1 35, 1 36) . In like
mediastinum is shifted toward the side of the effusion manner, there is no decrease in the level of dyspnea at a
(Fig. 1 0.3), the pleural pressure is more negative on given workload in many patients ( 1 35, 1 36) .
the side of the effusion. Pleurodesis is then unlikely to
be successful because the ipsilateral lung is unable to
Prognostic Factors for Successful
expand. In such patients, a bronchoscopic examina
Chemical Pleurodesis
tion should be performed to assess the patency of
the major bronchi. If neoplastic obstruction of the In the past, it has been shown that if the pleural fluid
bronchi is discovered, radiotherapy, laser therapy, or pH or glucose levels are reduced, pleurodesis is less
likely to be successful (74, 1 37) . In one large series in
which pleurodesis was attempted with talc insuffia
tion during thoracoscopy, pleurodesis failed in 6 of
14 patients (43%) if the pleural fluid pH level was
below 7.2 but only in 8 of 92 patients (9%) if the
pleural fluid pH level was above 7.2 ( 1 37) . Compa
rable findings have been reported when the pleurode
sis was attempted with intrapleural tetracycline (74) .
Similar results are reported when a glucose measure
ment of less than 60 mg/dL was used as a predictor of
pleurodesis failure. One review of 433 patients under
going pleurodesis found that the pleural fluid pH
level was the only independent predictor of pleurode
sis failure (7 1 ) . Interestingly, the receiver operator
curves (ROC) for the pleural fluid pH, pleural fluid
glucose, and pleural fluid LDH levels were virtually
superimposable (7 1 ) . Nevertheless, a low pleural fluid
pH level should not be taken as an absolute contra
indication to chemical pleurodesis because 40% of
patients with a pleural fluid pH below 7 will still have
FIGURE 1 0.3 • Poste roa nte rior chest r a d i o g r a p h a successful pleurodesis and 68% of the patients with
f r o m a p a t i e n t w i t h a m a l i g n a nt p l e u ra l effu s i o n . pleural fluid pH below 7.3 will also have a successful
N ote t h a t the m e d i a sti n u m is s h i fted towa rd the
pleurodesis (7 1 ) . Interestingly, in this large series,
s i d e of the effu s i o n .
there was no association between pleurodesis success
1 68 PLE U RAL D I S EASES
A B
FIGURE 1 0.4 • A: Poste roa nte r i o r chest ra d i o g ra p h from a patient with a m a l i g n a nt p l e u r a l effus i o n .
N ote t h at the m e d i a sti n u m is s h i fted away from the s i d e w i t h the p l e u ra l effu s i o n . B: Poste roante r i o r ch est
rad i o g r a p h from the s a m e patient after i nsert i o n of a ch est t u b e i nto the l eft p l e u ra l space. N ote that the
l eft lung i s not we l l expa n d e d . Accord i n g ly, a sclerosi ng agent should not be i nj ected i nto the pleural space.
and the pleurodesis technique (thoracoscopy vs. tube lung is unlikely to expand with the removal of the
thoracostomy) , pleurodesis agent (talc, bleomycin, or pleural fluid.
tetracycline derivative) , or tumor type (7 1 ) . The success rate with pleurodesis is also related
Th e success rate with pleurodesis i s related t o the to the site of the primary tumor. Bielsa et al. ( 1 39)
changes in pleural pressure during thoracentesis. Lan reviewed the outcome of pleurodesis attempts in 1 26
et al. ( 1 38) measured the change in pleural pressure patients who received doxycycline and 447 patient who
after 500 mL of pleural fluid had been withdrawn received talc poudrage. They reported that in the talc
from 65 patients with a malignant pleural effusion. group patients with lung cancer and mesothelioma had
They then inserted a chest tube and continued to drain a significantly lower complete response rate (63% and
the lung until (a) the drainage was less than 1 50 mL/ 6 1 %, respectively) than did patient with breast (77%)
day, (b) the drainage was less than 250 mL/day for or other metastatic effusions (74%) ( 1 39) . Similar con
4 consecutive days, or (c) the drainage had continued clusions could be made in the doxycycline group ( 1 39).
for 1 0 days. At this time, they attempted pleurodesis
if the lung had expanded. They found that 14 patients
Mechanisms for Pleurodesis
had a pleural elastance greater than 1 9 cm H 2 0, that
is, the pleural pressure decreased by more than 9 . 5 cm Originally, antineoplastic agents such as nitrogen mus
H 2 0 when the 500 mL of pleural fluid was withdrawn. tard ( 1 40) or radioisotopes ( 1 4 1 ) were injected into the
They found that 1 1 of the 1 4 patients had a lung that pleural space in the hope that these agents would kill the
had not expanded. Pleurodesis was attempted in the tumor cells and control the pleural effusion. It was sub
remaining three patients with bleomycin and failed sequently shown that the injection of these agents often
in all. In contrast, only 3 of the 5 1 patients with pleu controlled the pleural effusion when tumor cells per
ral elastance less than 1 9 cm H 2 0 had a trapped lung sisted and that the effectiveness of intrapleural therapy
and bleomycin pleurodesis was successful at 1 month was related more to the creation of a pleurodesis that
in 42 of 43 (98%) of the patients who returned for prevented the accumulation of the pleural fluid than
reevaluation. to any antineoplastic effect of the agent administered
The results of the study mentioned in the pre ( 1 42 ,1 43) . The effectiveness of intracavitary nitrogen
ceding text could have been anticipated because mustard is much greater when the instillation of this
pleurodesis will fail if the two pleural surfaces cannot agent is combined with tube thoracostomy (4), because
be brought into approximation. The rapid fall in the the apposition of the two pleural surfaces allows the
pleural pressure is an indication that the underlying fibrotic process to obliterate the pleural space.
C H APTE R 1 0 I PLE U RAL E F F U S I O N S R E LAT E D TO M ETASTATIC MALI G N A N C I E S 1 69
Subsequent to the demonstration of the impor various cytokines in inducing either fibrosis or repair
tance of the chemical pleuritis in controlling pleu remains to be determined. In the future, it is likely
ral effusions, nonspecific irritants such as talc ( 1 44) , that pleurodesis will be produced by the intrapleural
tetracycline derivatives ( 1 45), silver nitrate, iodopo injection of cytokines. One cytokine that is an excel
vidone, and quinacrine ( 1 46) were combined with lent candidate as an effective pleurodesis-producing
tube thoracostomy in an attempt to control malig agent is transforming growth factor-/3 (TGF-/3) .
nant effusions. The initial event in the production of TGF-/3 has several characteristics that would be
a pleurodesis by these agents without question is an important for a pleurodesis agent: (a) TGF-/3 is a
injury to the pleura. An acute exudative pleural effu potent fibrogenic cytokine that regulates extracellu
sion develops within 1 2 hours of the instillation of lar matrix production. In situations in which there
essentially all the agents that are currently used for is too much TGF-{3, fibrosis results ( 1 56). The tran
pleurodesis including talc ( 1 47) , tetracycline deriva sient overexpression of TGF-/3 in the rat lung leads
tives ( 1 48), quinacrine ( 1 49), mitoxantrone ( 1 50) , to marked pleural and interstitial fibrosis ( 1 57) . (b)
and bleomycin ( 1 49) . Th e pleural fluid that accumu Once present, TGF-/3 can induce its own transcrip
lates after the intrapleural injection of these agents is tion ( 1 58), which suggests that a single injection
initially characterized by relatively high protein and may be sufficient. (c) Mesothelial cells express and
LDH levels, and neutrophil counts ( 1 5 1 ) . However, secrete TGF-/3; therefore, one intrapleural injec
injury to the pleura, as evidenced by the production tion of TGF-/3 might result in prolonged secretion of
of an acute exudative pleural effusion, is not sufficient TGF-{3, which could result in pleurodesis. (d) The
to induce a pleurodesis because many agents, when incubation of human pleural mesothelial cells with
injected intrapleurally, produce an acute exudative TGF-/3 results in secretion of increased levels of plas
effusion but do not produce a pleurodesis ( 1 49). minogen activator inhibitor 1 (PAI- 1 ) ( 1 59). This
The response of the pleura to an injury is a complex could facilitate pleurodesis because inhibition of the
and incompletely understood multifactorial process fibrinolytic system is thought to be necessary for the
that can result in the development of fibrosis with the production of a pleurodesis ( 1 55).
obliteration of the pleural space, or it can result in res Our preliminary studies in both rabbits ( 1 5 1 ) and
toration of the pleura to its normal state. The mecha sheep ( 1 60, 1 6 1 ) demonstrate that the intrapleural
nisms of pleurodesis seem to differ from agent to agent. injection of small amounts ofTGF-/3 results in a bet
The histologic appearance is much different with mito ter pleurodesis than does the intrapleural injection of
xantrone ( 1 50) than it is with talc ( 1 47) or tetracycline either doxycycline or talc slurry. The pleurodesis after
derivatives ( 1 48) . The pleurodesis that follows talc, but TGF-/3 occurs faster than after talc ( 1 62) . Moreover,
not tetracycline, can be blocked if corticosteroids are the pleural fluid that results from the intrapleural
given systemically ( 1 52) or if tumor necrosis factor injection ofTGF-/3 is characterized by a much lower
alpha-blocking antibodies ( 1 53) are given intrapleu WBC count and LDH level than the fluid that results
rally immediately after talc is administered. from the intrapleural injection of doxycycline or talc
The balance between the procoagulant system slurry ( 1 5 1 ) . The pleurodesis following intrapleural
and the fibrinolytic system is also important in TGF-/3 is not inhibited by corticosteroids ( 1 63). We
determining whether a pleurodesis will result after believe that TGF-/3 produces a fibrotic reaction in
the intrapleural injection of a substance. If the pro the pleural space without the necessity for a pleural
coagulant system dominates, then pleurodesis will injury. If indeed this is the situation, TGF-/3 will be
result, whereas if the fibrinolytic system dominates, an ideal agent for pleurodesis. Presently, the effective
no pleurodesis will result. When rabbits are given tet ness and safety of TGF-/3 as an agent for pleurodesis
racycline intrapleurally, the number of pleural adhe in humans is awaiting clinical trials.
sions that occur is reduced if the rabbits are given
either heparin or urokinase intrapleurally ( 1 54) . In
Vascular Endothelial Growth Factor
humans, pleurodesis occurs after talc insuffiation only
if the intrapleural fibrinolytic activity decreases ( 1 5 5). When mesothelial cells are incubated with TGF-{3,
the mesothelial cells produce increased amounts of
VEGF ( 1 64) . In addition, the pleural fluid levels
Transforming Growth Factor f3
of VEGF are significantly correlated with the levels
Without a doubt, cytokines are involved in the ofTGF-/3 in patients with pleural effusions ( 1 65). We
production of a pleurodesis, but the importance of noted that when pleurodesis was induced in rabbits
1 70 PLE U RAL D I S EASES
by the intrapleural administration of TGF-/3, there rabbits (400 mg/kg) is much higher than that typi
was much more pleural fluid than when pleurodesis cally used in patients (- 1 00 mg/kg) ( 1 70) . Bresticker
was induced with doxycycline or talc slurry ( 1 1 0) . We et al. ( 1 7 1 ) reported that talc insuffiation ( 1 g) and
therefore hypothesized that the increased amounts mechanical abrasion were essentially equivalent in
of pleural fluid with TGF-/3 were due to increased producing pleurodesis in dogs, whereas Jerram et al.
amounts of VEGF. However, when we administered ( 1 72) reported that mechanical abrasion was superior
anti-VEGF antibodies to the rabbits, the amount to talc slurry in producing pleurodesis in dogs.
of pleural fluid was just slightly diminished, but no It appears that talc in a slurry is effective in pro
pleurodesis occurred ( 1 1 1 ) . When the pleural tissues ducing a pleurodesis in humans ( 1 72-1 76) . Dresler
were assessed for the amount of vasculature, there was et al. ( 1 76) randomized 482 patients to receive 4 to
a close correlation between the amount of vasculature 5 g of talc, either administered as a slurry in 1 00 mL
and the pleurodesis scores ( 1 1 1 ) . This study dem saline through a chest tube or insuffiated during tho
onstrated that angiogenesis is very important in the racoscopy. The results in this study were much poorer
production of a pleurodesis and pleurodesis should than those reported previously (Fig. 1 0.5). Of the 1 63
not be attempted in patients who are receiving anti patients that received talc slurry and whose lung was
VEGF regimens. more than 90% expanded, only 56% were alive with
out recurrence at 30 days. Moreover, 29% of those
alive had a recurrence by 30 days ( 1 7 6) . Burgers et al.
Choice of Sclerosing Agent
( 1 67) reported that 36% of 64 patients had a recur
Currently, the agents that are most commonly recom rence at a mean of 1 7 days. Other uncontrolled stud
mended are talc (either insuffiated or as a slurry) , the ies or studies with fewer patients have reported better
tetracycline derivatives (minocycline or doxycycline), results with talc slurry. Adler and Sayek ( 1 44) treated
the antineoplastic agents (bleomycin or mitoxantrone), 44 hemithoraces with malignant pleural effusion with
silver nitrate and iodopovidone. A brief discussion of 1 0 g of talc in 250 mL of saline and reported con
the various agents proposed as sclerosing agents follows. trol of the effusion in 41 (93%) . Kennedy et al. ( 1 75)
reported that 10 g of talc mixed in 1 50 to 250 mL of
saline effectively controlled 38 of 47 (8 1 %) malig
Talc
nant pleural effusions. Haddad et al. ( 1 77) adminis
The sclerosing agent that is most commonly used for tered 4 g of talc slurry to 37 patients with malignant
chemical pleurodesis in English speaking countries effusions and reported that the success rate at 60 days
is talc ( 1 66) . In a survey of 84 1 pulmonologists in was 85%.
English speaking countries, talc slurry was the agent Aydogmus et al. ( 1 78) analyzed the factors asso
of choice in 5 5 .6%, whereas insuffiated talc was the ciated with successful pleurodesis with talc slurry.
agent of choice in 1 2.2%. Talc is also the agent rec They found the success rate was significantly higher
ommended by the Dutch Society of Pulmonologists if the time period between the diagnosis of effu
( 1 67) and the British Thoracic Society ( 1 68) . The sion and administration of talc slurry was less than
reason that talc is used most commonly is that it is 30 days or spontaneous expansion was attained after
widely available, inexpensive, and is perceived to be chest tube drain of if the daily drainage was less than
effective. Talc can be instilled into the pleural space 200 ml before talc slurry ( 1 78). The success rate was
either as slurry (suspended in saline) or insuffiated not related to the source of the primary tumor or the
(as an aerosol) . Talc insuffiation is usually done at the amount of drainage when chest tube drainage was
time of thoracoscopy and this procedure is discussed terminated ( 1 78).
later in the section on Thoracoscopy for Pleurodesis. The primary concern with talc is that its intrapleu
Animal studies have demonstrated that the intra ral administration has been associated with the devel
pleural administration of talc slurry can produce a opment of the acute respiratory distress syndrome
pleurodesis if the pleura is normal (see Chapter 4) . (ARDS) . The incidence of ARDS following intrapleu
However, there is no convincing evidence that talc ral talc has varied markedly from series to series, and
is superior to other agents. In our rabbit model, talc most of the reported cases have been from the United
slurry at a dose of 400 mg/kg produces a pleurodesis States. The highest incidence was that reported by
( 1 69), but doxycycline, 1 0 mg/kg, is at least as effec Rehse et al. ( 1 79), who retrospectively reviewed
tive ( 1 5 1 ) . Moreover, it should be noted that the dose their experience of 89 talc pleurodesis procedures
of talc slurry necessary to produce a pleurodesis in in 78 patients after 1 patient developed fulminant
C H APTE R 1 0 I PLE U RAL E F F U S I O N S R E LAT E D TO M ETASTATIC MALI G N A N C I E S 171
1 .0
0.8
Cl
c
·;;: Talc s l u rry
·� 0.6
::::l
en
c
0
t
0
0. 0.4
e Talc i nsufflation
a..
0.2
0.0
0 50 1 00 1 50
Days
S l u rry 221 1 35 89 66 59 44 26
l nsufflati on 228 1 44 54 61 51 38 19
Total 449 279 1 73 1 27 110 82 45
pneumonia after rece1vmg talc. They reported that Ben-Zeev ( 1 80) observed no cases of ARDS in 360
seven patients (9%) developed ARDS requiring cases who received 2 g talc intrapleurally (some with
mechanical ventilation, and one of their patients died insuffiation and some with slurry) and concluded
( 1 79). All their patients received 5 g of talc, and three that the ARDS was dose related. This is not necessar
patients had received insuffiated talc, whereas four ily the case, however, because of the four fatal cases
received talc slurry. In the multicenter study from the reported by Campos et al. ( 1 8 1 ) all had received only
United States reported by Dresler et al. ( 1 76) , respi 2 g of insuffiated talc. It should be noted, however,
ratory failure developed in 8 (4%) of the patients that when Montes et al. ( 1 82) injected rabbits with
that received talc slurry and 1 8 (8 . 1 %) of the patients 50 or 200 mL/kg talc intrapleurally, there was more
that received insuffiated talc. Eleven of the patients with systemic talc deposition in the animals that received
respiratory failure died ( 1 76) . Brant and Eaton ( 1 26) the high dose than in the rabbits that received the low
reviewed 33 instances of talc pleurodesis performed in dose. In Germany, Schulze et al. ( 1 83) treated 1 0 5
29 patients and reported that major complications of patients with insuffiated talc and reported n o cases of
hypoxemia and hypotension occurred in 7 patients, ARDS. Janssen and coworkers ( 1 84) in a multicenter,
and 2 of these patients died. open-label, prospective cohort study of 5 5 8 patient
In contrast, studies from Europe have reported who received 4 g of calibrated French large-particle
virtually no cases of ARDS or death following the size talc for malignant effusion reported that there
intrapleural administration of talc. Weissberg and no instances of ARDS. However, seven patients did
1 72 PLE U RAL D I S EASES
develop pulmonary infiltrates which they attributed to 8.36 and 1 2.0 ,um in rabbits. They demonstrated
reexpansion pulmonary edema in 2, cardiogenic pul that the intrapleural injection of the smaller talc
monary edema in 1 and respiratory failure unrelated elicited greater pulmonary and systemic talc particle
to talc in 1 ( 1 84) . I am not convinced that the pul deposition ( 1 9 1 ) . Rossi et al. ( 1 92) have shown that
monary infiltrates in some of these patients were not rabbits receiving small talc (median 6.4 ,um) intra
related to the talc. pleurally develop a more intense systemic inflamma
The mechanism responsible for the ARDS after tory response than do rabbits that receive large talc
intrapleural talc is not definitely known. The acute (2 1 .2 ,um) . Maskell et al. ( 1 87) provided support for
lung injury could be due either to talc itself or its this hypothesis in humans when they randomized
contaminants such as bacteria, fungi, endotoxin, 48 patients to receive mixed talc (most talc particles
dolomite, quartz, kaolinite, calcite, or chlorite. One < 1 5 ,um) or graded talc (most particles < 1 0 ,um
hypothesis is that the acute pneumonitis is related removed) . They reported that with the mixed talc
to the systemic absorption of talc with the subse there was a significantly greater increase in the alveo
quent elaboration of inflammatory mediators. This lar to arterial oxygen gradient, a significantly greater
hypothesis is supported by the observations in the decrease in the arterial Po 2 , and a significantly greater
case reported by Rinaldo et al. ( 1 85) in which there increase in the C-reactive protein. We have demon
were large quantities of talc in the bronchoalveolar strated (unpublished observations) that when the
lavage fluid of their patient. Talc particles were also same two-talc preparations are injected in rabbits,
found in the bronchoalveolar lavage in all four of the those receiving the graded talc have much less systemic
patients reported by Ribas-Milanez de Campos et al. deposition of talc than those receiving the mixed talc.
( 1 86) . In addition, one of the patients reported by Moreover, all the talc particles detected systemically
Ribas-Milanez de Campos et al. died, and this patient were < 1 0 ,um in their maximum diameters.
had talc particles present in almost every organ at From the discussion in the preceding text, the evi
autopsy, including the ipsilateral and contralateral dence seems strong that the intrapleural injection of
lung, brain, liver, kidney, heart, and skeletal muscle. talc can cause ARDS and death in a small percentage
The systemic effects of talc are greater than those of patients. The ARDS is probably more common
of the tetracycline derivatives. Maskell et al. ( 1 87) if smaller talc particles are used. Therefore, if one is
randomized 20 patients with malignant effusions to going to use talc to produce a pleurodesis, it is impor
intrapleural tetracycline or mixed talc (most talc par tant to not use talc preparations that have many small
ticles < 1 5 ,um) and measured changes in the systemic talc particles ( < 1 0 ,um) .
inflammation from lung clearance scans, oxygen satu Another complication of talc pleurodesis is pul
rations, and C-reactive protein concentrations from monary embolism. Montes-Worboys et al. ( 1 93)
baseline to 48 hours after pleurodesis. The changes in reviewed 23 1 patients who received talc poudrage
the systemic inflammation, oxygen saturations, and and reported that 1 7 patients (7.4%) died within
the C-reactive protein were all significantly greater 1 5 days and thrombotic events were observed in six
in the group that received the talc ( 1 87) . of these patients. Patients who survived less than
One hypothesis to explain the varying incidences 1 5 days tended to have higher serum IL-8 levels and
of ARDS from country to country is that the inci these higher levels were correlated with the thrombin
dence of ARDS is higher in countries where talc prep antithrombin levels ( 1 93) . It is unknown whether the
arations with smaller particles are used. There would same events happen with talc slurry.
be more systemic inflammation from the talc prepa The intrapleural injection of talc induces chronic
rations with smaller diameters because the smaller inflammation of the pleural space as assessed via PET
talc particles would be more likely to be absorbed scan with F- 1 8 fluorodeoxyglucose ( 1 94) .
through the stomas in the lymphatics of the parietal
pleura. Then their systemic distribution would lead to
Tetracycline Derivatives
systemic inflammation. When talc preparations from
various manufacturers are analyzed, there is a marked Tetracycline derivatives are now the second most
variation in particle size ( 1 88). Animal studies have common agents used for sclerosis ( 1 66) . In the
also demonstrated that talc is deposited extrapleu survey by Lee et al. ( 1 66) , tetracycline derivatives
rally after it is administered intrapleurally ( 1 89, 1 90) . were the sclerosing agent of choice for 25.8% of the
Ferrer et al. ( 1 9 1 ) intrapleurally injected two dif respondents. During the 1 980s, tetracycline was
ferent sizes of talc with mean maximum diameters probably the most commonly used agent for creating
C H APTE R 1 0 I PLE U RAL E F F U S I O N S R E LAT E D TO M ETASTATIC MALI G N A N C I E S 1 73
a pleurodesis. Tetracycline, 35 mg/kg, is effective in than their usual doses. Dikensoy et al. (207) demon
creating a pleurodesis in rabbits ( 1 49). Tetracycline strated that administration of one half (5 mg/kg doxy
is also effective in treating malignant pleural effu cycline plus 200 mg/kg talc) or one fourth (2. 5 mg/kg
sions. Sherman et al. ( 1 95) reported that tetracycline, doxycycline plus 1 00 mg/kg talc) of both the usual
1 ,500 mg, effectively controlled 94.4% of 1 0 8 malig doses of doxycycline ( 1 0 mg/kg) and talc ( 400 mg/kg)
nant pleural effusions. In a review of 1 1 reports in rabbits resulted in a mean pleurodesis score that
involving 359 patients, the success rate with tetracy was better than the mean pleurodesis score with full
cline was 67% ( 1 96) . dose talc and similar to the mean pleurodesis score
Parenteral tetracycline is no longer available in the with full-dose doxycycline.
United States, although it remains available in some
countries such as Germany ( 1 97) . It has recently
Antineoplastics Including Bleomycin,
been shown that oral forms of tetracycline and its
Nitrogen Mustard, and Mitoxantrone
derivatives can be used for pleurodesis if they are dis
solved in saline and then passed through a 0.2-,um Bleomycin is another agent that is sometimes used
sterile and nonpyrogenic polyethersulfone membrane as a sclerosing agent for malignant pleural effusions.
to remove infectious materials and other particulate In the survey of 84 1 pulmonologists, it was the third
particles ( 1 9 8). most commonly used agent but was used only by
Because parenteral tetracycline is no longer avail 6.5% of the respondents ( 1 66) . The popularity of
able, the tetracycline derivatives minocycline and bleomycin is due in part to an older randomized con
doxycycline have been evaluated for their effective trolled study by Ruckdeschel et al. (208) comparing
ness in producing a pleurodesis. In the rabbit, mino the results with 60 units of bleomycin and 1 ,000 mg
cycline, 7 mg/kg ( 1 99), or doxycycline, 1 0 mg/kg of tetracycline in 44 patients. The rate of success with
( 1 48), produces a pleurodesis that is comparable to bleomycin at 30 days (64%) was significantly better
that produced by tetracycline, 35 mg/kg. One dis than that with tetracycline (33%) . It should be noted
turbing aspect of the intrapleural administration of that less than an optimal dose of tetracycline was used
tetracycline derivatives in animals is that it is asso in this study, the rate of success with tetracycline was
ciated with a high incidence of hemothorax, which much less than that generally reported and the rate
is frequently fatal ( 1 99) . The hemothoraces and the of success with bleomycin was not particularly good.
mortality, however, are prevented if chest tubes are Overall, it appears that bleomycin is probably less
inserted into the animals ( 1 48) . effective than talc ( 1 77) or the tetracycline deriva
Doxycycline and minocycline are effective in pro tives in producing a pleurodesis. In a review of eight
ducing pleurodesis in patients with malignant pleural reports with a total of 1 99 patients using bleomy
effusion. When five reports (200-204) with a total of cin to treat malignant pleural effusions, the overall
1 1 0 patients are combined, there was control of the success rate was only 54% ( 1 96) . In a randomized
effusion at 30 days in 9 1 of the patients (83%) . The controlled study, bleomycin failed in 4 1 % of patients
usual dose of doxycycline is 500 mg. There has also at 30 days, 59% at 90 days, and 65% at 1 80 days
been one report in which the administration of mino (209) . In the rabbit model, bleomycin is ineffective
cycline, 300 to 500 mg, produced a complete response in producing a pleurodesis (2 1 0) . Much better results
at 30 days in 62. 5 % of patients and a partial response were reported in a more recent randomized study in
(no need for further thoracentesis) in an additional which 1 60 patients received bleomycin 0.75 mg/kg
25% (205). The primary side effect when pleurodesis or interferon (2 1 1 ) . If the output from the chest tube
is performed with a tetracycline derivative is severe was more than 1 00 ml/day on the third day after the
chest pain (206) . Although the chest pain tends to be bleomycin was administered, a second dose of bleo
worse in patients who receive the tetracycline deriva mycin was administered (2 1 1 ) . In this study, 70 of
tive for a pneumothorax than for a pleural effusion, the 83 patients (84%) responded (2 1 1 ) . Bleomycin is
it is sometimes very severe in patients with malignant much more expensive (�US $ 1 ,000/patient) than the
pleural effusions. It is recommended that patients tetracycline derivatives or talc. Therefore, it cannot be
who receive a tetracycline derivative for pleurode recommended.
sis be given lorazepam or midazolam in addition to Nitrogen mustard was one of the first antineo
systemic pain medications before the injection. plastic agents to be used intrapleurally to treat malig
It is possible that pleurodesis can be induced by nant pleural effusions. Interestingly, the results with
using combinations of sclerosing compounds at less nitrogen mustard on the average are better than those
1 74 PLE U RAL D I S EASES
with bleomycin. For example, Leininger et al. (2 1 2) evaluated for the treatment of pleural effusions, and
administered 1 0 mg of nitrogen mustard through a the response rates have been less than 50% for almost
chest tube to 1 8 patients and reported that the treat all of them ( 1 96) . Therefore, their use cannot be
ment was effective in 1 7 (94%) . Kinsey et al. (2 1 3) recommended.
administered 30 mg of nitrogen mustard through a
tube thoracostomy in 62 patients and reported that
Silver N itrate
the effusion was completely controlled in 57 (92%) .
Interestingly, nitrogen mustard is the only antineo Silver nitrate was probably the first agent used to
plastic agent other than mitoxantrone that is effective produce a pleurodesis and is very effective (2 1 9) . In
in producing pleurodesis in animals (2 1 4) . Because the 1 980s, silver nitrate was replaced by tetracycline
nitrogen mustard is at least as effective as bleomy probably because of severe side effects seen after the
cin and costs less than US$ 1 00 per patient, it is the intrapleural injection of high concentrations of this
agent of choice if an antineoplastic agent is going to agent. Subsequent studies in rabbits demonstrated
be used. that a lower concentration of silver nitrate (0.5%)
Another antineoplastic agent that has shown some was as effective as tetracycline, 35 mg/kg, in produc
promise as a pleural sclerosant is mitoxantrone. In one ing a pleurodesis (220) . Moreover, this dose of silver
report, 1 1 4 patients with malignant pleural effusions nitrate is superior to talc, 400 mg/kg, in producing
from breast carcinoma were given 40 mg mitoxantrone pleurodesis in rabbits (22 1 ) .
intrapleurally and there was complete response of the There has been two clinical study evaluating the
effusion in 53.5% and partial response of the effusion efficacy of silver nitrate in inducing pleurodesis in
in another 25% (2 1 5) . In another report, the intra patients with malignant pleural effusions (222) . Pas
pleural administration of 40 mg mitoxantrone to choalini et al. (222) randomized 60 patients with
2 1 patients with malignant pleural effusions resulted in malignant pleural effusion to receive either 5 g talc
complete control of the effusions in all cases at 60 days slurry or 20 mL of 0 . 5% silver nitrate through a chest
(2 1 6) . One advantage of mitoxantrone over bleomy tube. In the 47 patients who returned for follow-up,
cin is that mitoxantrone binds to cell membranes the silver nitrate tended to be more effective than talc
and is therefore likely to remain in the pleural space (222) . There was no recurrence of the effusion in
longer (2 1 7) . Interestingly, the intrapleural injection 96% of the patients who received silver nitrate and
of mitoxantrone in rabbits leads to proportionately 84% of the patients who received talc slurry (222) .
more inflammation than does the intrapleural injec In a later study (223), pleurodesis was performed as
tion of tetracycline derivatives or talc ( 1 50). In the an outpatient in 65 patients with malignant pleural
animal model, the intrapleural administration of high effusion and at 30 days recurrences had occurred in 2
doses of mitoxantrone produces a pleurodesis, but of 48 hemithoraces (4%) . These studies suggest that
there is a significant amount of cardiotoxicity ( 1 50) . silver nitrate should be considered a reasonable alter
Cardiotoxicity has also been reported i n patients who native to other commonly used pleurodesing agents
have received mitoxantrone intrapleurally. Currently, such as the tetracycline derivatives or talc slurry.
mitoxantrone is not recommended as a sclerosing
agent because it is no more effective than talc or the
lodopovidone (Betadine)
tetracycline derivatives, is associated with more toxic
ity and is very expensive (US$4,000/patient) . It appears that iodopovidone is also an effective agent
Another agent that has shown some promise is for producing pleurodesis. Olivares-Torres et al.
intrapleural cisplatin. Seto et al. (2 1 8) administered (224) injected 1 00 mL of 2% iodopovidone into the
25 mg cisplatin in 500 ml distilled water intrapleu pleural spaces of 40 patients at the end of a thoraco
rally in 80 patients with non-small-cell lung cancer. scopic procedure and 1 2 patients through tube tho
At 4 weeks, they reported that 34% of the patient racostomy. They reported that a complete response
achieved a complete response and 49% achieved with no reaccumulation of the fluid was obtained in
a partial response and the median survival was 50 of the 52 patients (96%) . Three patients devel
239 days (2 1 8) . If these results can be confirmed, oped intense pleuritic pain and systemic hypotension
ciplatins might become a reasonable alternative to after the intrapleural instillation of the iodopovidone,
produce a pleurodesis. but there were no fatalities (224) . In a second study,
Other antineoplastic agents such as doxorubicin, 37 patients were given 1 00 mL of 2% iodopovidone
etoposide, fluorouracil, and mitomycin-C have been through their chest tube and there was no recurrence
C H APTE R 1 0 I PLE U RAL E F F U S I O N S R E LAT E D TO M ETASTATIC MALI G N A N C I E S 1 75
of the pleural effusion in 32 (86.5%) (225) . In a used for malignant pleural effusion in 98 patients
third study (226) , 6 1 procedures were performed in with a success rate of 86% (232) . In a more recent
54 patients and there was only one recurrence with report, quinacrine 500 mg was instilled through tube
a mean follow-up of 5 . 6 months. In a fourth study thoracostomy in 54 patients (233) . The injection was
(227) , which was a randomized controlled study, repeated if there was more than 50 mL pleural drain
45 patients were randomized to receive 4 grams of age per day after 3 days. This treatment was successful
insufHated talc or 50 ml 2% iodopovidone through at 2 months in 47 of the 54 patients (87%) (233) .
a chest tube The recurrence rates were similar in The intrapleural injection of quinacrine causes a sys
each group and the side effects tended to be greater temic inflammatory response similar to that seen after
in the patients who received insuffiated talc (227) . talc injection (234) .
In rabbits, the intrapleural administration of 2% or
4% iodopovidone is comparable in effectiveness to
Agent of Choice
doxycycline 1 0 mg/kg in producing a pleurodesis
(228) . There is one report (229) from Germany in The following recommendations for the selection of
which three patients who received 200 to 500 ml of an agent for pleurodesis in patients with malignant
a 1 0% iodopovidone solution developed blindness. pleural effusions are made based on the informa
However, this dose is about 25 times larger than the tion provided in the preceding text. If a patient has a
recommended dose. None of the other studies have malignant pleural effusion and is being treated with
reported visual problems. Iodopovidone should also tube thoracostomy, then the agent of choice is doxy
be considered a reasonable alternative to other com cycline, 500 mg. Alternative tetracycline derivatives
monly used pleurodesing agents such as the tetracy are tetracycline, 1 ,500 mg, or minocycline, 300 mg.
cline derivatives or talc slurry. If a tetracycline derivative is unavailable, 20 mL of
0.5% silver nitrate or 1 00 mL of 2% iodopovidone
are reasonable alternatives. If a patient has a malig
Other Agents
nant pleural effusion diagnosed during thoracoscopy
Over the years, there have been many other articles or at thoracotomy, the patient should be subjected to
evaluating the capability of various other agents to pleural abrasion or parietal pleurectomy. Talc would
produce a pleurodesis. Several agents that presumably be recommended if it were not associated with the
work by being immune modulators have been evalu development of the ARDS. One should only use talc
ated. Dried killed Corynebacterium parvum, an anaer if it is large particle talc.
obic gram-positive bacterium, has been evaluated in
nine reports with a total of 1 69 patients who had an
lntrapleural Injection of Sclerosing Agent
overall success rate of 7 6% ( 1 96) . The availability of
C. parvum worldwide is very limited, and this limits Pleurodesis should be performed by injecting the
its overall use. sclerosant through a chest tube. The reason to use
OK-432 is obtained from the SU strain of Strepto a chest tube rather than just injecting the sclerosing
coccuspyogenes and has properties similar to C. parvum agent into the pleural effusion is that fusion of the
in that it is both immunostimulating and cytotoxic. visceral and the parietal pleura to create a pleurodesis
In Japan, it is considered by some to be the sclerosing requires that the two pleural surfaces be next to each
agent of choice (230) . Response rates as high as 75% other. If they are not in close apposition, pleurodesis
have been obtained. In one study, the administration is unlikely to occur.
of the combination of OK-432 and 30 mg of doxo The first question in performing a pleurodesis
rubicin resulted in complete control of the effusion in is on the size of chest tube to be used. There is no
80% of the patients (23 1 ) . In another study, the com evidence that the use of a large tube provides better
bination of OK-432 and cisplatin resulted in com results than does the use of a small tube (235-238) .
plete control of the effusion for 1 80 days in 87% of There have been three randomized controlled stud
15 patients whereas OK-432 by itself resulted in control ies (235-237) with a total of 1 66 patients comparing
in only 47% of 17 patients ( 1 20). To my knowledge, large-bore to small-bore chest tubes for pleurodesis.
OK-432 is available only in Asia. All three concluded that the two different tube sizes
Quinacrine, the antimalarial agent, has been used were equivalent (235-237) .
for decades for pleurodesis in Scandinavian countries. The chest tube is connected to a water-sealed drain
Until 1 993, there were reports of quinacrine being age system, and the effusion is allowed to drain (2 1 2) .
1 76 PLE U RAL D I S EASES
It is recommended that negative pressure not be et al. (244) injected urokinase 1 00,000 daily for at
applied initially to the chest tube in this situation least 3 days in 12 patients with trapped lungs and 36
because the combination of a chronic pleural effu patients with loculated pleural fluid. They reported
sion and the application of negative pleural pressure can that they obtained lung reexpansion in 29 of the 48
cause reexpansion pulmonary edema (see Chapter 24) . patients (60.4%) (244) . These 29 patients subse
If the lung has not expanded within 24 hours, then quently underwent minocycline pleurodesis which
negative pressure should be applied to the chest tube. was successful at 1 month in 27 patients (93%) (244) .
The pleurodesis is less likely to be successful if If the lung does not expand with tube thoracostomy
the patient is on corticosteroids or nonsteroidal anti with or without the instillation of fibrinolytics, the
inflammatory drugs (NSAIDs) . The induction of a pleural fluid can be drained on a chronic basis with
pleurodesis usually involves the creation of intense the PleurX catheter or with a pleuroperitoneal shunt
intrapleural inflammation, which then leads to fibrosis. (see the discussion of pleuroperitoneal shunt later in
Therefore, it is not surprising that anti-inflammatory this chapter) . One of these options should definitely
drugs decrease the efficacy of pleurodesis. In animal be performed if the mediastinum is shifted away from
studies, the efficacy of pleurodesis is decreased in rab the side of the effusion (Fig. 1 0.4A) .
bits given doxycycline and corticosteroids (239) or The injection of any of the tetracycline derivatives
the NSAID diclofenac (240) , and in rabbits given talc produces an intense pleuritis that can be very painful.
slurry and corticosteroids (24 1 ) , and in pigs treated Accordingly, patients should be given systemic medi
with mechanical abrasion and diclofenac (242) . cation to control the pain. We currently use loraz
Accordingly, it is recommended that corticosteroids epam or midazolam to produce conscious sedation.
and NSAIDs not be administered to patients undergo Sherman et al. (245) have suggested that the patient
ing pleurodesis. should be given local anesthesia such as lidocaine
Once the chest tube has been inserted, how long hydrochloride intrapleurally. There are no controlled
should one wait before injecting the sclerosing agent? studies evaluating the efficacy of intrapleural lido
It is important to make certain that the underlying caine, and, currently, no intrapleural anesthetic is rec
lung has fully expanded before the injection is made. ommended because the patient remembers no pain
If the underlying lung has not expanded, then the due to the conscious sedation.
injection of a sclerosing agent will lead only to addi After the sclerosant is injected, the catheter is
tional thickening of the visceral pleura, which will flushed with an additional 50 to 1 00 mL of saline and
further compromise the function of the underlying the chest tube is clamped for at least 1 hour. Although
lung. Some authors have advocated that the scleros in the past it has been recommended that the
ing agent not be injected until the drainage from the patient be moved into different positions so that the
chest tube is less than 1 50 mL/day (208). However, sclerosant contacts all the pleural surfaces, this does
there is no supporting data for this practice. Vil not appear to be necessary. In animals, the disper
lanueva et al. (243) randomly assigned patients to a sal of radioisotopes injected intrapleurally is similar
group in which tetracycline was not instilled until the whether the animals are rotated or not (246) . In
drainage was less than 1 50 mL/day and a group in humans, the dispersal of the injected radioisotopes is
which 1 ,500 mg of tetracycline was instilled as soon similar whether the patients are rotated or not (247) .
as the lung had reexpanded. The rate of success was Rotation did not have a statistically significant effect
the same in each group (80%) , but the duration of on the results of pleurodesis in one randomized study
the chest tube drainage was much less in the latter with tetracycline derivatives (248) . The rate of suc
group (2 days) than in the former group (7 days) . In cess with rotation was 73.7% whereas that with no
view of this study, it is recommended that the scle rotation was 6 1 .9% (248) . However, because rotating
rosant be injected as soon as the lung has reexpanded. the patient certainly does not decrease the likelihood
If successful reexpansion of the lung cannot be of pleurodesis and because the patient has much less
accomplished with pleural drainage, as shown in Fig pain with rotation if the small catheters are used, it
ure 1 0.4B, sclerosing agents should not be injected into is recommended that all patients be rotated unless it
the pleural space. The injection can only thicken the is particularly uncomfortable for the patient.
visceral pleura and allow lesser lung expansion. Another After 1 to 2 hours, the chest tube is undamped
reason for lack of expansion of the lung is loculation of and negative pressure (-1 5 to -20 cm H 2 0) is applied
the pleural fluid. One approach in both of these situa to the chest tube. Suction is maintained for at least
tions is to inject a fibrinolytic agent intrapleurally. Hsu 24 hours and until the pleural drainage is less than
C H APTE R 1 0 I PLE U RAL E F F U S I O N S R E LAT E D TO M ETASTATIC MALI G N A N C I E S 1 77
1 50 ml/day. The chest tube is removed after 96 hours the procedure of choice to achieve pleurodesis (25 1 ) .
regardless of the volume of pleural fluid. The keys to However, the available evidence does not support this
the success of this procedure are the pleuritis produced conclusion. The randomized study by Dresler et al.
by the sclerosant and the approximation of the visceral ( 1 76) (Fig. 1 0.5) showed that there was no significant
and the parietal pleura by the chest tubes so that a difference in recurrence rates between patients treated
pleural symphysis can occur. There appears to be no with thoracoscopy and talc insuffiation and those
advantage if the sclerosant is injected twice. In one treated with tube thoracostomy with injection of talc
study, 25 patients received one injection of tetracy slurry through the chest tube. Moreover, Heffner et al.
cline, 20 mg/kg, and 25 patients received instillations (7 1 ) reported that the results with thoracoscopy were
of tetracycline, 20 mg/kg, on 2 consecutive days. Effu not significantly better than those with tube thoracos
sions recurred in four patients in each group (249) . tomy. Lastly, Yim et al. (252) randomized 28 patients
It would be ideal if pleurodesis could be performed to pleurodesis with thoracoscopy and talc insuffiation,
on an outpatient basis. Pleurodesis accomplished on an and 27 patients to pleurodesis with tube thoracos
outpatient basis has two advantages. First, the patient tomy and talc slurry. They found that there were no
is not hospitalized. Because the mean hospitalization significant differences in the results with the two dif
time of patients treated in hospital is approximately ferent methods (249). Moreover, thoracoscopy adds
6.5 days ( 1 23) and the patient has a life expectancy of significantly to the expense of the procedure.
only approximately 90 days, this hospitalization repre However, if thoracoscopy is performed for an
sents 5% of and probably the best days of the patient's undiagnosed recurrent pleural effusion, an attempt
remaining life. Second, it decreases the cost of the should be made to induce a pleurodesis at the time of
treatment because there are no hospitalization costs. the procedure. It should be noted that if thoracoscopy
Terra et al. (223) performed pleurodesis with silver is performed and no pleural sclerosis is attempted,
nitrate as an outpatient in 65 patients with malignant 62% of the patients have no recurrence of their pleu
pleural effusion and reported that at 30 days recur ral effusion for the remainder of their life (253,254) .
rences had occurred in 2 of 48 hemithoraces (4%) . In nonrandomized studies, thoracoscopy with the
The study was performed by inserting a pigtail cath insuffiation of talc has been reported to be very effec
eter on day 1 and then draining the pleural fluid into a tive in creating a pleurodesis in patients with malig
collection bag. Then on the following day, 30 ml 0.5% nant pleural effusion. Ribas-Milanez de Campos et al.
silver nitrate was injected and allowed to remain for ( 1 86) treated 383 patients with malignant pleural
1 hour. Then the patients returned after 7 days and effusions with 2 g of insuffiated talc and reported a
had the catheter removed (223). Reddy et al. (250) success rate of93.4%. Kolschmann et al. (255) treated
combined talc poudrage and insertion of an indwell 1 02 patients with 8 g of insuffiated talc and reported
ing catheter in 30 patients. After the patients had the that the treatment was successful (no requirement for
talc poudrage, an indwelling catheter was placed and repeat thoracentesis) in 89% of 85 surviving patients
the catheter was drained every other day until the at 30 days and in 83% of 46 patients surviving at
drainage was less than 1 50 ml on two consecutive days 1 80 days (25 5). Arapis et al. (256) reported their results
and then it was removed. They reported that pleurode in 24 1 patients who received 5 g of insuffiated talc at
sis was successful in 92% of the cases, the median hos the time of thoracoscopy. Of the 1 72 patients seen at
pitalization following the procedure was 1 . 79 days follow-up at 1 month, the chest radiograph showed
and that the catheters were removed at a median of no fluid in 1 22 (7 1 %), a small effusion in 39 (23%) ,
7.54 days. This procedure has the advantage that the and a recurrence in 1 1 (6%) (253) . The results in some
hospitalizations are decreased compared with usual studies have been significantly worse ( 1 76,257) . Love
pleurodesis and that the indwelling catheter remains et al. (257) reviewed 60 patients treated with 4 to 5 g
in place for a much shorter time. insuffiated talc and reported that complete control of
the effusion was obtained until death in only 52% of
patients. These results are similar to those reported by
Thoracoscopy for Pleurodesis
Dresler et al. ( 1 76) as shown in Figure 1 0 . 5 .
Some have recommended that pleurodesis should Th e insuffiation o f talc at the time of thoracos
be performed through thoracoscopy. Indeed, it has copy is not without significant side effects. In the
been stated that video-assisted thoracoscopic surgery multicenter study reported by Dresler et al. ( 1 76) , 1 8
(VATS) with talc poudrage has replaced conventional o f 223 patients (8 . 1 % ) developed respiratory failure
instillation of talc slurry through tube thoracostomy as and 6 (3%) died. Froudarakis et al. (258) reported
1 78 PLE U RAL D I S EASES
that patients who underwent thoracoscopy with Other than the study by Putnam et al. ( 1 23), there
talc insuffiation had a significantly greater tempera have been no randomized controlled studies compar
ture elevation, peripheral blood WBC increase, and ing the two modalities and each has its advantages
C-reactive protein level increase than did patients and disadvantages and each has its supporters. The
who underwent thoracoscopy without talc insuf primary advantage of the indwelling catheter is that
flation. In a more recent study of 84 patients from its insertion does not require hospitalization. Indeed
Chicago, 5 (5.9%) developed acute lung injury and the total hospital days associated with treatment of
severe hypoxemia developed in 25 (29 .8%) (259) . the pleural effusion is less when the indwelling cath
Because the study by Dresler et al. ( 1 76) indicates eter is used than when pleurodesis is performed. Fysh
that insuffiated talc can cause death, and talc is not et al. (264) compared the number of in hospital days
significantly more effective than other agents, what in 34 patients who elected to receive the indwelling
are the alternatives? The three best alternatives appear catheter and 3 1 patients who elected talc pleurode
to be pleural abrasion, the intrapleural instillation of sis. They reported that median number of hospital
2% iodopovidone, or the intrapleural instillation of days was significantly greater in the pleurodesis group
collagen. ( 1 8 days) than in the indwelling pleural catheter
In animal studies, mechanical abrasion of the group (6.5 days) . Moreover, the median number of
pleura is at least as effective in producing pleurodesis hospital days related to the effusions was significantly
as is talc ( 1 7 1 , 1 72) . Mechanical abrasion is also effec greater in the pleurodesis group ( 1 0 days) than in
tive in controlling malignant effusions in humans. the indwelling pleural catheter group (3 days) . The
Crnjac (260) treated 44 patients with malignant patients in the indwelling catheter group also spent
pleural effusion with mechanical pleural abrasion a smaller percentage of their remaining lives (8.0%)
and reported that at 6 months there was no recur in the hospital compared with the pleurodesis group
rence of the effusion in 93% of the patients. In a ( 1 1 .2%) . In a second study, Hunt et al. (265) retro
subsequent report, Crnjac et al. (26 1 ) randomized spectively compared the course of 59 patients who
87 patients with breast cancer to receive either mech had indwelling catheter placed and 50 who were
anical pleurodesis or insuffiated talc. They reported treated with thoracoscopic talc pleurodesis. They
that both methods were equally effective in patients reported that repeat interventions for recurrent
with pleural fluid pH level above 7.3 but that ipsilateral pleural effusions were significantly more
mechanical abrasion was more effective in patients common with talc pleurodesis ( 1 6%) than with the
with pleural fluid pH level below 7.3 (8 1 % vs. 55%). indwelling catheter (2%) (265). In addition, they
Moreover, the duration of chest tube drainage and the reported that the length of hospitalization was sig
duration of hospitalization were significantly longer nificantly greater in the group that received talc
in the patients who received talc insuffiation (26 1 ) . pleurodesis (265). They concluded that placement of
Single center studies have shown the effectiveness an indwelling catheter should be considered for pal
of intrapleural iodopovidone and collagen. Olivares liation of malignant pleural effusions (265) .
et al. (224) instilled 1 00 mL of 2% iodopovidone in Another advantage of the indwelling catheter is
52 patients with pleural effusions and reported that that if spontaneous pleurodesis occurs, only rarely
there was no recurrence of the effusions in 50 (96%) . are additional interventions necessary to treat recur
Akopov et al. (262) instilled 1 g of bovine dermal col rent pleural effusions ( 1 29). The indwelling pleu
lagen powder in 45 patients with malignant pleural ral catheter is certainly the treatment of choice for
effusions and reported that this was successful in 89% patients who have trapped lungs (266) or who fail
of patients. pleurodesis (264) .
In view of the discussion in the preceding text, it There are also several advantages to control
appears that mechanical abrasion of the pleura is the ling the effusion with pleurodesis. The procedure is
best method to induce a pleurodesis in patients who completed within 1 week, and the patient does not
undergo thoracoscopy for an undiagnosed pleural have to worry about draining his or her pleural space
effusion. intermittently for prolonged periods. Available home
health or a very reliable patient and/or family are not
necessary with pleurodesis.
Choosing Between Indwelling Catheter
Side effects occur with both procedures. With the
and Pleurodesis
indwelling catheter, the pleural space may become
As mentioned earlier, these are the two main options infected or the catheter may become occluded.
for managing symptomatic pleural effusions (263) . Skin infections occur. Some patients experience
C H APTE R 1 0 I PLE U RAL E F F U S I O N S R E LAT E D TO M ETASTATIC MALI G N A N C I E S 1 79
pain when they drain their pleural fluid. This can be If the primary symptom is dyspnea, the patient
diminished if the fluid is removed slowly. Patients should be given opiates or oxygen, or both, if they
who undergo pleurodesis frequently have chest pain are not candidates for insertion of the PleurX cath
or fever after the sclerosing agent is instilled. eter or chemical pleurodesis. Both opiates and oxy
Since there are no good randomized controlled gen relieve the dyspnea. The disadvantage of opiates
studies, it is probably best to present both options is that their administration can be associated with an
to the patient and let them choose. In the study of increase in the Paco 2 , which, in turn, will lead to a
Fysh et al. (264) , approximately 50% of the patients decrease in the Pao 2 " The disadvantage of oxygen is
chose each option. I gave a talk at M. D. Anderson in that it is very expensive and is not portable. Opiates
20 1 0 on the treatment of malignant pleural effusions are probably underused in the treatment of dyspnea
discussing both the indwelling catheter and pleurode associated with pleural effusions. It is recommended
sis. I was surprised when one of the physicians in the that they be titrated for the degree of dyspnea as they
audience asked me why I even discussed pleurodesis are titrated for the degree of pain.
virtually all malignant pleural effusions at that institu
tion are treated with the indwelling catheter. Pleuroperitoneal Shunt
An alternative approach to the management of
Alternatives to Pleurodesis and the
patients with malignant pleural effusions is the place
Indwelling Catheter
ment of a pleuroperitoneal shunt (268-270) , which
If the patient is not a good candidate for chemical is marketed commercially by CareFusion Corpo
pleurodesis or insertion of the PleurX catheter, there ration, San Diego, CA. This device consists of two
are several options available that include symptom catheters connected with a valved pump chamber
atic treatment, the implantation of a pleuroperitoneal (Fig. 1 0.6) . The two one-way valves in the pump
shunt, serial thoracentesis, and pleurectomy. chamber are positioned such that fluid can only flow
from the pleural space to the pump chamber to the
peritoneal cavity. These valves open at a positive pres
Symptomatic Treatment
sure of approximately 1 cm H 2 0. Because the pleural
The two primary symptoms associated with a malig pressure is almost always more negative than the peri
nant pleural effusion are chest pain and shortness of toneal pressure, the pump chamber must be used to
breath. If the patient has chest pain, sufficient analge move fluid from the pleural cavity to the peritoneal
sics should be given to control the pain. There is no cavity. The capacity of the pump chamber is approxi
reason to worry about narcotic addiction because the mately 1 . 5 mL. When it is compressed, fluid is
life expectancy of the patient is so short. forced from the chamber into the peritoneal cavity.
First Second
one-way valve one-way valve
Pump chamber Radiopaque
blue stripe
Fenestrated pleural Fenestrated peritoneal
catheter (proxi mal) catheter (d istal)
Then when the pump chamber is released, negative Since this original report (269), there have been sev
pressure created in the pump chamber draws fluid eral additional reports (270-273) . In these latter four
from the pleural cavity to the pump chamber. studies, a total of 84 patients with malignancy were
When the pleuroperitoneal shunt was first treated and 76 had either failed pleurodesis or had a
released, the pump chamber was always placed in a trapped lung. Alleviation of dyspnea and control of the
subcutaneous pocket caudal to a skin incision in the effusion was obtained in more than 90% of the patients.
lateral part of the inframammary crease and fixed to These results are particularly impressive when one real
the tissues with sutures through the sewing holes in izes that previous treatment failed in more than 90% of
the base of the pump chamber. Subsequently, a modi the patients in these four series. The primary complica
fication of the shunt has been developed in which the tion with the pleuroperitoneal shunt is that in approxi
pump chamber is placed exteriorly and the volume of mately 1 5% of patients, the shunt becomes occluded.
the chamber is 2.5 mL rather than 1 . 5 mL. There In such a situation, it should be replaced (274) .
fore, less pumping is required and the pumping is What should be the place of the pleuroperitoneal
easier with the pump chamber exterior. The pleural shunt in the management of patients with malig
catheter is inserted at the lateral and superior aspect nant pleural effusions? The shunt should certainly
of the incision with a Seldinger-type introducer kit. be considered along with the indwelling catheter in
The abdominal catheter is tunneled subcutaneously the patient in whom the lung does not expand after
across the costal margin and inserted, through a 3-cm tube thoracostomy. Petrou et al. (275) inserted
skin incision in the anterior abdominal wall, into pleuroperitoneal shunts in 63 patients who had a
the peritoneal cavity through a small incision in the trapped lung at thoracoscopy and reported effective
peritoneum. Pump compression by physicians and palliation in more than 95%. The pleuroperitoneal
nurses is initiated in the recovery room and is then shunt should also be considered in patients in whom
started by the patient or the patient's family on the pleurodesis has failed.
first postoperative day. Within 3 days of shunt place There are no studies comparing the efficacy of the
ment, pump compression usually causes no discom shunt and chemical pleurodesis. The advantages of
fort. Most patients are ready for discharge from the the shunt include the following: (a) the total hospi
hospital within 48 hours of the operation. Selected talization time is less than with chemical pleurodesis,
patients can have the procedure performed on an out (b) the amount of pain is probably less than with
patient basis. pleurodesis, (c) the procedure can be performed on
The first sizable series of patients receiving the an outpatient basis, and (d) the patient may benefit
pleuroperitoneal shunt was reported by Little et al. psychologically from using the pump when he or she
(269) in 1 988. They inserted shunts into 29 patients is dyspneic. The disadvantages of the shunt include
with excellent results overall. Eight of the patients the following: (a) the shunt becomes obstructed in
previously had chest tube placement with attempted some patients, (b) insertion frequently requires gen
sclerosis with tetracycline. In most instances, shunt eral anesthesia, (c) the shunt must be inserted by a
implantation was performed in the operating room surgeon, and (d) the patient must use the pump daily.
using general anesthesia, but in four patients, local Currently, when outpatient therapy of malignant
anesthesia was used. Three patients never pumped pleural effusions is undertaken, the PleurX catheter is
on their shunts, one obese patient was unable to recommended because it is easier to insert. However,
localize the pump chamber, a 90-year-old man suf if a chylothorax is present, the pleuroperitoneal shunt
fering from senility was simply unable to comply, is recommended because the nutritional status of the
and one patient developed a cancer phobia and patient is preserved with this method. The pleuroperi
refused to touch her shunt. Two patients died shortly toneal shunt is also preferred in patients in whom the
after implantation from disseminated malignant underlying lung is trapped, because in this instance, the
disease. Excellent results were obtained in 20 of the patient will require some type of pleural drainage for
remaining 24 patients. The 1 4 patients with malig the rest of his or her life.
nant effusions had a median survival of 4 months,
and there were no instances of peritoneal tumor seed
Pleurectomy
ing. In 5 patients, the shunts became occluded by
fibrinous debris between 3 weeks and 2 months after In carefully selected individuals, pleurectomy can
the operation. Replacement was uneventful in all five be of use in controlling malignant pleural effusions.
instances. Pleurectomy may be attempted in two different
C H APTE R 1 0 I PLE U RAL E F F U S I O N S R E LAT E D TO M ETASTATIC MALI G N A N C I E S 181
situations (276) . The first is in the patient who under 4. Anderson CB, Philpott GW, Ferguson TB. The treatment of
goes a diagnostic thoracotomy for an undiagnosed malignant pleural effusions. Cancer. 1 974;33 : 9 1 6-922.
5 . Naito T, Satoh H, Ishikawa H , et al. Pleural effusion as a
pleural effusion. If malignant disease is found, an significant prognostic factor in non-small cell lung cancer.
immediate parietal pleurectomy is useful to prevent Anticancer Res. 1 997; 1 7:4743-4746.
recurrence of the effusion (276) . Parietal pleurectomy 6. Johnston WW. The malignant pleural effusion: a review of
consists of stripping all of the parietal pleura from the cytopathologic diagnoses of 584 specimens from 472 consecu
tive patients. Cancer. 1 9 8 5 ; 5 6:905-909.
rib cage and the mediastinum. The second situation is
7. Chernow B, Sahn SA. Carcinomatous involvement of the
in the symptomatic patient with a persistent pleural pleura. Am ] Med. 1 977;63:69 5-702.
effusion and trapping of the ipsilateral lung so that 8 . Niho S, Kubota K, Yoh K, et al. Clinical outcome of chemo
the injection of sclerosing agents is contraindicated. radiation therapy in patients with limited-disease small cell
The surgical procedure involves decortication of lung cancer with ipsilateral pleural effusion. ] Thorac Oncol.
2008;3:723-727.
the trapped lung in conjunction with parietal pleu 9. Lai CL, Tsai CM, Tsai TT, et al. Presence of serum anti-p5 3
rectomy. Pleurectomy controls the pleural effusion antibodies i s associated with pleural effusion and poor progno
in more than 90% of cases (276) . However, it is a sis in lung cancer patients. Clin Cancer Res. l 998;4:3025-3030.
substantial operation with a mortality rate of approxi 10. Shimizu K, Yoshida }, Nagai K, et al. Visceral pleural invasion
mately 1 0% (276) . A pleurectomy combined with is an invasive and aggressive indicator of non-small cell lung
cancer. ] Thorac Cardiovasc Surg. 200 5 ; 1 30: 1 60- 1 6 5 .
decortication is recommended only in a patient who 1 1 . Riquet M, Badoual C, L e Pimpec Barthes F, e t al. Visceral
is symptomatic from the pleural effusion, who is in pleura invasion and pleural lavage tumor cytology by lung
good overall condition, and whose primary tumor is cancer: a prospective appraisal. Ann Thorac Surg. 2003;75 :
either under control or is progressing slowly. I have 353-3 5 5 .
yet to see such a patient, but an occasional patient 1 2 . Renshaw AA , Madge R , Sugarbaker DJ, e t al. Malignant pleu
ral effusions after resection of pulmonary adenocarcinoma.
with breast carcinoma meets these criteria. One recent Acta Cytol. 1 998;42: 1 1 1 1 -1 1 1 5 .
study concluded that thoracotomy with decortica 1 3 . Sugiura S, Ando Y, Minami H, e t al. Prognostic value o f pleu
tion was not indicated in patients with a malignant ral effusion in patients with non-small cell lung cancer. Clin
effusion because of poor survival, a high frequency of Cancer Res. 1 9 97;3 :47-5 0 .
1 4 . Detterbeck F C , Boffa DJ , Tanoue LT. The new lung cancer
complications, and prolonged hospital stay (277) .
staging system. Chest. 2009; 1 36 :260-27 1 .
1 5 . Rodriguez-Panadero F. Lung cancer and ipsilateral pleural
Thoracentesis effusion. Ann Oncol. 1 995;6(suppl 3):S25-S27.
16. Decker DA, Dines DE, Payne WS, et al. The significance of a
In the past, many patients with malignant pleural cytologically negative pleural effusion in bronchogenic carci
effusion were managed with repeated therapeutic noma. Chest. 1 978;74:640-642.
17. Ruffini E, Rena 0 , Bongiovanni M, et al. The significance
thoracenteses for symptomatic relief This regimen of intraoperative pleural effusion during surgery for broncho
has several drawbacks. After therapeutic thoracen genic carcinoma. EurJ Cardiothorac Surg. 2002;2 l :508-5 1 3 .
tesis, malignant pleural effusions reaccumulate rap 1 8 . Tremblay A , Robbins S, Berthiaume L , e t al. Natural history
idly, usually within 1 to 3 days (4) . Therefore, the of asymptomatic pleural effusions in lung cancer patients.
] Bronchol. 2007; 14:98-100.
patient must visit the physician frequently for the
1 9 . Kimura H, Fujiwara Y, Sone T, et al. EGFR mutation status in
procedure. In addition, repeated thoracenteses often tumour-derived DNA from pleural effusion fluid is a practical
lead to loculation of the pleural fluid, which makes basis for predicting rhe response to gefitinib. BrJ Cancer. 2006;
subsequent pleurodesis difficult (4) . In view of these 9 5 : 1 390- 1 3 9 5 .
disadvantages, serial therapeutic thoracentesis should 2 0 . S o h J, Toyooka S, Aoe K , e t al. Usefulness of EGFR mutation
screening in pleural fluid to predict rhe clinical outcome of
be performed only in moribund patients in whom the gefitinib treated patients with lung cancer. IntJ Cancer. 2006;
procedure offers symptomatic relief. 66:78 54-78 5 8 .
2 1 . Fracchia AA , Knapper WH, Carey JT, e t al. Intrapleural che
motherapy for effusion from metastatic breast carcinoma.
REFERENCES Cancer. 1 970;26:626-629.
22. Goldsmith HS, Bailey HD, Callahan EL, et al. Pulmonary
I . Light RW, Erozan YS, Ball WC. Cells in pleural fluid: their lymphangitic metastases from breast carcinoma. Arch Surg.
value in differential diagnosis. Arch Intern Med. 1 973; 1 32: 1 967;94 :483-488.
854-860. 23. Banerjee AK, Willetts I, Robenson JF, et al. Pleural effusion
2 . Marel M, Arustova M, Stasny B, et al. Incidence of pleural in breast cancer: a review of the Nottingham experience. Eur]
effusion in a well-defined region: epidemiologic study in Surg Oncol. l 994;20:33-36.
central Bohemia. Chest. 1 993; 1 04: 1 486- 1 4 8 9 . 24. Apffelstaedt JP, Van Zyl JA, Muller AG . Breast cancer compli
3 . Spriggs A I , Boddington MM. The Cytology ofEjfosions, 2 n d ed. cated by pleural effusion: patient characteristics and results of
New York, NY: Grune & Stratton; 1 968. surgical management. J Surg Oncol. 1 99 5 ; 5 8 : 1 73- 1 7 5 .
1 82 PLE U RAL D I S EASES
2 5 . Van Galen KP, Visser HP, Van Der Ploeg T, et al. Prognostic 48. Collins PO, Connolly OT, Williams TJ. Characterization of
facrors in patients with breast cancer and malignant pleural increase in vascular permeability induced by vascular perme
effusion. Breast]. 20 1 0; 1 6:675-677. ability factor in vivo. Br J Pharmacol. 1 993; 1 09: 1 95-199.
26. Fentiman IS, Millis R, Sexton S, et al. Pleural effusion in breast 49. Yano S, Herbst RS, Shinohara H, et al. Treatment for malig
cancer: a review of 1 0 5 cases. Cancer. 1 9 8 1 ;47:2087-2092. nant pleural effusion of human lung adenocarcinoma by inhi
27. Hirata T, Yonemori K, Hirakawa A, et al. Efficacy of pleurode bition of vascular endothelial growth factor receptor tyrosine
sis for malignant pleural effusions in breast cancer patients. kinase phosphorylation. Clin Cancer Res. 2000;6:957-96 5 .
Eur Respir]. 20 1 1 ;3 8 : 1 425-1430. 5 0 . Yano S, Shinohara H, Herbst R S , e t al. Production of experi
28. Das DK. Serous effusions in malignant lymphomas: a review. mental malignant pleural effusions is dependent on invasion
Diagn Cytopathol. 2006;34:33 5-347. of the pleura and expression of vascular endothelial growth
29. Romano M, Libshitz HI. Hodgkin disease and non-Hodgkin factor/vascular permeability factor by human lung cancer
lymphoma: plain chest radiographs and chest computed cells. Am ] Pathol. 2000; 1 57: 1 893- 1 903.
tomography of thoracic involvement in previously untreated 5 1 . Wang NS. The preformed stomas connecting the pleural cavity
patients. Radio/ Med (Torino). 1 998;95 :49-5 3 . and the lymphatics in the parietal pleura. Am Rev Respir Dis.
30. Tateishi U, Muller N L , Johkoh T, e t al. Primary mediastinal 1 975 ; 1 1 1 : 1 2-20.
lymphoma: characteristic features of the various hisrological 52. Light RW, Jenkinson SG, Minh V, et al. Observations on pleu
subtypes on CT. J Comput Assist Tomogr. 2004;28:782-789. ral pressures as fluid is withdrawn during thoracentesis. Am
3 1 . Vieta JO, Craver LE Intrathoracic manifestations of the lym Rev Respir Dis. 1 9 80; 1 2 1 :799-804.
phomatoid diseases. Radiology. 1 9 4 1 ;37: 1 3 8- 1 5 8 . 53. Rodriguez-Panadero F, Borderas Naranjo F, Lopez Mejias J.
32. Stolberg HO, Patt N L , MacEwen KF, e t al. Hodgkin's disease Pleural metastatic tumours and effusions. Frequency and
of the lung: roentgenologic-pathologic correlation. AJR Am J pathogenic mechanisms in a post-mortem series. Eur Respir].
Roentgenol. 1 964;92:96- 1 1 5 . 1 989;2:366-369.
3 3 . Berkman N , Breuer R, Kramer MR, e t al. Pulmonary involve 54. Mare! M , Stastny B, Melinova L, et al. Diagnosis of pleural
ment in lymphoma. Leuk Lymphoma. 1 996;20:229-237. effusions: experience with clinical studies, 1 9 86- 1 990. Chest.
34. Xaubet A, Diumenjo MC, Marin A, et al. Characteristics and 1 99 5 ; 1 07: 1 598-1 603.
prognostic value of pleural effusions in non-Hodgkin's lym 55. Maher GG, Berger HW. Massive pleural effusion: malignant
phomas. Eur J Respir Dis. 1 98 5 ;66: 1 3 5-140. and nonmalignant causes in 46 patients. Am Rev Respir Dis.
3 5 . Elis A, Blickstein D, Mulchanov I, et al. Pleural effusion in 1 972; 1 0 5 :458-460.
patients with non-Hodgkin's lymphoma: a case-controlled 56. Poree! JM, Vives M . Etiology and pleural fluid characteristics
study. Cancer. 1 998;83: 1 607- 1 6 1 1 . of large and massive effusions. Chest. 2003; 1 24:978-983.
36. Okada F, Ando Y, Kondo Y, et al. Thoracic CT findings of 57. Weick JK, Kiely JM, Harrison EG Jr, et al. Pleural effusion in
adult T-cell leukemia or lymphoma. AJR Am J Roentgenol. lymphoma. Cancer. 1 973;3 1 : 848-8 53.
2004; 1 82:76 1 -767. 5 8 . Bruneau R, Rubin P. The management of pleural effusions and
37. Brixey AG, Light RW Pleural effusions due to dasatinib. Curr chylothorax in lymphoma. Radiology. 1 965; 8 5 : 1 085-1 092.
Opin Pulm Med. 20 1 0 ; 1 6:36 1 -366. 5 9 . Celikoglu F, Teirstein AS, Krellenstein DJ, et al. Pleural effusion
38. Park BB, Ryoo BY, Lee JH, et al. Clinical features and treat in non-Hodgkin's lymphoma. Chest. 1 992; 1 0 1 : 1 357- 1 360.
ment outcomes of angioimmunoblastic T-cell lymphoma. 60. Yilmaz U, Polar G, Sahin N, et al. CT in differential diagnosis
Leuk Lymphoma. 2007;48:71 6-722. of benign and malignant pleural disease. Monaldi Arch Chest
39. Yamagata T, Okamoto Y, Yamagata Y, et al. Angioimmuno Dis. 2005;63: 1 7-22.
blastic lymphadenopathy with dysproteinaemia accompanied 6 1 . Trail ZC, Davies RJ, Gleeson FV. Thoracic computed tomog
by pleural effusion. Respirology. 2005 ; 1 0 : 1 24- 1 27. raphy in patients with suspected malignant pleural effusions.
40. Cullen MH, Stansfeld AG, Oliver RT, et al. Angioimmuno Clin Radio!. 200 1 ; 56: 1 93- 1 96.
blastic lymphadenopathy: report of ten cases and review of the 62. Light RW, MacGregor MI, Luchsinger PC, et al. Pleural effu
literature. QJ Med. 1 979; 1 8 1 : 1 5 1 - 1 77. sions: the diagnostic separation of transudates and exudates.
4 1 . Kamble R, Wilson C, Fassas A, et al. Malignant pleural effu Ann Intern Med. l 972;77:5 07-5 1 3 .
sion of multiple myeloma: prognostic factors and outcome. 6 3 . Assi Z , Caruso JL, Herndon J, e t al. Cytologically proved
Leuk Lymphoma. 2005 ;46: 1 1 37- 1 1 42. malignant pleural effusions: distribution of transudates and
42. Light RW, Hamm H . Malignant pleural effusion: would the exudates. Chest. 1 998; 1 1 3 : 1 302- 1 304.
real cause please stand up? Eur Respir]. 1 997; 1 0 : 1 7 0 1 - 1 702. 64. Ashchi M, Golish J, Eng P, et al. Transudative malignant
43. Meyer PC. Metastatic carcinoma of the pleura. Thorax. 1 966; pleural effusions: prevalence and mechanisms. South Med ].
2 1 :437-443. 1 998;9 l :23-22.
44. Rodriguez RM , Purnam JB, Glassfor D, et al. Predicting spon 65. Ferrer J, Roldan J, Teixidor J, et al. Predictors of pleural malig
taneous pleurodesis in patients with malignant pleural effu nancy in patients with pleural effusion undergoing thoracos
sion treated with an indwelling catheter. Am J Respir Crit Care copy. Chest. 200 5 ; 1 27: 1 0 1 7- 1 022.
Med. 1 99 8 ; 1 57:A6 5 . 66. Kalomenidis I, Light RW. Eosinophilic pleural effusions. Curr
4 5 . Leckie WJH , Tothill P. Albumin turnover in pleural effusions. Opin Pulm Med. 2003;9:254-260.
Clin Sci. 1 965 ;29:339-352. 67. Krenke R, Nasilowski J, Korczynski P, et al. Incidence and
46. Ishimoto 0 , Saij o Y, Narumi K, et al. High level of vascular etiology of eosinophilic pleural effusion. Eur Respir ]. 2009;
endothelial growth factor in hemorrhagic pleural effusion of 34: 1 1 1 1 - 1 1 1 7.
cancer. Oncology. 2002;63 :70-75. 68. Martinez-Moragon E, Aparicio J, Sanchis J, et al. Malignant
47. Cheng C-S, Rodriguez RM , Perkett EA, et al. Vascular pleural effusion: prognostic factors for survival and response
endothelial growth factor in pleural fluid. Chest. 1 999; 1 1 5 : to chemical pleurodesis in a series of 1 2 0 cases. Respiration.
760-765. 1 998;65: 1 08-1 1 3 .
C H APTE R 1 0 I PLE U RAL E F F U S I O N S R E LAT E D TO M ETASTATIC MALI G N A N C I E S 1 83
69. Light RW, Ball WC. Glucose and amylase in pleural effusions. 9 1 . Salyer WR , Eggleston J C , Erozan YS . Efficacy o f pleural
JAMA. 1 973;225:257-260. needle biopsy and pleural fluid cytopathology in the diag
70. Rodriguez-Panadero F, Lopez-Mejias J . Low glucose and nosis of malignant neoplasm involving the pleura. Chest.
pH levels in malignant pleural effusions. Am Rev Respir Dis. 1 975;67:536-539.
1 989; 1 39:663-667. 92. Frist B , Kahan AV, Koss LG. Comparisons of the diagnostic
7 1 . Heffner JE, Nietert PJ, Barbieri C. Pleural fluid pH as a pre values of biopsies of the pleura and cytologic evaluation of
dictor of survival for patients with malignant pleural effusions. pleural fluids. Am J Clin Pathol. l 979;72:48-5 1 .
Chest. 2000; 1 1 7:79-86. 93. Canto A, Rivas J , Saumench J , et al. Points to consider when
72. Good JT Jr, Taryle DA, Sahn SA. The pathogenesis of low choosing a biopsy method in cases of pleurisy of unknown
glucose, low pH malignant effusions. Am Rev Respir Dis. origin. Chest. 1 9 83;84: 1 76- 1 79.
1 9 8 5 ; 1 3 1 :737-74 1 . 94. Maskell N, Gleeson FV, Davies RJ. Standard pleural biopsy
73. Light RW, MacGregor MI, Ball WC Jr, et al. Diagnostic signif versus CT-guided cutting-needle biopsy for diagnosis of
icance of pleural fluid pH and Pco 2 • Chest. 1 973;64 : 5 9 1-596. malignant disease in pleural effusions: a randomised con
74. Sahn SA, Good JT Jr. Pleural fluid pH in malignant effusions. trolled trial. Lancet. 2003;36 1 : 1 326-1 330.
Ann Intern Med. 1 9 88; 1 08 : 345-349. 95. Poe RH, Ortiz C, Israel RH, et al. Sensitivity, specificity, and
75. Ende N. Studies of amylase activity in pleural effusions and predictive values of closed pleural biopsy. Arch Intern Med.
ascites. Cancer. 1 960; 1 3 :283-287. 1 984; 1 44:325-328 .
76. Kramer MR, Saidana MJ, Cepero RJ, et al. High amylase 9 6 . Hucker J, Bhatnagar NK, al-Jilaihawi AN, e t al. Thoraco
levels in neoplasm-related pleural effusion. Ann Intern Med. scopy in the diagnosis and management of recurrent pleural
1 9 89; 1 1 0: 567-5 69. effusions. Ann Thorac Surg. 1 9 9 1 ;52: 1 1 45-1 1 47.
77. Jarvi OH, Kunnas RJ, Laitio MT, et al. The accuracy and 97. Menzies R, Charbonneau M . Thoracoscopy for the diagnosis
significance of cytologic cancer diagnosis of pleural effusions. of pleural disease. Ann Intern Med. 1 99 1 ; 1 14:27 1 -276.
Acta Cytol. 1 972; 1 6 : 1 52- 1 57. 98. Antman KH. Clinical presentation and natural history
78. Grunze H . The comparative diagnostic accuracy, efficiency of benign and malignant mesothelioma. Semin Oncol.
and specificity of cytologic techniques used in the diagnosis 1 9 8 1 ;8 : 3 1 3-320.
of malignant neoplasm in serous effusions of the pleural and 99. Staats BA, Ellefson RD, Budahn LL, et al. The lipoprotein
pericardia! cavities. Acta Cytol. 1 964;8 : 1 50-1 64. profile of chylous and nonchylous pleural effusions. Mayo
79. Dekker A, Bupp PA. Cytology of serous effusions. An inves Clin Proc. 1 980;55:700-704.
tigation into the usefulness of cell blocks versus smears. Am J 1 00 . Bonnefoi H, Smith IE. How should cancer presenting as a
Clin Pathol. 1 978;70 : 8 5 5-860. malignant pleural effusion be managed? Br J Cancer. 1 996;
80. Prakash URS , Reiman HM. Comparison of needle biopsy 74:832-8 3 5 .
with cyrologic analysis for the evaluation of pleural effusion: 1 0 1 . Bielsa S, Salud A , Martinez M , e t al. Prognostic significance
analysis of 4 1 4 cases. Mayo Clin Proc. 1 9 8 5 ;60 : 1 58-1 64. of pleural fluid data in patients with malignant effusion. Eur
8 1 . Bueno CE, Clemente G, Castro BC, et al. Cytologic and bac J Intern Med. 2008 ; 1 9 :334-339.
teriologic analysis of fluid and pleural biopsy specimens with 1 02 . Burrows CM, Mathews WC, Colt HG. Predicting survival
Cope's needle. Arch Intern Med. 1 990; 1 5 0 : 1 1 90- 1 1 94. in patients with recurrent symptomatic malignant pleural
82. Naylor B, Schmidt RW. The case for exfoliative cytology of effusions: an assessment of the prognostic values of physi
serous effusions. Lancet. 1 964; 1 :7 1 1 -7 1 2 . ologic, morphologic, and quality of life measures of extent of
83. Ordonez N G . What are the current best immunohistochemi disease. Chest. 2000; 1 1 7:73-78.
cal markers for the diagnosis of epithelioid mesothelioma? 1 03 . Bielsa S, Martin-Juan J, Poree! JM, et al. Diagnostic and
A review and update. Hum Pathol. 2007;3 8 : 1 - 1 6. prognostic implications of pleural adhesions in malignant
84. Ordonez NG. The immunohistochemical diagnosis of meso effusions. ] Thorac Oncol. 2008;3 : 1 2 5 1 - 1 2 5 6 .
thelioma: a comparative study of epithelioid mesothelioma and 1 04. H s u I L , Su WC, Yan JJ, et al. Angiogenetic biomarkers i n
lung adenocarcinoma. Am J Surg Pathol. 2003;27: 1 0 3 1 - 1 0 5 1 . non-small cell lung cancer with malignant pleural effusion:
8 5 . Miedouge M , Rouzaud P, Salama G , et al. Evaluation o f seven correlations with patient survival and pleural effusion con
tumour markers in pleural fluid for the diagnosis of malignant trol. Lung Cancer. 2009;65 :37 1 -376.
effusions. Br J Cancer. 1 999;8 1 : 1 059- 1 06 5 . 1 0 5 . Tremblay A, Robbins S, Berthiaume L, et al. Natural hisrory
86. Lee YC, Knox B S , Garrett JE. Use of cytokeratin fragments 1 9 . 1 of asymptomatic pleural effusions in lung cancer patients.
and 1 9 . 2 1 (Cyfra 2 1 - 1 ) i n the differentiation of malignant and J Bronchol. 2007; 1 4 :98-100.
benign pleural effusions. Aust N ZJ Med. 1 999;29:765-769. 1 06. Spector M , Pollak JS. Management of malignant pleural
87. Lee YC, Chern JH, Lai SL, et al. Sialyl stage-specific embryonic effusions. Semin Respir Crit Care Med. 2008;29 :405-4 1 3 .
antigen- I : a useful marker for differentiating the etiology of 1 07. Jones SE, Durie BGM, Salmon S E . Combination che
pleural effusion. Chest. 1 998 ; 1 1 4: 1 542- 1 54 5 . motherapy with adriamycin and cyclophosphamide for
8 8 . Poree! J M , Vives M, Esquerda A , e t al. Use o f a panel of tumor advanced breast cancer. Cancer. 1 975;36:90-97.
markers (carcinoembryonic antigen, cancer antigen 1 2 5 , car 1 08 . Livingsron RB, McCracken JD, Trauth CJ, et al. Isolated
bohydrate antigen 1 5-3, and cytokeratin 1 9 fragments) in pleural effusion in small cell lung carcinoma: favorable prog
pleural fluid for the differential diagnosis of benign and malig nosis. Chest. 1 9 82;8 1 :208-2 1 1 .
nant effusions. Chest. 2004; 1 26: 1 757- 1 763. 1 09 . Albain KS, Crowley JJ, LeBlanc M , et al. Determinants of
89. Light RW. Tumor markers in undiagnosed pleural effusions. improved outcome in small-cell lung cancer: an analysis of the
Chest. 2004; 1 26: 1 72 1 - 1 722. 2,580-patient Southwest Oncology Group data base. J Clin
90. Moriarty AT, Wiersema L, Snyder W, et al. Immunopheno Oncol. 1 990;8 : 1 563-1 574.
typing of cytologic specimens by flow cytometry. Diagn Cyto 1 1 0 . Yano S, Matsumori Y, Ikuta K, et al. Current status and
pathol. l 993;9:252-25 8 . perspective of angiogenesis and antivascular therapeutic
1 84 PLE U RAL D I S EASES
strategy: non-small cell lung cancer. Int J Clin Oneal. 2006; 129. Warren WH , Kalimi R, Khodadadian LM, et al. Management
1 1 :73-8 1 . of malignant pleural effusions using the Pleur(x) catheter.
1 1 1. Guo YB , Kalomenidis I , Hawthorne M , et al. Pleurodesis Ann Thorac Surg. 2008 ;8 5 : 1 049-1 0 5 5 .
is inhibited by anti-vascular endothelial growth factor anti 1 30. Van Meter M E , McKee KY, Kohlwes RJ. Efficacy and
body. Chest. 200 5 ; 1 28 : 1 790- 1 797. safery of tunneled pleural catheters in adults with malignant
1 1 2. Herrstedt J, Clementsen P, Hansen OP. Increased myelosup pleural effusions: a systematic review. J Gen Intern Med.
pression during cytostatic treatment and pleural effusion 2 0 1 0;26:70-76.
in patients with small cell lung cancer. Eur J Cancer. 1 992; 1 3 1 . Janes SM, Rahman NM, Davies RJ, et al. Catheter-tract
28A: l 070-1 073. metastases associated with chronic indwelling pleural cath
1 1 3. Li J, Gwilt P. The effect of malignant effusions on metho eters. Chest. 2007; 1 3 1 : 1 232- 1 234.
trexate disposition. Cancer Chemother Pharmacol. 2002; 5 0 : 1 32. Morel A, Mishra E, Medley L, et al. Chemotherapy should
373-382. not be withheld from patients with an indwelling pleural
1 1 4. Oickgreber NJ, Sorensen JB, Paz-Ares LG, et al. Pemetrexed catheter for malignant pleural effusion. Thorax. 20 1 1 ;66:
safery and pharmacokinetics in patients with third-space 448-449.
fluid. Clin Cancer Res. 20 1 O ; 1 6 :2872-2880. 1 33 . Chee A, Tremblay A. The use of tunneled pleural catheters
1 1 5. Su WC, Lai WW, Chen HH, et al. Combined intrapleural in the treatment of pleural effusions. Curr Opin Pulm Med.
and intravenous chemotherapy, and pulmonary irradiation, 20 1 1 ; 1 7:237-24 1 .
for treatment of patients with lung cancer presenting with 1 34. Fysh ET, Wrightson JM, Lee YC, et al. Fractured indwelling
malignant pleural effusion: a pilot study. Oncology. 2003; pleural catheters. Chest. 20 1 2; 1 4 1 : 1 090- 1 094.
64: 1 8-24. 1 3 5 . Shinto RA, Stansbury OW, Brown SE, et al. Does therapeu
1 1 6. Ren S, Terman OS, Bohach G, et al. lntrapleural staphylo tic thoracentesis improve the exercise capaciry of patients
coccal superantigen induces resolution of malignant pleural with pleural effusion? Am Rev Respir Dis. 1 987; 1 3 5 :A244.
effusions and a survival benefit in non-small cell lung cancer. 1 36. Shinto RA, Stansbury OW, Fischer CE, et al. The effect of
Chest. 2004; 1 26: 1 529- 1 539. thoracentesis on central respiratory drive in patients with
1 1 7. Schmidt HH, Renner H, Linkesch W. !ntrapleural instilla large pleural effusions. Am Rev Respir Dis. 1 9 88; 1 37:Al 1 2 .
tion of rituximab for the treatment of malignant pleural effu 1 37. Sanchez-Armengol A , Rodriguez-Panadero F. Survival and
sions in NHL. Haematologica. 2004;89:ECR39. talc pleurodesis in metastatic pleural carcinoma, revisited.
1 1 8. Sartori S, Trevisani L, Nielsen I, et al. Intracavitary bleomy Report of 1 2 5 cases. Chest. 1 993; 1 04 : 1 482- 1 4 8 5 .
cin vs interferon in the management of malignant pleural 1 3 8 . Lan RS, L o S K , Chuang M L , e t al. Elastance o f the pleural
effusions. Chest. 1 998 ; 1 1 3 : 1 1 45-1 146. space: a predictor for the outcome of pleurodesis in patients
1 1 9. Lissoni P, Barni S, Tancini G, et al. !ntracavitary therapy with malignant pleural effusion. Ann Intern Med. 1 997;
of neoplastic effusions with cytokines: comparison among 1 26:768-774 .
interferon alpha, beta and interleukin-2. Support Care Can 1 39. Bielsa S, Hernandez P, Rodriguez-Panadero F, et al. Tumor
cer. 1 995;3:78-80. rype influences the effectiveness of pleurodesis in malignant
120. Ishida A, Miyazawa T, Miyazu Y, et al. Intrapleural cisplatin effusions. Lung. 2 0 1 1 ; 1 89: 1 5 1 - 1 5 5 .
and OK432 therapy for malignant pleural effusion caused by 1 40. Weisberger AS, Levine B, Storaasli JP. Use o f nitrogen mus
non-small cell lung cancer. Respirology. 2006; 1 1 :90-97. tard in treatment of serous effusions of neoplastic origin.
121. Sebastian M, Kiewe P, Schuette W, et al. Treatment of malig JAMA. 1 9 5 5 ; 1 5 9 : 1 704-1 706.
nant pleural effusion with the trifunctional antibody Catu 1 4 1 . Ariel IM, Oropeza R, Pack GT. lntracavitary administra
maxomab (Removab) (Anti-EpCAMxAnti-CD3) : results of tion of radioactive isotopes in the control of effusions due to
a phase 1 /2 study. ] Immunother. 2009;32 : 1 9 5-202. cancer. Cancer. 1 966; 1 9 : 1 096-1 1 0 1 .
1 22. Roy PH, Carr OT, Payne WS . The problem of chylothorax. 1 42. Austin EH, Flye MW. Th e treatment o f recurrent malignant
Mayo Clin Proc. 1 967;42:457-467. pleural effusion. Ann Thorac Surg. 1 979;28: 1 90-203.
1 23 . Putnam JB Jr, Light RW, Rodriguez RM , et al. A randomized 1 43 . Izbicki R, Weyhing BT, Baker L, et al. Pleural effusion in
comparison of indwelling pleural catheter and doxycycline cancer patients: a prospective randomized study of pleural
pleurodesis in the management of malignant pleural effu drainage with the addition of radioactive phosphorous to
sions. Cancer. 1 999;86: 1 992- 1 999. the pleural space vs. pleural drainage alone. Cancer. 1 975;36:
1 24. Putnam JB Jr, Walsh GL, Swisher SG, et al. Outpatient 1 5 1 1-1 5 1 8.
management of malignant pleural effusion by a chronic 1 44. Adler RH, Sayek I. Treatment of malignant pleural effusion:
indwelling pleural catheter. Ann Thorac Surg. 2000;69: a method using tube thoracostomy and talc. Ann Thorac
369-375. Surg. 1 976;22:8- 1 5 .
125. Tremblay A, Michaud G . Single-center experience with 250 1 4 5 . Rubinson RM , Bolooki H . Intrapleural tetracycline for
tunnelled pleural catheter insertions for malignant pleural control of malignant pleural effusion: a preliminary report.
effusion. Chest. 2006; 129 :362-368. South MedJ 1 972;65: 847-849.
1 26. Brant A, Eaton T. Serious complications with talc slurry 1 46. Dollinger MR, Krakoff IH, Karnofsky DA. Quinacrine
pleurodesis. Respirology. 200 1 ;6: 1 8 1 - 1 8 5 . (Atabrine) in the treatment of neoplastic effusions. Ann
1 27. Al-Halfawy A , Light R . Safery and efficacy o f using a surgivac Intern Med. 1 967;66:249-257.
pump for the drainage of chronic indwelling pleural cath 1 47. Xie C, Teixeira LR, Wang N-S, et al. Serial observations after
eters in malignant pleural effusions. Respirology. 2008; 1 3 : high dose talc slurry in the rabbit model for pleurodesis.
461-464. Lung. 1 998; 1 76:299-307.
128. Musani AI, Haas AR, Seijo L, et al. Outpatient management 148. Wu W, Teixeira LR, Light RW. Doxycycline pleurodesis in
of malignant pleural effusions with small-bore tunneled rabbits. Comparison of results with and without chest tube.
pleural catheters. Respiration. 2004;7 1 : 5 5 9-566. Chest. 1 998; 1 1 4: 563-5 68.
C H APTE R 1 0 I PLE U RAL E F F U S I O N S R E LAT E D TO M ETASTATIC MALI G N A N C I E S 1 85
1 49. Sahn SA, Good JT. The effect of common sclerosing agents Pleural Disease Guideline 20 1 0 . Thorax. 2 0 1 0;65(suppl 2) :
on the rabbit pleural space. Am Rev Respir Dis. 1 98 1 ; 1 24: ii32-ii40.
65-67. 1 69. Bethune N. Pleural poudrage: new technique for deliberate
1 5 0. Vargas FS, Teixeira LR, Antonangelo L, et al. Acute and production of pleural adhesions as preliminary to lobecromy.
chronic pleural changes after the intrapleural instillation of ] Thorac Cardiovasc Surg. 1 93 5 ;4:25 1 -26 1 .
mitoxantrone in rabbits. Lung. 1 99 8 ; 1 76:227-236. 1 70. Light RW, Wang N-S, Sassoon CSH, et al. Talc slurry is
1 5 1 . Light RW, Cheng D-S, Lee YC, et al. A single intrapleural an effective pleural sclerosant in rabbits. Chest. 1 99 5 ; 1 07:
injection of transforming growth factor-2 produces excel 1 702- 1 706.
lent pleurodesis in rabbits. Am J Respir Crit Care Med. 2000; 171. Bresticker MA, ObaJ, LoCicero J III, et al. Optimal pleurode
1 62:98-1 04. sis: a comparison study. Ann Thorac Surg. 1 993;55 :364-366.
1 52. Xie C, Teixeira LR, McGovern JP, et al. Systemic cortico 1 72. Jerram RM, Fossum Tw; Berridge BR, et al. The efficacy of
steroids decrease the effectiveness of talc pleurodesis. Am J mechanical abrasion and talc slurry as methods of pleurode
Respir Crit Care Med. 1 998; 1 57 : 1 44 1 - 1 444. sis in normal dogs. Vet Surg. l 999;28:322-332.
1 53 . Cheng D-S, Rogers J, Wheeler A, et al. The effects of intra 1 73 . Noppen M, Degreve J, Mignolet M, et al. A prospective,
pleural polyclonal anti-tumor necrosis factor alpha (TNF) fab randomized study comparing the efficacy of talc slurry and
fragments on pleurodesis in rabbits. Lung. 2000; 1 7 8 : 1 9-30. bleomycin in the treatment of malignant pleural effusions.
1 54 . Strange C, Baumann MH, Sahn SA, et al. Effects of intra Acta Clin Belg. 1 997;52:258-262.
pleural heparin or urokinase on the extent of tetracycline 1 74. Webb WR, Ozmen V, Moulder PV, et al. Iodized talc
induced pleural disease. Am ] Respir Crit Care Med. 1 995; pleurodesis for the treatment of pleural effusions. J Thorac
1 5 1 : 508-5 1 5 . Cardiovasc Surg. 1 992; 1 03 : 8 8 1-886.
1 5 5 . Rodriguez-Panadero F, Segado A , Martin Juan ], e t al. Failure 175. Kennedy L, Rusch VW, Strange C, et al. Pleurodesis using
of talc pleurodesis is associated with increased pleural fibri talc slurry. Chest. 1 994; 1 06:342-346.
nolysis. Am ] Respir Crit Care Med. 1 995; 1 5 1 :78 5-790. 1 76. Dresler CM, Olak J, Herndon JE II, et al. Phase III inter
1 56 . Grande J P. Role of transforming growth factor- 13 in tissue group study of talc poudrage vs talc slurry sclerosis for malig
injury and repair. Proc Soc Exp Biol Med. 1 997;2 1 4:27-40. nant pleural effusion. Chest. 2005 ; 1 27:909-9 1 5 .
1 57. Sime PJ, Xing Z, Graham FL, et al. Adenovector-medi 1 77. Haddad FJ, Younes RN, Gross JL, e t al. Pleurodesis in
ated gene transfer of active transforming growth factor-/3 1 patients with malignant pleural effusions: talc slurry or bleo
induces prolonged severe fibrosis in rat lung. ] Clin Invest. mycin? Results of a prospective randomized trial. World ]
1 997; 1 00:768-776. Surg. 2004;28:749-752.
1 5 8 . Perkett EA. Role of growth factors in lung repair and dis 178. Aydogmus U, Ozdemir S, Cansever L, et al. Bedside talc
eases. Curr Opin Pediatr 1 995;7:242-249. pleurodesis for malignant pleural effusion: factors affecting
1 5 9. Idell S, Zwieb C, Kumar A, et al. Pathways of fibrin turnover success. Ann Surg Oneal. 2009; 1 6:745-750.
of human pleural mesothelial cells in vitro. Am J Respir Cell 1 79. Rehse DH, Aye RW, Florence MG. Respiratory failure
Mot Biol. 1 992;7:4 1 4-426. following talc pleurodesis. Am ] Surg. 1 999; 1 77:437-440.
1 60. Lee YCG, Lane KB, Parker RE, et al. Transforming growth 1 80. Weissberg D, Ben-Zeev I . Talc pleurodesis. Experience
factor (TGF)-/3 2 produces effective pleurodesis in sheep with with 360 patients. J Thorac Cardiovasc Surgery. 1 993; 1 06:
no systemic complications. Thorax. 2000; 5 5 : 1 0 5 8- 1 062. 689-69 5 .
1 6 1 . Lee YCG, Yasay JR, Johnson JE, et al. Comparing transform 181. Campos JR, Werebe EC, Vargas F S , e t al. Respiratory failure
ing growth factor-/3 , , talc and bleomycin as pleurodesing due to insuffiated talc. Lancet. 1 997;349:25 1 -252.
agents in sheep. Respirology. 2002;7:209-2 1 6 . 1 82. Montes JF, Ferrer J, Villarino MA, et al. Influence of talc
1 62. Lee YCG, Teixeira LR, Devin CJ, e t al. Transforming growth dose on extrapleural talc dissemination afrer talc pleurodesis.
factor-beta(2) induces pleurodesis significantly faster than Am ] Respir Crit Care Med. 2003 ; 1 68:348-3 5 5 .
talc. Am ] Respir Crit Care Med. 200 1 ; 1 63:640-644 . 1 83 . Schulze M, Boehle AS, Kurdow R , e t al. Effective treat
1 63 . Lee YCG, Devin CJ, Teixeira LR, et al. Transforming growth ment of malignant pleural effusion by minimal invasive
factor /3 2 induced pleurodesis is not inhibited by corticoste thoracic surgery: thoracoscopic talc pleurodesis and pleuro
roids. Thorax. 200 1 ;56:643-648 . peritoneal shunts in 1 0 1 patients. Ann Thorac Surg. 200 1 ;
1 64. Lee YCG, Malkerneker D, Thompson PJ, e t al. Transforming 7 1 : 1 809- 1 8 1 2 .
growth factor /3 induces vascular endothelial growth factor 1 84. Janssen JP, Collier G, Asroul P, e t al. Safety of pleurodesis
elaboration from pleural mesothelial cells in vivo and in vitro. with talc poudrage in malignant pleural effusion: a prospec
Am ] Respir Crit Care Med. 2002; 1 6 5 : 88-94. tive cohort study. Lancet. 2007;369 : 1 5 3 5- 1 539.
1 6 5 . Cheng D-S, Lee YC, Rogers JT, et al. Vascular endothelial 185. Rinaldo JE, Owens GR, Rogers RM. Adult respiratory distress
growth factor level correlates with transforming growth syndrome following intrapleural instillation of talc. J Thorac
factor-/3 isoform levels in pleural effusions. Chest. 2000; Cardiovasc Surg. 1 983;85: 523-526.
1 1 8 : 1 747- 1 753. 1 86. de Campos JR, Vargas FS, de Campos Werebe E, et al. Tho
1 66. Lee YC, Baumann MH, Maskell NA, et al. Pleurodesis racoscopic talc poudrage: 15 years experience. Chest. 200 1 ;
practice for malignant pleural effusions in five English 1 1 9 : 8 0 1 -806.
speaking countries: survey of pulmonologists. Chest. 2003; 1 87. Maskell NA, Lee YC, Gleeson FV, et al. Randomised tri
1 24:2229-2238. als describing lung inflammation after pleurodesis with talc
1 67. Burgers JA, Kunst PW, Koolen MG, et al. Pleural drainage of varying particle size. Am ] Respir Crit Care Med. 2004;
and pleurodesis; implementation of the Dutch guidelines in 1 70 :377-3 82.
four hospitals. Eur Respir]. 2008; 1 7:237-24 1 . 1 88. Ferrer J, Villarino MA, Tura JM, et al. Talc preparations
1 68 . Roberts ME, Neville E , Berrisford RG, et al.Management used for pleurodesis vary markedly from one preparation to
of a malignant pleural effusion: British Thoracic Society another. Chest. 200 1 ; 1 1 9 : 1 90 1 - 1 90 5 .
1 86 PLE U RAL D I S EASES
1 8 9. Werebe EC, Pazetti R, De Campos JRM, et al. Systemic malignant pleural effusions. Am J Resp Dis Crit Care Med.
distribution of talc after intrapleural administration in rats. 2000 ; 1 62: 1 445-1 449.
Chest. 1 999; 1 1 5 : 1 90- 1 93. 2 1 0. Vargas FS, Wang NS, Despars JA, et al. Effectiveness of
1 90. Kennedy L, Harley RA, Sahn SA, et al. Talc slurry pleurode bleomycin in comparison to tetracycline as pleural sclerosing
sis. Pleural fluid and hisrologic analysis. Chest. 1 99 5 ; 1 07: agent in rabbits. Chest. 1 993; 1 04: 1 5 82- 1 5 84.
1 707- 1 7 1 2 . 211. Sartori S, Tassinari D, Ceccotti P, et al. Prospective random
1 9 1 . Ferrer J, Montes JF, Villarino MA, e t al. Influence o f par ized trial of intrapleural bleomycin versus interferon alfa-2b
ticle size on extrapleural talc dissemination after talc slurry via ultrasound-guided small-bore chest tube in the pallia
pleurodesis. Chest. 2002 ; 1 22: 1 0 1 8- 1 027. tive treatment of malignant pleural effusions. J Clin Oncol.
1 92. Rossi VF, Vargas FS, Marchi E, et al. Acute inflammatory 2004;22: 1 228-1 233.
response secondary ro intrapleural administration of two 2 1 2. Leininger BJ, Barker WL, Lanstron HT. A simplified method
types of talc. Eur Respir]. 2 0 1 0;35 :396-40 1 . for management of malignant pleural effusion. J Thoracic
1 93 . Montes-Worboys A, Rodriguez-Portal JA, Arellano-Orden E, Cardiovasc Surg. 1 969;58:758-763 .
et al. Interleukin-8 activates coagulation and correlates with 213. Kinsey D L , Carter D, Klassen KP. Simplified management
survival after talc pleurodesis. Eur Respir]. 20 1 0 ; 3 5 : 1 60- 1 66. of malignant pleural effusion. Arch Surg. l 964; 89:3 89-39 1 .
1 94. Nguyen NC, Tran I, Hueser CN, et al. F- 1 8 FDG PET/CT 2 1 4. Marchi E , Vargas FS, Teixeira LR, et al. Comparison of
Characterization of talc pleurodesis-induced pleural changes nitrogen mustard, cytarabine and dacarbazine as pleural
over time: a retrospective study. Clin Nucl Med. 2009; sclerosing agents in rabbits. Eur Respir]. 1 997; 1 0 : 5 9 8-602.
34:886-890. 215. Barbetakis N, Anroniadis T, Tsilikas C. Results of chemical
195. Sherman S, Grady KJ, Seidmen JC. Clinical experience with pleurodesis with mitoxantrone in malignant pleural effusion
tetracycline pleurodesis of malignant pleural effusions. South from breast cancer. World] Surg Oncol. 2004;2: 1 6 .
Med]. 1 987;80:7 1 6-7 1 9 . 216. Morales M, Exposiro M C . Intrapleural mitoxantrone fo r the
1 96. Walker-Renard P B , Vaughan LM, Sahn SA. Chemical palliative treatment of malignant pleural effusions. Support
pleurodesis for malignant pleural effusions. Ann Intern Med. Care Cancer. 1 995;3: 1 47- 1 49.
1 994; 1 2 0 : 5 6-64. 2 1 7. Maiche AG, Virkkunen P, Kontkanen T, et al. Bleomycin and
1 97. Costabel U. Adieu, tetracycline pleurodesis; (but not in miroxantrone in the treatment of malignant pleural effusions.
Germany) . Chest. 1 993; 1 03:984. Am ] Clin Oncol. 1 993; 1 6 : 5 0-5 3.
198. Bilaceroglu S, Guo Y, Hawthorne ML, et al. Oral forms 218. Seto T, Ushijima S, Yamamoto H, et al. Intrapleural hypo
of tetracycline and doxycycline are effective in producing ronic cisplatin treatment for malignant pleural effusion
pleurodesis. Chest. 2005 ; 1 28 :3750-3756. in 80 patients with non-small-cell lung cancer: a multi
1 99. Light RW, Wang NS, Sassoon CS, et al. Comparison of institutional phase II trial. Br] Cancer. 2006;9 5 : 7 1 7-72 1 .
the effectiveness of tetracycline and minocycline as pleural 2 1 9. Vargas FS, Carmo AO, Teixeira LR. A new look at old agents
sclerosing agents in rabbits. Chest. 1 994; 1 06:577-582. for pleurodesis. Nitrogen mustard, sodium hydroxide and
200. Robinson LA, Fleming WH, Galbraith TA. Intrapleural dox silver nitrate. Curr Opin Pulm Med. 2000;6:28 1 -286.
ycycline control of malignant pleural effusions. Ann Thorac 220. Vargas FS, Teixeira LR, Silva LMMF, et al. Comparison of
Surg. 1 993;5 5 : 1 1 1 5- 1 1 2 1 . silver nitrate and tetracycline as pleural sclerosing agents in
20 1 . Heffner JE, Standerfer RJ, Torsrveit J , e t al. Clinical efficacy rabbits. Chest. 1 99 5 ; 1 0 8 : 1 080- 1 083.
of doxycycline for pleurodesis. Chest. 1 994; 1 0 5 : 1 743-1 747. 22 1 . Vargas FS, Teixeira LR, Vaz MAC, et al. Silver nitrate is supe
202. Mansson T. Treatment of malignant pleural effusion with rior to talc slurry in producing pleurodesis in rabbits. Chest.
doxycycline. Scand] Infect Dis Suppl. 1 9 88;53 :29-34. 2000; 1 1 8 : 808-8 1 3 .
203. Seaton KG, Patz EF Jr, Goodman PC. Palliative treatment of 222. Paschoalini Mda S, Vargas F S , e t al. Prospective randomized
malignant pleural effusions: value of small-bore catheter tho trial of silver nitrate vs talc slurry in pleurodesis for sympto
racosromy and doxycycline sclerotherapy. A]R Am J Roent matic malignant pleural effusions. Chest. 2005 ; 1 28 : 684-689.
genol. 1 99 5 ; 1 64 : 5 89-5 9 1 . 223. Terra RM, Kim SY, Pego-Fernandes PM, et al. Is silver nitrate
204. Pulsiripunya C , Youngchaiyud P, Pushpakom R, e t al. Th e effi pleurodesis for patients with malignant pleural effusion
cacy of doxycycline as a pleural sclerosing agent in malignant feasible and safe when performed in an outpatient setting?
pleural effusion: a prospective study. Respirology. 1 996; 1 :69-72. Ann Surg Oncol. 20 1 1 ; 1 8 : 1 145-1 1 5 0 .
205. Peng M-J, Kuo H-T, Chen P-J, et al. Minocycline pleurode 224. Olivares-Torres CA, Laniado-Laborin R , Chavez-Garcia C,
sis for malignant pleural effusions. Thorac Med. 1 99 5 ; 1 0 : et al. Iodopovidone pleurodesis for recurrent pleural effu
243-248 . sion. Chest. 2002; 1 22: 5 8 1 -5 8 3 .
2 0 6 . Light RW, O'Hara V S , Moritz T E , e t a l . Intrapleural tetracy 225. Agarwal R , Aggarwal AN, Gupta D . Efficacy and safety o f
cline for the prevention of recurrent spontaneous pneumo iodopovidone pleurodesis through tube thoracosromy. Respi
thorax. JAMA. l 990;264:2224-2230. rology. 2006; 1 1 : 1 05-1 08.
207. Dikensoy 0, Zhu Z, Donnelly E, et al. Combination 226. Nero JD, de Oliveira SF, Vianna SP, et al. Efficacy and safety
therapy with intrapleural doxycycline and talc in reduced of iodopovidone pleurodesis in malignant pleural effusions.
doses is effective in producing pleurodesis in rabbits. Chest. Respirology. 20 1 O; 1 5 : 1 1 5- 1 1 8 .
2005; 1 2 8 : 373 5-3742. 227. Mohsen TA, Zeid AA, Meshref M, et al. Local iodine
208. Ruckdeschel JC, Moores D, Lee JY, et al. Intrapleural therapy pleurodesis versus thoracoscopic talc insuffiation in recurrent
for malignant pleural effusions. A randomized comparison of malignant pleural effusion: a prospective randomized control
bleomycin and tetracycline. Chest. 1 99 1 ; 1 00 : 1 528- 1 5 3 5 . trial. EurJ Cardiothorac Surg. 2 0 1 1 ;40:282-286.
2 0 9 . Diacon AH, Wyser C , Bollinger CT, e t al. Prospective ran 228. Guo Y, Tang K, Bilaceroglu S, et al. Iodopovidone is as
domized comparison of thoracoscopic talc poudrage under effective as doxycycline in producing pleurodesis in rabbits.
local anaesthesia vs. bleomycin instillation for pleurodesis in Respirology. 20 1 0; 1 5 : 1 1 9- 1 2 5 .
C H APTE R 1 0 I PLE U RAL E F F U S I O N S R E LAT E D TO M ETASTATIC MALI G N A N C I E S 1 87
229. Wagenfeld L, Zeitz 0, Richard G. Visual loss after povidone 249 . Landvater L, Hix WR, Mills M, et al. Malignant pleural effu
iodine pleurodesis. N Engl] Med. 2007;3 57: 1 264- 1 265. sion treated by tetracycline sclerotherapy: a comparison of
230. Kataoka M, Morishita R, Hiramatsu J, et al. OK-432 single vs repeated instillation. Chest. 1 988;93: 1 1 96-1 1 98 .
induces production of neutrophil chemotactic factors in 2 5 0 . Reddy C, Ernst A , Lamb C, e t al. Rapid pleurodesis fo r
malignant pleural effusion. Intern Med. 1 99 5 ;34:3 52-3 56. malignant pleural effusions: a pilot study. Chest. 20 1 1 ; 1 3 9:
23 1 . Kishi K, Homma S , Sakamoto S, et al. Efficacious pleurode 1 4 1 9- 1 423.
sis with OK-432 and doxorubicin against malignant pleural 25 1 . Tassi GF, Cardillo G, Marchetti GP, et al. Diagnostic and
effusions. Eur Respir J 2004;24:263-266. therapeutical management of malignant pleural effusion.
232. Dikensoy 0, Light RW. Alternative widely available, inex Ann Oncol. 2006; 1 7 (suppl l ) :iil l -ii l 2 .
pensive agents for pleurodesis. Curr Opin Pulm Med. 2005; 2 5 2 . Yim A P, Chan AT, Lee T w, e t al. Thoracoscopic talc insuf
1 1 :340-344. flation versus talc slurry for symptomatic malignant pleural
233. Ukale V, Agrenius V, Hillerdal G, et al. Pleurodesis in recur effusion. Ann Thorac Surg. 1 9 96;62: 1 65 5 - 1 6 5 8 .
rent pleural effusions: a randomized comparison of a classical 253. Groth G, Gatzemeier U, Haubingen K , e t al. Intrapleural
and a currently popular drug. Lung Cancer. 2004;43 :323-328. palliative treatment of MPEs with mitoxantrone versus pla
234. Ukale V, Agrenius V, Widstrom 0, et al. Inflammatory cebo (pleural rube alone) . Ann Oncol. 1 9 9 1 ;2 :2 1 3-2 1 5 .
parameters after pleurodesis in recurrent malignant pleu 254. Sorensen PG, Svendsen TL, Enk B . Treatment o f MPE with
ral effusions and their predictive value. Respir Med. 2004; drainage, with and without instillation of talc. Eur J Respir
98: 1 1 66- 1 1 72. Dis. 1 9 84;65 : 1 3 1 - 1 3 5 .
235. Clementsen P, Evald T, Grode G, et al. Treatment of malig 2 5 5 . Kolschmann S, Ballin A , Gillissen A . Clinical efficacy and
nant pleural effusion: pleurodesis using a small percutaneous safety of thoracoscopic talc pleurodesis in malignant pleural
catheter. A prospective randomized study. Respir Med. 1 998; effusions. Chest. 2005 ; 1 2 8 : 1 4 3 1 - 1 4 3 5 .
92:5 93-596. 256. Arapis K , Caliandro R , Stern JB, e t a l . Thoracoscopic pallia
236. Caglayan B, To run E, Turan D, et al. Efficacy ofiodopovidone tive treatment of malignant pleural effusions: results in 273
pleurodesis and comparison of small-bore catheter versus patients. Surg Endosc. 2006;20 : 9 1 9-923.
large-bore chest rube. Ann Surg Oncol. 2008 ; 1 5 :25 94-2599. 257. Love D, White D, Kiroff G . Thoracoscopic talc pleurodesis
237. Parulekar W, Di Primio G, Matzinger F, et al. Use of small for malignant pleural effusion. ANZJ Surg. 2003;73: 1 9-22 .
bore vs large-bore chest rubes for treatment of malignant 2 5 8 . Froudarakis ME, Klimathianaki M, Pougounias M. Systemic
pleural effusions. Chest. 200 1 ; 120: 1 9-25 . inflammatory reaction after thoracoscopic talc poudrage.
2 3 8 . Light RW Pleural controversy: optimal chest rube size fo r Chest. 2006; 1 29:356-36 1 .
drainage. Respirology. 2 0 1 1 ; 1 6 :244-248. 259. Kuzniar TJ, Blum MG, Kasibowska-Kuzniar K, et al. Predic
239. Teixeira LR, Wu W, Chang DS, et al. The effect of cortico tors of acute lung injury and severe hypoxemia in patients
steroids on pleurodesis induced by doxycycline in rabbits. undergoing operative talc pleurodesis. Ann Thorac Surg.
Chest. 2002; 1 2 1 :2 1 6-2 1 9 . 2006;82: 1 976- 1 98 1 .
240. Ors Kaya S, Bir F, Atalay H, e t al. Effect o f diclofenac on 260. Crnjac A. The significance of thoracoscopic mechanical
experimental pleurodesis induced by tetracycline in rabbits. pleurodesis for the treatment of malignant pleural effusions.
J Investig Med. 2005;53:267-270. Wien Klin Wochenschr. 2004; 1 1 6 (suppl 2):28-32.
24 1 . Xie C, Teixeira LR, McGovern JP, et al. Systemic cortico 26 1 . Crnjac A, Sok M, Kamenik M . Impact of pleural effusion
steroids decrease the effectiveness of talc pleurodesis. Am J pH on the efficacy of thoracoscopic mechanical pleurodesis
Respir Crit Care Med. 1 998; 1 57: 1 44 1 - 1 444. in patients with breast carcinoma. Eur J Cardiothorac Surg.
242. Lardinois D, Vogt P, Yang L, et al. Non-steroidal anti 2004;26:432-436.
inflammatory drugs decrease the quality of pleurodesis after 262. Akopov AL, Egorov VI, Varlamov VV, et al. Thoracoscopic
mechanical pleural abrasion. Eur] Cardiothorac Surg. 2004; collagen pleurodesis in the treatment of malignant pleural
2 5 : 8 6 5-87 1 . effusions. Eur] Cardiothorac Surg. 2005;28:750-753.
243 . Villanueva AG, Gray AW Jr, Shahian DM, et al. Efficacy 263. Maceachern P, Tremblay A. Pleural controversy: pleurodesis
of short term versus long term tube thoracostomy drainage versus indwelling pleural catheters for malignant effusions.
before tetracycline pleurodesis in the treatment of malignant Respirology. 20 1 1 :747-754.
pleural effusions. Thorax. 1 9 94;49:23-2 5 . 264. Fysh ET, Waterer GW, Kendall P et al. Indwelling pleural
244. H s u L H , Soong TC, Feng AC, e t al. Intrapleural urokinase catheters reduce inpatient days over pleurodesis for malig
for the treatment of loculated malignant pleural effusions nant pleural effusion. Chest. 20 1 2 ; 1 42:394-400.
and trapped lungs in medically inoperable cancer patients. 265. Hunt BM, Farivar AS, Vallieres E, et al. Thoracoscopic talc
] Thorac Oncol. 2006; 1 :460-467. versus tunneled pleural catheters for palliation of malignant
245 . Sherman S, Ravikrishnan KP, Patel AS, et al. Optimum anes pleural effusions. Ann Thorac Surg. 2 0 1 2;94: 1 0 53- 1 0 5 9 .
thesia with intrapleural lidocaine during chemical pleurode 2 6 6 . Efthymiou CA , Masoudi T I , Thorpe JA, e t al. Malignant
sis with tetracycline. Chest. 1 988;93: 533-536. pleural effusion in the presence of trapped lung. Five-year
246 . Vargas FS, Teixeira LR, Coelho IJC, et al. Distribution of experience of PleurX tunnelled catheters. Interact Cardiovasc
pleural injectate: effect of volume of injectate and animal Thorac Surg. 2009;9:96 1-964.
rotation. Chest. 1 994; 1 06: 1 246- 1 249. 267. Thornton RH, Miller Z, Covey AM, et al. Tunneled pleural
247. Lorch DG, Gordon L, Wooten S, et al. Effect of patient posi catheters for treatment of recurrent malignant pleural effu
tioning on distribution of tetracycline in the pleural space sion following failed pleurodesis. ] Vase Interv Radio!. 20 I O ;
during pleurodesis. Chest. 1 988;93: 527-529. 2 1 : 696-700.
248. Dryzer SR, Allen ML, Strange C, et al. A comparison of 268. Cimochowski GE, Joyner LR, Fardin R, et al. Pleuroperito
rotation and nonrotation in tetracycline pleurodesis. Chest. neal shunting for recalcitrant pleural effusion. J Thorac Car
1 993; 1 04: 1 763- 1 766. diovasc Surg. l 986;92:866-870.
1 88 PLE U RAL D I S EASES
269. Little AG, Kadowaki MH, Ferguson MK, et al. Pleuroperiro 274. al-Kattan KM, Kaplan DK, Goldstraw P. The non-functioning
neal shunting: alternative therapy for pleural effusions. Ann pleuro-perironeal shunt: revise or replace? 1horac Cardiovasc
Surg. 1 9 8 8;208 :443-450. Surg. 1 994;42:3 1 0-3 1 2 .
270. Tzeng E, Ferguson MK. Predicting failure following shunt 2 7 5 . Petrou M, Kaplan 0, Goldstraw P. Management of recurrent
ing of pleural effusions. Chest. 1 990;98: 890-893. malignant pleural effusions. The complementary role talc
27 1 . Lee KA, Harvey JC, Reich H, et al. Management of malig pleurodesis and pleuroperironeal shunting. Cancer. 1 995;
nant pleural effusions with pleuroperitoneal shunting. J Am 75:80 1-805.
Coll Surg. 1 994; 1 7 8 : 5 86-5 8 8 . 276. Martini N, Bains MS, Beattie EJ Jr. Indications for pleu
272. Tsang V, Fernando H C , Goldscraw P. Pleuroperironeal shunt recromy in malignant pleural effusion. Cancer. 1 975;35:
for recurrent malignant pleural effusion. Thorax. 1 990; 734-73 8 .
4 5 :369-372. 277. Bernard A , d e Dompsure R B , Hagry 0, e t a l . Early and lace
273 . Ponn RB, Blancaflor J, D'Agoscino RS, et al. Pleuroperironeal mortality after pleurodesis for malignant pleural effusion.
shunting for intractable pleural effusions. Ann 1horac Surg. Ann 1horac Surg. 2002;74:2 1 3-2 1 7 .
1 9 9 1 ; 5 1 :605-609 .
Pr i mary Tumors of the Pleura
Ma ligna n t Meso theliomas, Solita ry Fibro us Tum ors, Body Ca vity
Lymphoma, a n d Pyo th orax-Associa ted Lymphoma
1 89
1 90 PLE U RAL D I S EASES
any of the different types of asbestos results in the mesothelioma ( 1 9) . A follow-up study in 2003 of 66 1
production of mesotheliomas in 8% to 66% of ani adult villagers alive in 1 979 revealed that 1 1 9 ( 1 8%)
mals, depending upon the dose. Mesotheliomas can had died of mesothelioma, and 1 9% of the meso
also be induced after the inhalation of various types thelioma deaths were in villagers less than 40 years
of asbestos ( 1 2) . These mesotheliomas are histologi old (20) . Asbestos does not occur in the local soil
cally identical to tumors in humans ( 1 3 , 1 4) . or rocks, nor is it handled in the village. The atmo
Th e incidence o f mesotheliomas following asbes sphere in the area does contain increased amounts of
tos exposure increases linearly with the intensity of erionite, a mineral of the zeolite family. This mineral
the exposure but exponentially (to the third or fourth is a major contributor to the clouds in the area. This
power) with time from the first asbestos exposure. report demonstrates that the inhalation of airborne
The risk of malignant mesothelioma can be estimated respirable fibers other than asbestos can be associated
from the following mathematic equation: with the subsequent development of pleural meso
theliomas ( 1 9) . In North Dakota, erionite has been
found to contaminate gravel used to pave more than
R = K X F X TP
300 miles of roads (2 1 ) . It remains to be seen whether
this contamination will lead to cases of mesothe
where R is the risk of mesothelioma, K is a coef lioma in the future. When erionite fibers are admin
ficient dependent on the fiber size and type (highest istered intrapleurally to rats, they are two orders of
for crocidolite and lowest for chrysotile) , F is the magnitude more carcinogenic than crocidolite (22) .
number of fibers per milliliter, T is the time after the There are probably other factors related to the devel
first exposure, and p is the exponent that is thought opment of pleural mesothelioma. Antman et al. (23)
to be between 3 and 4 ( 1 ,2) . The most recent esti reported that mesothelioma developed in proxim
mates for Kfor chrysotile and crocidolite were 0.0 1 3 ity to a field of therapeutic radiation administered
and 1 .0, respectively (2) . The presence o f pleural 1 0 to 3 1 years previously in four patients. As subse
plaques or pleural thickening does not increase the quent review of 9,342 patients who received radio
risk of mesothelioma when the length and duration therapy for breast cancer revealed the three developed
of exposure are taken into consideration ( 1 5) . mesothelioma in the ipsilateral hemithorax (24) .
At times, the asbestos exposure may not be obvious. Roviaro et al. (25) reviewed 35 cases of pleural me
In one report ( 1 6), five cases of mesothelioma devel sothelioma and found that 3 of the patients had cal
oped in a Native American pueblo of approximately cified posttuberculous fibrothorax. There is no clear
2,000 persons. Epidemiologic investigation revealed evidence of a familial tendency to develop mesothe
that asbestos mats were used to insulate worktables lioma (26) . There is no definite evidence that smok
against the intense heat of brazing torches and molten ing increases the risk of mesothelioma ( 1 ) .
metal in the preparation of silver jewelry. In addition, One factor that has received much attention i n the
the villagers scrubbed leather with cakes of asbestos to last 20 years as a possible etiologic factor in the devel
make their leggings and moccasins a brilliant white. opment of malignant mesothelioma is the simian virus
The mechanism by which asbestos fibers induce 40 (SV40) . It should be stated that the role of SV40 in
malignant changes is not known. The asbestos fiber mesothelioma is controversial and the possibility that
appears to have two major sources of genotoxicity: technical factors can produce false-positive results sug
generation of reactive oxygen species and mechanical gestive of SV40 infection has been raised (27) . Two
effects such as interference with mitotic spindle for more recent well-designed studies have shown that
mation and the segregation of chromosomes ( 1 7, 1 8) . the polymerase chain reaction (PCR) primers used to
Exposure t o asbestos can also damage the cellular detect SV40 in many studies were targeting sequences
DNA, and if the cells with the damaged DNA either within the SV40 genome that are also present in com
do not undergo apoptosis or undergo cell cycle ar monly used laboratory plasmids, leading to the false
rest, a malignant transformation may occur. It should positive detection of SV40 (28,29) . A recent study
be noted that, however, so far no consistent abnor using highly sensitive RT-PCR-based assays that are
malities in oncogenes or suppressor genes have been specific for SV40-encoded micro RNAs found none
found in human mesothelioma ( 1 ) . in 94 malignant mesotheliomas (30) . Moreover, age
Mineral fibers other than asbestos can induce me specific trends in the pleural mesothelioma incidence
sotheliomas. In one area of Turkey, approximately rates are not consistent with an effect of exposure to
1 % of the population dies each year of malignant SV40-contaminated poliovirus vaccine (3 1 ) .
CHAPT E R 1 1 I P R I MARY TU M O RS OF TH E PLE U RA 191
A B
should be obtained for all patients in whom a malig cases, the pulmonary margin is irregular, and separate
nant mesothelioma is considered (Fig. 1 1 . l C) . With nodules representing either metastases or lymph node
mesothelioma, the disease is unilateral in almost all infiltration may be seen in the juxtamediastinal tis
cases (46) . On CT scan, the pleura is thickened, with sue. In a patient with a malignant pleural effusion,
an irregular, often nodular internal margin that serves features that suggest malignant mesothelioma rather
to distinguish this tumor from other types of pleu than metastatic pleural disease are rind-like pleural
ral thickening. These changes are most pronounced involvement, mediastinal pleural involvement, and
at the base of the lung. The CT scan usually reveals pleural thickness more than 1 cm (49) .
marked thickening of the major fissure due to a The volume of the hemithorax with malignant
combination of fibrosis, tumor, and associated Buid. mesothelioma is quite varied. If the patient has a
The fissure may also appear nodular because of tu pleural effusion with pleural thickening and de
mor infiltration (47) . At times, pleural thickening is creased volume of the ipsilateral hemithorax, it is
seen predominantly along the mediastinum. In such suggestive of mesothelioma. In one series, the volume
CHAPT E R 1 1 I P R I MARY TU M O RS OF TH E PLE U RA 1 93
of the hemithorax was reduced in 42% of 50 cases the remainder. This fluid is exudative. In approxi
of mesothelioma (45), whereas in another series, mately one third of patients, the pleural fluid glucose
the volume of the hemithorax was reduced in 30% level is below 50 mg/dL and the pleural fluid pH level
of 50 cases (46) . It should be emphasized, however, is below 7.20 ( 5 5 ) . Patients with a low pleural fluid
that contralateral mediastinal shift is seen in approxi pH or low pleural fluid glucose level tend to have a
mately 1 5 % to 25% of cases, and this is usually due poorer prognosis ( 5 5 ) . The pleural fluid is generally
to a large effusion (45,46) . cellular and contains a mixture of normal mesothelial
The CT scan is also useful in demonstrating dis cells, differentiated and undifferentiated malignant
ease beyond the pleura and is thereby quite useful in mesothelial cells, and varying numbers of lympho
staging the disease. It often reveals intrapulmonary cytes and polymorphonuclear leukocytes ( 56 ) . Al
nodules that are not apparent on the standard chest though measurement of the levels of tumor markers
radiograph (47) . The CT scan may reveal chest wall in pleural fluid is generally not recommended (see
invasion, diaphragmatic invasion, or extension of the Chapter 7) , malignant mesotheliomas tend to have
tumor to the liver or retroperitoneal space. However, high levels of CYFRA 2 1 - 1 and low levels of carcino
CT scans are not without problems with regard to embryonic antigen (CEA) in comparison with meta
mesothelioma; CT fails to identify chest wall and static adenocarcinoma ( 5 7 ) .
mediastinal invasion in some patients who undergo It has been proposed that the demonstration of
surgical resections (50) . In addition, it is often very elevated levels of soluble mesothelin related protein
difficult to distinguish pleural disease alone from asso (SMRP) are useful in the diagnosis of malignant me
ciated pericardia! disease, and extensive pleural disease sotheliomas (5 8-60) . Creaney et al ( 6 1 ) measured
often envelops and obscures the nodal anatomy in the the SMRP levels in pleural fluids from 1 92 patients
hilar and middle mediastinal nodal groups (45). presenting to a respiratory clinic including 52 with
Magnetic resonance imaging (MRI) and fluoro malignant mesothelioma, 56 with nonmesothelioma
deoxyglucose (FDG) positron emission tomography tous malignancies and 84 benign effusions. The pleu
(PET) are other imaging modalities that are useful in ral effusions from the patients with mesothelioma
the evaluation of patients with suspected malignant had significantly higher concentrations of SMRP
mesothelioma (5 1 ,52 ) . MRI is superior to CT in de than did the other patients ( 6 1 ) . However, the SMRP
termining the extent of a malignant mesothelioma in patients with sarcomatoid mesothelioma did not
particularly when the tumor invades local structures differ significantly from nonmalignant effusions ( 6 1 ) .
such as the ribs and the diaphragm (52 ) . PET scans Davies et al ( 60 ) measured pleural fluid SMRP in 24
are useful in differentiating malignant pleural disease patients with mesothelioma, 67 patients with pleural
from benign pleural disease (5 1 ,53). In one series, 28 metastases and 75 patients with benign conditions.
patients with pleural thickening, including 24 with Using ROC curve analysis, pleural fluid SMRP had
malignant and 4 with benign disease, were subjected an AUC of 0.878 in its ability to differentiate between
to FDG-PET scanning. The uptake of FDG was sig patients with mesothelioma and all other diagnoses
nificantly higher in the malignant lesions than in the at an optimal cutoff value of 20 nM ( 60 ) . At this
benign lesions. In addition, the FDG-PET images cutoff, the diagnostic sensitivity and specificity were
provided excellent delineation of the extent of the 0.71 and 0.90, respectively ( 60 ) . Again, the levels of
disease (5 1 ) . However, it does not appear that meta SMRP were lower in patients that had sarcomatoid
static adenocarcinoma can be differentiated from ma mesothelioma ( 60 ) . Adenocarcinomas accounted for
lignant mesothelioma with this imaging technique. 12 of the 13 false positives. The above two studies
In a paper reporting on PET scan results in 63 pa demonstrate that pleural fluid SMRP measurement
tients with mesothelioma, the authors concluded that provide additional information to cytology. However,
the PET scan defines metastatic disease and delin tissue confirmation of mesothelioma is indicated in
eates the extent oflocal disease as well (54) . It has also most situations.
been shown that the intensity of the FDG uptake on At times, the pleural fluid of patients with malig
the PET scan is a significant factor in the prediction nant mesothelioma is viscid, owing to the presence of
of patient survival (53) . large amounts of hyaluronate, which was previously
called hyaluronic acid. Nurminen et al. ( 62 ) assessed
the diagnostic utility of hyaluronate levels by assaying
Pleural Fluid
the levels in 1 ,039 samples of pleural fluids including
The pleural fluid with mesotheliomas is yellow in ap 50 from mesothelioma. They found that with a cut
proximately 50% of patients and serosanguineous in off of 75 mg/L for hyaluronate, the assay specificity
1 94 PLE U RAL D I S EASES
for malignant mesothelioma was 1 00% and the biopsy (70) . However, the diagnosis of mesothelioma
sensitivity was 56% (62) . It should be noted that an can be established in more than 80% of patients
other study (63) demonstrated that the hyaluronate with mesothelioma through a image-guided cutting
measurements were much less useful. It appears that needle biopsy (7 1-73) (see Chapter 28) .
the poor results in the latter study are probably at More invasive procedures are often necessary to
tributable to procedural mistakes (64) . The results by provide a larger tissue sample so that a definitive di
Nurminen et al. (62) were obtained by high-pressure agnosis can be made. If the patient has skin depos
liquid chromatography (HPLC) , and this assay is not its, these should be biopsied. However, usually the
generally available in the United States. diagnosis must be made based on a thoracoscopy or
open biopsy. Thoracoscopy establishes the diagnosis
of mesothelioma in more than 90% of cases. When
Diagnosis
two series (74-75) were combined, the diagnosis was
The diagnosis of malignant mesothelioma should be established in 5 1 of the 56 (90%) patients. These
considered in all patients with exudative pleural ef results were similar to those with thoracotomy and
fusions. The suspicion of mesothelioma should be open biopsy (43) . Thoracoscopy is therefore the pro
higher in middle-aged or older patients with persis cedure of choice because its diagnostic yield is simi
tent chest pain or shortness of breath, particularly lar to that of open thoracotomy, but the procedure
if there is a history of asbestos exposure. The chest is less invasive (43). It should be noted that the clas
CT scan is frequently suggestive of the diagnosis. Al sification of the mesothelioma with thoracoscopy is
though a diagnosis of malignancy can be established wrong in about 1 5% of cases (76) . Most cases of mis
by cytologic smears or needle biopsies of the pleura, classification are diagnosed as an epithelial subtype
these procedures usually cannot distinguish between at thoracoscopy and a biphasic subtype as the final
a metastatic adenocarcinoma and a mesothelioma. diagnosis (76) .
In one series (65) , 80 patients with mesothelioma Malignant mesothelioma often infiltrates needle
had pleural fluid cytology. In 20 patients (25%) , cy tracts, thoracotomy scars, and chest tube drainage
tologic examination of the pleural fluid established sites after diagnostic or therapeutic procedures. In
that the patient had malignant disease, but in none one study (73) , the incidence of seeding was 4% for
could the diagnosis of mesothelioma be established image-guided core-needle biopsy and 22% for tho
definitely with only cytology. In the remaining racoscopic biopsy. The role of small amounts of ir
60 patients, the diagnosis of malignancy could not radiation after the procedure to prevent such seeding
be established (65 ) . is controversial. Randomized studies have come to
There are, however, certain cytologic features different conclusions (77) . A recent review (77) on
that assist in making this differentiation. One report the subject concluded that prophylactic radiation was
(66) compared the cytologic features of 44 cases of not justified.
malignant mesothelioma and 46 cases of metastatic Histologic examination of hematoxylin and eo
adenocarcinomas, and the authors concluded that sin (H&E)-stained tissue section remains the pri
the following five features separate malignant me mary method by which the diagnosis of malignant
sothelioma from adenocarcinoma with better than mesothelioma is established. Malignant mesothe
95 .4% accuracy. Mesotheliomas tend to have true lioma can have many different histologic patterns
papillary aggregation, multinucleation with atypia, (78) . However, it is frequently difficult to distin
and cell-to-cell apposition, whereas adenocarcinomas guish malignant mesothelioma from metastatic ad
tend to have acinus-like structures and balloon-like enocarcinomas on the H &E-stained slides (36,78) .
vacuolation (66) . In addition, immunohistochemical Currently, immunohistochemical procedures have
studies on cell blocks from pleural fluid are useful in gained widespread acceptance as valuable adjuncts
distinguishing mesothelioma from adenocarcinoma in establishing the diagnosis of malignant mesothe
(67-69) (see Chapter 7) . lioma (36,79) . Some of the immunohistochemical
Blind needle biopsy of the pleura is usually not markers are positive with adenocarcinomas, whereas
diagnostic of mesothelioma. In one report (65), the others are positive with malignant mesotheliomas.
needle biopsy was diagnostic of mesothelioma in only When one wishes to differentiate metastatic adeno
1 8 of 84 cases (2 1 %) . Also, there is poor concordance carcinoma from malignant mesothelioma, the tissue
among different pathologists when the diagnosis of sections should be stained with a panel of immuno
mesothelioma is based on specimens from needle histochemical markers (see Chapter 7) . Currently,
CHAPT E R 1 1 I P R I MARY TU M O RS O F TH E PLE U RA 1 95
the best markers for mesothelioma appear to be cal (range 1 .3-4: 1 ) . SEM can be used when glutaralde
retinin, keratin 5/6, podoplanin, and WT l , whereas hyde-fixed, plastic-embedded tissue is not available
the best markers for metastatic adenocarcinomas are for transmission EM (84) . A number of subcellular
CEA, MOC-3 1 , B72.3, Ber-EP4, BG-8, and TTF - 1 structures, such as mucin granules, myelinosomes,
(79) (see Chapter 7) . Th e sarcomatous type of me microvilli coated by a filamentous glycocalyx, and
sothelioma usually does not stain positively with microvillous rootlets, may be observed in some ad
calretinin (80) . Moreover, it is important to realize enocarcinomas. The presence of any of these features
that nonmalignant mesothelial cells also stain posi excludes mesothelioma (37) .
tive for cytokeratin 5/6 and calretinin (8 1 ) . Indeed, Flow cytometry does not appear to be particu
one of the most difficult differentials is to distin larly useful in establishing the diagnosis of malignant
guish malignant mesothelioma cells from benign mesothelioma. Burmer et al. (85) performed flow
reactive mesothelial hyperplasia (82) . The most re cytometry on 46 cases of malignant pleural meso
liable criterion for determining that a mesothelial thelioma and 3 1 nonmesothelioma malignancies
proliferation is malignant is the demonstration of of the pleural space. They reported that 65% of the
invasion (82) . mesotheliomas were diploid in DNA content, with
The diagnosis of malignant mesotheliomas is usu intermediate-to-low proliferative rates. In contrast,
ally established by the combination of the histology 8 5 % of the nonmesothelial malignant neoplasms
and the immunohistochemical stains. However, when were aneuploid.
doubt exists, two tests that have been used for decades
can at times still be useful. The periodic acid-Schiff
Blood Markers of Mesothelioma
(PAS) stain can still be used to distinguish mesothelio
mas from adenocarcinomas. The presence of strongly In the last fifteen years, there have been many pa
positive vacuoles after diastase digestion effectively es pers written on blood markers for mesothelioma.
tablishes the diagnosis of adenocarcinoma, although The three blood markers that have received the most
not all adenocarcinomas have this staining character attention are soluble mesothelin-related peptides
istic. In addition, most mesotheliomas contain large (SMRP), osteopontin, and megakaryocyte-potentiat
amounts of hyaluronate that stain positively with col ing factor (MPF) (86) . Mesothelin is a glycoprotein
loidal iron or alcian blue stains. To be unequivocally that is expressed on the surface of normal mesothe
positive, absence or attenuation of blue staining after lial cells, but is overexpressed in mesothelioma and
pretreatment of a serial section with bovine testicular various carcinomas (87) . SMRP are released from
hyaluronidase overnight is required (83) . the cell surface into the serum. Osteopontin is also
The diagnosis of pleural mesothelioma is made ac a glycoprotein that is overexpressed in lung, breast,
curately in most cases without resorting to electron colorectal, gastric, and ovarian carcinoma.
microscopic (EM) examination (36) . However, be The reported studies have demonstrated that the
cause EM still plays a decisive role in some cases with mean serum levels of SMRP, osteopontin, and MPF
unusual morphology or anomalous histochemical are all higher in patients with malignant mesotheli
or immunohistochemical reactions, a portion of the oma than in normal individuals and patients exposed
pleural specimen should be routinely fixed at the time to asbestos but without mesothelioma (86,88) . How
of pleural biopsy for possible subsequent processing ever, osteopontin lacks specificity for mesothelioma
for EM if malignant mesothelioma is suspected (36) . while both SMRP and MPF lack sensitivity for de
Epithelial mesotheliomas are characterized by the tecting nonepithelial subtypes (86) . Moreover, there
presence of tonofilaments, desmosomes, and micro is significant overlap in the values in the different
villi. The appearance of the microvilli is important groups (87,89-92) . In one study (88) in which the
in distinguishing mesotheliomas from adenocarci three markers were all measured, the SMRP was by
nomas. In mesothelioma, the microvilli are numer far the most accurate in diagnosing mesothelioma.
ous and are characteristically long and thin, whereas However, to have a specificity of 95%, the sensitivity
in adenocarcinoma they are typically much less fre for SMRP was only 73% (88) . The SMRP is also not
quent and are usually short and stubby (36,84) . In particularly efficient at detecting malignant mesothe
one study (84) , with scanning electron microscopy lioma early. In one study (93) of 1 06 asbestos exposed
(SEM) , the mean length-to-diameter ratio of the individuals who developed mesothelioma, only 1 7
microvilli in mesotheliomas was 1 9.7: 1 (range 1 3.7- o f the 1 06 had elevated levels o f SMRP before they
23. 5 : 1 ) , whereas that for adenocarcinomas was 2.5 : 1 were diagnosed. The practical value of using any of
1 96 PLE U RAL D I S EASES
these three serum markers as a diagnostic test for the sarcomatous-type mesothelioma (98,99, 1 02) . Pa
mesothelioma must await the assessment of their lev tients with larger tumor volumes and more advanced
els in other disorders that are common after extensive stages of the disease also have shorter survival times
exposure to asbestos, such as diffuse pleural thicken ( 1 03) . Survival time is also poorer in patients with
ing with or without benign effusion, lung cancer with regional lymph node involvement and in patients
pleural involvement and rounded atelectasis, and who have a high standard uptake value (SUV) on
metastatic adenocarcinoma of the pleura. The ERS/ their PET scan ( 1 04) . Pleural fluid findings may also
ESTS task force (94) concluded that at the present be associated with the expected survival time. It has
time there is no place for screening for malignant me been shown that an elevated pleural fluid hyaluranon
sothelioma. In one study (95) of 1 00 patients with ( 1 05) or a high pleural fluid pH level (> 7 .32) ( 1 06)
malignant mesothelioma, 1 39 with lung cancer and are associated with increased survival. Survival is also
75 with benign asbestosis, the best statistical cutoff less in patients with mesotheliomas that stain posi
for SMRP only had a sensitivity of 53% and a speci tively for vascular endothelial growth factor (VEGF)
ficity of 82.7%. Interestingly, Scherpereel et al. (90) ( 1 07) or have a high microvessel density ( 1 08). Re
reported that in patients with pleural effusions, the cently, researchers at Harvard Medical School have
pleural fluid SMRP levels were much higher than the developed a four-gene expression ratio test ( 1 09).
simultaneous serum SMRP levels. Moreover, they When this test was applied to 1 20 patients undergo
reported that the pleural fluid levels of SMRP were ing surgery, the mean survival in the 70 low-risk pa
better at distinguishing mesothelioma from pleural tients by the gene test was 1 6.8 months compared to
metastatic disease than were the serum levels (90) . 9 . 5 months in the 50 high-risk patients ( 1 1 0) . Musk
These serum markers may be useful prognostically et al. ( 1 1 O) recently reported that the median over
in patients with malignant mesothelioma. Cappia et al. all survival with malignant mesothelioma gradually
(96) showed that osteopontin levels were significantly increased from 1 960-1 970 (64 days) to 2000-2005
lower in 32 long-term survivors (>24 months) than in (30 1 days) .
69 short-term survivors ($24 months) . Higher SMRP
levels have also been associated with a poorer progno Staging
sis (97) . Wheatley-Price et al. (97) have shown than
changes in SMRP levels but not osteopontin are indica When a patient is suspected of having a mesothelioma,
tors of the response to treatment of multiple myeloma. the extent of the disease should be staged because the
stage of the disease dictates the therapeutic approach.
There have been several different staging systems pro
Management and Prognosis
posed. The staging scheme recommended by the Can
The prognosis of patients with pleural mesothelioma cer Committee of the College ofAmerican Pathologists
is more dependent on the so-called pretreatment fac is that developed by the American Joint Committee on
tors than on the effect of therapeutic intervention. Cancer (AJCC) and the International Union Against
In general, the prognosis of patients with malignant Cancer (UICC) . It is based on evaluation of tumor,
mesothelioma is not good, with an overall median nodes, and metastases (TNM) as shown in Table 1 1 . 1 .
survival time of approximately 8 to 1 2 months af To stage the disease, the following studies are required:
ter diagnosis (98,99) . There is a small fraction of a barium swallow to assess esophageal involvement; a
patients with mesothelioma who have a prolonged bronchoscopic examination to assess involvement of
survival, up to 1 0 years ( 1 00) . It should be noted, the tracheobronchial tree; a chest CT scan to assess
however, that the life expectancy is higher in patients mediastinal or chest wall involvement; brain, liver, and
with mesothelioma than in patients with metastatic bone scans to look for distant metastases; and possibly,
malignancy, in which case, the median survival is a pneumoperitoneogram to look for diaphragmatic
approximately 4 to 5 months after diagnosis. The penetration ( 1 1 1 ) . Rusch and Venkatraman ( 1 1 2) in
following have been identified as adverse prognostic vestigated prognostic factors in 23 1 patients who un
factors: poor performance status, age greater than derwent thoracotomy between 1 983 and 1 998. They
75 years, chest pain, nonepithelioid histology, el found that the median survival time was 29.9 months
evated serum lactate dehydrogenase, elevated platelet for stage I tumors, 1 9 months for stage II, 1 0.4 months
counts, male sex, and elevated peripheral white blood for stage III, and 8 months for stage IV ( 1 1 2) . A sim
cell count (> 8.3/mm3) ( 1 00, 1 0 1 ) . The prognosis is pler staging system devised by Butchart et al. ( 1 1 1 ) is
better with the epithelial type when compared with shown in Table 1 1 .2.
CHAPT E R 1 1 I P R I MARY TU M O RS OF TH E PLE U RA 1 97
. . . . . .
TO No evidence of p r i m a ry t u m o r
Tu m o r confi ned with i n the ca p s u l e of the
T1 Tu m o r l i m ited t o i p s i l ate ral pa rieta l o r vis pa rietal p l e u ra , i . e . , i nvolvi n g o n l y the
cera l p l e u ra i p s i l atera l p l e u ra , l u n g , pericard i u m , a n d
T2 Tu m o r i nvades any of the fo l l owi n g : i p s i l at d i a p h ra g m
eral l u n g , e n d othoracic fascia, d i a p h ra g m , Tu m o r i nvolvi n g ch est wa l l or m e d i asti n a l
o r pericard i u m structu res; poss i b l e lym p h n o d e i nvolve
T3 Tu m o r i nvades any of the fo l l owi n g : i p s i l at ment i n s i d e the ch est
eral ch est wa l l m uscle, ribs, or med iasti n a l Ill Tu m o r penetrati n g d i a p h ra g m to i nvolve
o r g a n s o r tissues the perito n e u m ; contralatera l p l e u ra l
T4 Tu mor d i rectly extends to any of the fol l ow i n volvement; lym p h n o d e i nvolvement
i n g : contra l atera l pleural, contra l atera l l u n g , o utsi d e the chest
perito n e u m , i ntra-abdom i n a l organs, or IV D ista nt bloodborne m etastases
cervica l tissue
M o d ified from B utch a rt EG, Ashcroft T, B a r n s ley W C ,
Regional Lymph Nodes (N) e t a l . T h e r o l e o f s u rgery i n d iffuse m a l i g n a n t m eso
NO N o reg i o n a l lym p h node m etastasis t h e l i o m a of t h e p l e u ra . Semin Oneal. 1 98 1 ; 8 : 3 2 1 -
3 2 8 , with p e r m i ss i o n .
N1 M etastasis i n i ps i late ra l peri b ro n c h i a l o r
i p s i l atera l h i la r lym p h nodes, i n c l u d i n g
i ntra p u l mo n a ry nodes i n vo lved b y d i rect
Palliative Therapy
exte n s i o n of the p r i m a ry t u m o r
N2 M etastasis i n i ps i late ra l m e d iasti n a l o r s u b No satisfactory treatment exists for malignant meso
cari n a l lym p h nodes thelioma. The beneficial effects of treatment of ma
N3 M etastasis i n contra lateral med iasti
lignant mesothelioma are moot ( 1 1 3) . In a previous
nal, contralatera l h i l a r, i psi latera l or edition of this book, it was written that there were no
contralateral sca lene, o r s u p raclavi c u l a r controlled studies to demonstrate that any treatment
lym p h nodes is effective in prolonging survival. However, recently
Distant Metastasis (M) it has been shown that systemic chemotherapy with
cisplatin and pemetrexed does prolong survival (see
MO N o evi d e n ce o f d i stant m etastasis
the discussion on chemotherapy in this chapter) .
M1 D ista nt m etastasis The futility of treatment of mesothelioma in the
AJCC/UICC TNM Stage Groupings past was demonstrated by a study of Law et al. ( 1 1 4)
Sta g e I T1 NO MO who compared the survival rates in 64 untreated pa
tients and 52 treated patients seen at the Brampton
T2 NO MO
and the Royal Marsden Hospitals between 1 97 1
Sta g e I I T1 N1 MO and 1 980. The two groups o f patients had compa
T2 N1 MO rable clinical conditions at the time of presentation.
Sta g e I l l T1 N2 MO Whether the patient received treatment depended
on the attending physician; some treated all patients,
T2 N2 MO
whereas others managed all patients symptomati
T3 NO, 1 , 2 MO cally. The survival curves for the 64 untreated pa
Sta g e IV Any T N3 MO tients and the 1 2 patients who received radiotherapy,
T4 Any N MO the 28 patients who received decortication, and the
1 2 patients who received chemotherapy were virtu
Any T Any N M1
ally identical. The median survival time was approxi
TN M , p r i m a ry t u m o r, reg i o n a l l y m p h nodes, a n d d is mately 1 8 months; 1 0% of patients survived more
tant m etastas i s . AJ C C , A m e r i c a n J o i nt C o m m ittee on than 4 years, including 7 of the 64 ( 1 1 %) untreated
C a n cer; U I C C , I nternati o n a l U n i o n A g a i nst C a n ce r. patients ( 1 1 4) . These results indicate that controlled
1 98 PLE U RAL D I S EASES
cooperative studies are needed to assess the effective malignant mesothelioma between 1 959 and 1 972
ness of the various treatment modalities proposed and reported that 2 patients (7%) were alive, with
for malignant mesothelioma. out evidence of recurrence, 3.5 and 6 years after the
Shortness of breath and chest pain are the two operation. The surgical resections performed by this
most troublesome symptoms in patients with ma group were extensive, with removal of the pleura,
lignant mesothelioma. The shortness of breath can lung, lymph nodes, ipsilateral pericardium, and dia
be either due to the presence of a large pleural ef phragm. The in-hospital postoperative mortality rate
fusion or to invasion of the lung or mediastinum was 3 1 %. These workers concluded that radical extra
by the tumor. If the patient is breathless and has a pleural pneumonectomy was only indicated for pa
pleural effusion, a therapeutic thoracentesis should tients younger than the age of 60 who are fit and who
be performed. If the shortness of breath is relieved have stage I tumors of the epithelial type.
by the thoracentesis, a pleurodesis should be at In recent years, Sugarbaker et al. ( 1 1 8) have re
tempted (65), or indwelling pleural catheter ( 1 1 5) ported good results in a series of 1 83 patients who
should be inserted. Pleurodesis should not be at underwent extrapleural pneumonectomy, followed
tempted if the mesothelioma prevents the under by chemotherapy and radiotherapy. Patients were
lying lung from expanding. In this situation, an operated on only if they had a Karnofsky perfor
indwelling pleural catheter should be inserted if a mance status greater than 70%, a creatinine level and
thoracentesis relieved the patient's dyspnea. Details liver function test results within normal limits, and
concerning both these procedures are delineated in a tumor that was judged to be completely resectable
Chapter 1 0 . If the shortness of breath is not relieved on the basis of CT and MRI scan. The extrapleural
by the thoracentesis, then oxygen or opiates should pneumonectomy entailed resection of the pleura,
be prescribed. lung, diaphragm, and pericardium en bloc ( 1 1 8) .
The other main symptom in patients with malig I n the series reported b y Sugarbaker, there were
nant mesothelioma is chest pain, frequently caused seven (3 . 8 %) perioperative deaths and the median
by tumor invasion of the chest wall. In such persons, postoperative length of stay was 9 days ( 1 1 8) . In a
local palliative radiotherapy may relieve the symp more recent report by the same group on 496 con
toms (48), but frequently the response is minimal or secutive patients, the operative death was 4% ( 1 1 9) .
nonexistent (46, 1 1 6) . More often, strong analgesics Overall, the median survival fo r these patients was
must be administered to control the pain. If the pain 1 9 months and the 2- and 5-year survival rates were
is severe, consideration should be given to performing 38% and 1 5 % , respectively. The subset of 3 1 pa
a percutaneous cervical cordotomy. In one series, this tients with epithelial cell type, negative resection
procedure was performed on 52 patients with intrac margins, and negative extrapleural nodal status
table chest pain who were taking a median of 1 00 mg had a median survival of 5 1 months, a 2-year sur
morphine per day ( 1 1 7) . After the procedure, 38% vival rate of 68%, and a 5-year survival rate of 46%
of the patients were able to stop taking morphine ( 1 1 9) . In a second study, Aziz et al. ( 1 20) performed
and another 37% were able to reduce their morphine extrapleural pneumonectomies on 64 of 3 0 1 pa
intake by more than 50%. Two patients experienced tients seen at their hospital. They reported that the
troublesome dysesthesia following the procedure, median survival was no different in the 1 3 patients
and four had persistent motor weakness. No patient subjected to surgery and no chemotherapy than it
became hemiplegic or was unable to walk. was in the 238 patients not subjected to surgery
Another troublesome symptom in approximately ( 1 20). However, the median survival in the 5 1 pa
one third of patients is intermittent fever and sweat tients who received both surgery and chemotherapy
ing (65). Law et al. (65) reported that the administra was 35 months ( 1 20) .
tion of prednisolone was of some benefit in alleviating It should be emphasized that the series reported
the fever and sweating and usually improved the ap by Sugarbaker et al. represent a select group of pa
petite and the well-being of the patient. tients and the overall median survival of 1 9 months
was disappointing. Only 1 % to 5% of patients with
malignant mesothelioma are candidates for extra
Surgical Treatment
pleural pneumonectomy ( 1 2 1 ) . Whether extrapleu
Surgical management appears to be the only form ral pneumonectomy combined with chemotherapy
of therapy that offers the patient any hope for cure. and/or radiotherapy improves survival time and the
Butchart et al. ( 1 1 1 ) operated on 29 patients with quality of life in patients with mesothelioma is yet
CHAPT E R 1 1 I P R I MARY TU M O RS OF TH E PLE U RA 1 99
to be proved ( 1 22, 1 23) . The combination of sur decortication who had participated in various trials.
gery, radiotherapy, and chemotherapy is expensive, They could find no difference in the survival rates
time consuming, and usually requires that the pa with the two surgical methods ( 1 1 2) . It should be
tients be away from home for considerable periods noted that in most series, surgery is combined with
of time. The ERS/ESTS (94) task force concluded another therapeutic modality. Controlled studies are
that extrapleural pneumonectomy should only be needed to compare pleurectomy with no surgery in
performed in clinical trials, in specialized centers trimodality therapeutic plans.
as part of multimodal treatment. In a recently re
ported study ( 1 24) from the United Kingdom, 5 0
Chemotherapy
patients eligible for extrapleural pneumonectomy
were randomized to receive extrapleural pneumo A role for chemotherapy in the treatment of malig
nectomy or no surgery. All patients received induc nant mesothelioma has been established in the past
tion platinum-based chemotherapy and all patients 1 0 years and cisplatin in combination with peme
received postoperative radiotherapy to the hemitho trexed or raltitrexed is now considered the first-line
rax ( 1 24) . In this study, the patients who received systemic therapy for select patients with mesothe
the extrapleural pneumonectomy lived a shorter lioma (94) . The definitive study was reported by
time ( 1 4.4 months vs. 1 9 . 5 months) ( 1 24) . After Vogelzang et al. ( 1 27) in 2003. In this study, 456
adj ustment for sex, histological subtype, state, and patients were randomly assigned to receive peme
age, the patients who did not undergo extrapleural trexed, a multitargeted antifolate, 500 mg/m 2 and
pneumonectomy had a significantly longer median cisplatin 75 mg/m2 or cisplatin 75 mg/m2 alone
survival ( 1 24) . There was also much higher morbid on day 1 and every 2 1 days thereafter ( 1 27) . The
ity associated with extrapleural pneumonectomy in doses of the drugs were decreased if toxicity oc
this study ( 1 24) . The results of this study suggest curred. To be eligible for the study, the patients
that radical surgery in the form of EPP within tri were required to have a Karnofsky performance
modal therapy offers no benefit and possibly harms status greater than 70 and have a life expectancy
the patient. of more than 3 months. In this study, the response
Other researchers have recommended that only rate in the combination group was significantly
pleurectomies be performed or that pleurectomies better than in the cisplatin-only group (4 1 . 3% vs.
be performed on patients who are not candidates for 1 6.7%) as was the median survival ( 1 2. 1 months
extrapleural pneumonectomies ( 1 25, 1 26) . Martin vs. 9 . 3 months) ( 1 27) . However, when the Kaplan
Ucar et al. ( 1 25) attempted surgical debulking in 5 1 Meier survival plots are examined (Fig. 1 1 .2), the
patients and were able to perform the procedure with differences between the two regimens are not par
VATS without resorting to thoracotomy in 1 7 pa ticularly impressive. The combination regimen had
tients. However, the 30-day mortality was 7.8% and significantly greater toxicities than did the cisplatin
the median survival was only 7 months ( 1 25). Hal alone regimen with the most severe toxicities be
stead et al. ( 1 26) attempted to do a VATS debulking ing severe neutropenia (27.9%) and severe leuko
pleurectomy-decortication in 79 patients and were penia ( 1 7. 7%) . A confirmatory study was reported
successful in 5 1 . If a decortication did not appear to by Van Meerbeeck et al. ( 1 28) . In this study,
be feasible, only a biopsy was attempted ( 1 26) . They 2 5 0 patients were randomized to receive 80 mg/m 2
reported that the median survival was significantly cisplatin intravenously or cisplatin plus raltitrexed,
greater in the pleurectomy group (4 1 6 days) than it another multitargeted antifolate, 3 mg/m 2 • Among
was in the biopsy-only group ( 1 27 days) ( 1 26) . How the 2 1 3 patients with measurable disease, the re
ever, the difference in survival may well have been sponse rate tended to be better in the combination
due to the greater tumor burden in the biopsy-only group than in the cisplatin-only group (23 .6% vs.
group. The ERS/ESTS task force (94) concluded that 1 3 . 6%, p = 0.056). The median survival was also
this treatment can be considered in patients to obtain greater in the combination group ( 1 1 .4 months vs.
symptom control, especially symptomatic patients 8 . 8 months) ( 1 28) . Again there was more toxicity
with entrapped lung syndrome who cannot benefit in the combination group ( 1 28 ) .
from chemical pleurodesis. I n view of the facts mentioned in the preceding
Rusch and Venkatraman ( 1 1 2) compared the sur text, it appears that the administration of the com
vival rate in 1 1 5 patients who had extrapleural pneu bination of cisplatin and either pemetrexed or ralti
monectomy and 59 patients with pleurectomy and trexed is associated with an increased survival time
200 PLE U RAL D I S EASES
and an increased response rate when compared there is a threefold- to fivefold-advantage on a loga
with cisplatin alone. However, it must be noted rithmic scale for pleural versus plasma area under the
that the increase in the survival times (Fig. 1 1 .2) concentration-time curves for cisplatin and mitomy
and the response rates are not great and the combi cin ( 1 30) . However, the results of clinical studies have
nation has significantly more side effects than does been disappointing ( 1 3 1 ) .
cisplatin. Nevertheless, these are the first studies to
demonstrate a survival benefit with combination
Radiotherapy
chemotherapy and raise hopes that more effective
treatments will be found in the future. It should The results with radiotherapy in the treatment of
also be noted that pemetrexed is quite expensive. malignant mesothelioma have been disappoint
An outpatient at Vanderbilt University Medical ing ( 1 32) . External radiotherapy does not control
Center would pay more than US $ 1 8,000 for each mesotheliomas locally and is associated with severe
treatment with pemetrexed, whereas they would toxicity in the underlying lung ( 1 33). Presently, it
only pay US $250 for each treatment with cisplatin. is not recommended ( 1 32) . There may be a place
Before combination chemotherapy is initiated, the for internal radiation therapy in the management
costs, benefits, and side effects of the drugs should of malignant mesothelioma ( 1 34) . In one report
be discussed with the patient. If systemic chemo from the Memorial Sloan-Kettering Cancer Center,
therapy is going to be administered, it is probably 33 patients were treated with the implantation of
better to do it as soon as the malignant mesothe permanent radioactive iodine- 1 25 ( 12 5I) sources in
lioma is diagnosed. O'Brien et al. ( 1 29) random residual tumor. This report concluded that local ra
ized 43 patients with malignant mesothelioma to diotherapy improved the length of survival ( 1 34) .
receive at the time of diagnosis or at the time when Many patients are given external radiotherapy after
they had progressive symptoms. They reported that extrapleural pneumonectomy or decortication, but
the mean survival was 1 4 months in the early group there are no controlled trials documenting whether
but only 1 0 months in the delayed group ( 1 29) . this additional therapy improves survival. The ERS/
Intrapleural chemotherapy probably has more ESTS task force recommended that radiotherapy
promise than systemic chemotherapy. It has been should not be performed after pleurectomy or
shown that when drugs are administered intrapleurally, decortication (94) .
MS MS
1 . 00 1 . 00
Pemetrexed/Cisplatin 1 2 . 1 mo -- Pemetrexed/Cisplatin 1 3.3 mo
9.3 mo - - - - C i s p l atin 1 0.0 mo
'
Cisplatin
'
Log rank p value 0.020 Log rank p val ue 0.051
I \
\
I
I
\
\ -.
0 . 75 I
I 0 . 75 I
I
I
Ql Ql
I I
.� -�
I
\ I
I \
(ii (ii \
\
I \
c \ c I
I
0 \ 0 \
\
t 0.50 t 0 . 50
\
0 0
\ \
c. ,_ c.
\
\
•-,
e \ e
\
\ I
\
a.. ,_ a.. \
\
,,
• ,, \
0.25 0.25
..
,_ _ ,__,
._ _ _ _
0 . 00 0 . 00
0 5 10 15 20 25 30 0 5 10 15 20 25 30
Survival time (mo) Survival time (mo)
Patients at risk Patients at risk
Pem/Cls 226 1 85 111 50 19 7 0 Pem/Cls 1 68 1 41 86 35 9 1 0
A Cls 222 1 73 91 32 19 3 0 Cls 1 63 1 28 69 20 9 0 0 B
FIGURE 1 1 .2 • Ka p l a n - M e i e r est i m ates of overa l l su rviva l t i m e for a l l patie nts (A) a n d for a l l patie nts
s u p p l e m e nted with fo l i c acid a n d Vita m i n B , 2 (B). M S, m e d i a n s u rviva l ; Pem, pemetrexed; Cis, c i s p l ati n .
CHAPT E R 1 1 I P R I MARY TU M O RS O F TH E PLE U RA 20 1
Recurrences with the benign sessile and the benign is hypoglycemia (Doege-Potter syndrome) . Of the
pedunculated tumors occur in less than 1 0% and approximately 1 50 extrapancreatic tumors causing
deaths are rare with these types of tumors ( 1 47) . hypoglycemia reported by 1 975, 1 0 were solitary fi
brous tumors of the pleura. In a review of 360 cases
of solitary fibrous tumors of the pleura, symptomatic
Clinical Manifestations
hypoglycemia was reported in 4% ( 1 49). The mecha
This tumor is evenly distributed between the sexes, nism responsible for the hypoglycemia appears to be
and the median age of presentation is 57 years ( 1 45). the production of high levels of insulin-like growth
Approximately 50% of patients with solitary fibrous factor II (IGF-II) by the tumor ( 1 50). The increased
tumors are asymptomatic, and the tumor is detected production of this insulin-like substance leads to an
on routine chest radiographs ( 1 46, 1 49). In the re increased glucose utilization by peripheral tissues and
maining patients, cough, chest pain, and dyspnea decreased production of glucose by the liver. Tumors
are the most frequent symptoms, each occurring in that are associated with hypoglycemia tend to be
approximately 40% of symptomatic patients. Ap large. The hypoglycemia is relieved with surgical re
proximately 25% of symptomatic patients are febrile moval of the tumor.
without any evidence of infection ( 1 49). The inci Radiologically, these tumors are manifested as
dence of hypemophic pulmonary osteoarthropathy is solitary, sharply defined, discrete masses located
approximately 20% in patients with solitary fibrous at the periphery of the lung or related to a fissure
tumors, and the incidence is much higher with larger (42, 1 44) . At times, the mass may become very large,
tumors. In one series, 1 0 of 1 1 patients (9 1 %) with occupying most of the hemithorax (Fig. 1 1 .3). The
lesions larger than 7 cm in diameter had hypemophic mass is frequently lobulated (42) . The mass has an
pulmonary osteoarthropathy, whereas none of 4 1 pa associated pleural effusion 1 0% to 20% of the time
tients with smaller lesions had the syndrome ( 1 46) . ( 1 46, 1 5 1 , 1 52) , but the presence or absence of an
When the tumors are surgically removed, the symp effusion apparently has no effect on the patient's
toms of hypertrophic pulmonary osteoarthropathy prognosis (42) . In one case, more than 1 70 L of
are relieved immediately in almost all patients ( 1 46) . transudative pleural fluid was produced by a solitary
Another paraneoplastic syndrome that sometimes fibrous tumor ( 1 5 1 ) . Calcifications are occasionally
accompanies solitary fibrous tumors of the pleura evident within the mass ( 1 44) .
A B
The appearance of the solitary fibrous tumors on in the serous body cavities in the absence of solid
CT scan is characteristic ( 1 04, 1 4 1 ) . The tumors are tumors ( 1 57, 1 58) . Primary effusion lymphoma is also
large, noninvasive, and tend to enhance with intrave known as body cavity-based lymphoma and is associ
nous contrast material, but the enhancement is fre ated with human herpesvirus 8 (HHV-8) or Kaposi's
quently nonhomogeneous. The intense enhancement sarcoma-associated herpesvirus (KSHV) ( 1 59).
of these tumors appears to be due to their high vascu HHV-8 is a gamma herpes virus with sequence ho
larity, whereas areas of low attenuation are due to foci mology to the Epstein-Barr virus ( 1 60) . The pres
of myxoid or cystic degeneration and hemorrhage in ence of the HHV-8 appears to be specific for primary
the lesion (34) . There is no associated mediastinal effusion lymphoma. Uphoff et al. ( 1 60) searched for
lymphadenopathy. PET scans of patients with soli the presence of HHV-8 sequences by PCR using a
tary fibrous tumors usually reveal little or no uptake panel of 1 3 3 human cell lines from a variety of solid
of the isotope by the tumor ( 1 53). tumors, 1 1 4 hematopoietic cell lines including 50
B-cell leukemia/lymphoma-derived cell lines, and 7
Diagnosis cell lines established from patients with primary effu
sion lymphoma. All of the cell lines from the patients
A VATS procedure or a thoracotomy is usually nec with primary effusion lymphoma were strongly posi
essary for diagnosis, although the diagnosis has been tive for HHV-8, whereas none of the other cell lines
established by transthoracic cutting-needle biopsy in were positive for HHV-8. This observation suggests
some patients ( 1 54). However, because the appropri an etiologic role for HHV-8 in primary effusion lym
ate treatment for the solitary fibrous tumor is surgical phoma. Castillo et al. ( 1 6 1 ) reviewed 1 47 patients
removal, most patients should be subjected to VATS with primary effusion lymphoma and reported that
or a thoracotomy for diagnosis and excision. In one 1 04 (7 1 %) , were HHV-8 positive. Many of these
series of 1 1 0 patients, the resection was performed via tumors are also characterized by the presence of the
VATS in 59 (54%) ( 1 43). The existence of solitary Epstein-Barr virus ( 1 58).
tumors, such as solitary fibrous tumors of the pleura, Primary effusion lymphoma most commonly
that can produce systemic symptoms underscores the occurs in homosexual patients with acquired immuno
importance of obtaining histologic proof of malignant deficiency syndrome (AIDS) ( 1 57). Primary effusion
disease in patients suspected of having malignant tu lymphoma is rare and accounts for less than 5% of all
mors before instituting radiotherapy or chemother AIDS-related non-Hodgkin's lymphoma ( 1 62) . Until
apy. Obviously, bronchoscopic and sputum cytologic 20 1 2 only 1 47 cases had been reported in the litera
tests are negative with solitary fibrous tumors. ture ( 1 6 1 ) . Mbulaiteye et al. ( 1 63) reviewed 304,439
patients with AIDS and were able to find only 4 cases
Treatment and Prognosis of primary effusion lymphoma of the pleura. These
tumors occasionally occur in patients who are not
The treatment of choice for solitary fibrous tumors is infected with the human immunodeficiency virus
surgical removal by thoracotomy or by VATS ( 1 5 5 ) . (HIV) , particularly in elderly men ( 1 59) or those who
I f the tumor originates i n the visceral pleura, substan are immunosuppressed ( 1 64, 1 65). The tumors have
tial amounts of lung parenchyma may also have to be a large cell morphology and their immunophenotype
removed ( 1 46, 1 56) . Surgical resection cures approxi is null. Nevertheless, they do have a B-cell genotype
mately 90% of these patients ( 1 43, 1 46, 1 49), but recur ( 1 59) . The cells from more than 90% of cases express
rent disease occurs in the remaining 1 0%. Complete the CD45 antigen ( 1 65). It has been shown that the
resection is the best way to prevent recurrences ( 1 56) . normal counterpart of the primary effusion lymphoma
Th e recurrences may occur more than 1 0 years after tumor cells is the mature B cell or preplasma cell.
the initial resection. It is recommended that annual The chest radiographs and CT scans show thicken
chest radiographs be obtained postoperatively in pa ing of the parietal pleura and pericardia! thickening
tients with solitary fibrous tumors to detect recur in many patients. Many patients also have pericardia!
rences early so that they can be surgically removed. effusions, and some patients also have ascites ( 1 59).
The pleural fluid is a lymphocytic exudate charac
terized by a very high lactate dehydrogenase level.
P R I M A RY E F F U S I O N LY M P H O M A
The diagnosis can usually be established with pleu
Primary effusion lymphoma i s an uncommon non ral fluid cytology. The primary effusion lymphoma
Hodgkin's lymphoma that grows in the liquid phase has a distinctive morphology, bridging large cell
204 PLE U RAL D I S EASES
20. Baris YI, Grandjean P. Prospective study of mesothelioma 4 1 . Hillerdal G. Malignant mesothelioma 1 982: review o f 4,7 1 0
mortality in Turkish villages with exposure to fibrous zeolite. published cases. B rJ Dis Chest. 1 983;77:32 1 -343.
J Natl Cancer Inst. 2006;9 8 : 4 1 4--4 1 7. 42. Antman KH . Clinical presentation and natural history
2 1 . Carbone M, Baris YI, Bertino P, et al. Erionite exposure in of benign and malignant mesothelioma. Semin Oncol.
North Dakota and Turkish villages with mesothelioma. Proc 1 9 8 1 ;8:3 1 3-320.
NatlAcad Sci USA. 20 1 1 ; 1 08 : 1 3 6 1 8- 1 3623. 43. Suzuki Y. Pathology of human malignant mesothelioma.
22. Carthew P, Hill RJ, Edwards RE, et al. Intrapleural admin Semin Oncol. 1 9 8 1 ;8:268-282.
istration of fibers induces mesothelioma in rats in the same 44. Pisani RJ, Colby TY, Williams DE. Malignant mesothelioma
relative order of hazard as occurs in man after exposure. Hum of the pleura. Mayo Clin Proc. 1 988;63: 1 234- 1 244.
Exp Toxicol. 1 992; 1 1 : 530-534. 45. Kawashima A, Libshitz HI. Malignant pleural mesothe
23. Antman KH , Corson JM, Li FP, et al. Malignant mesothelioma lioma: CT manifestations in 50 cases. Am J Roentgenol.
following radiation exposure. ] Clin Oncol. 1 9 83; 1 :695-700 . 1 990; 1 5 5 :965-969.
24. Deutsch M , Land S R, Begovic M, et a l . A n association 46. Yilmaz UM, Utkaner G, Yalniz E, et al. Computed tomo
between postoperative radiotherapy for primary breast cancer graphic findings of environmental asbestos-related malignant
in 1 1 National Surgical Adjuvant Breast and Bowel Project pleural mesothelioma. Respirology. 1 998;3 :33-3 8 .
(NSABP) studies and the subsequent appearance of pleural 47. Kreel L. Computed tomography in mesothelioma. Semin
mesothelioma. Am J Clin Oncol. 2007;30:294-296. Oncol. 1 98 1 ; 8 :302-3 1 2 .
2 5 . Roviaro GC, Sartori F, Calabro F, et al. The association of 48. Aisner J, Wiernik P H . Malignant mesothelioma: current sta
pleural mesothelioma and tuberculosis. Am Rev Respir Dis. tus and future prospects. Chest. 1 978;74:438--444.
1 982; 1 2 6 : 5 69-5 7 1 . 49. Metintas M , Ucgun I, Elbek 0, et al. Computed tomog
26. Huncharek M. Genetic factors in the aetiology of malignant raphy features in malignant pleural mesothelioma and
mesothelioma. EurJ Cancer. 1 995;3 1A: l 74 1 - 1 747. other commonly seen pleural diseases. Eur J Radio/. 2002;
27. Pilatte Y, Vivo C, Renier A, et al. Absence of SV40 large 4 1 : 1-9.
T-antigen expression in human mesothelioma cell lines. Am 5 0 . Ng CS, Munden RF, Libshitz HI. Malignant pleural mesothe
J Respir Cell Mol Biol. 2000;23 :788-793. lioma: the spectrum of manifestations on CT in 70 cases. Clin
28. Lopez-Rios F, Illei PB, Rusch V, et al. Evidence against a Radio/. 1 999;54: 1 5-2 1 .
role for SV40 infection in human mesotheliomas and high 5 1 . Benard F, Sterman D , Smith RJ, e t al. Metabolic imaging
risk of false-positive PCR results owing to the presence of of malignant pleural mesothelioma with fluoro
SV40 sequences in common laboratory plasmids. Lancet. deoxygl ucose positron emission tomography. Chest.
2004;364: 1 1 57-1 1 66. 1 998; 1 1 4 :7 1 3-722.
29. Manfredi JJ, Dong J, Liu WJ, et al. Evidence against 52. Wang ZJ, Reddy GP, Gorway MB, et al. Malignant pleural
a role for SV40 in human mesothelioma. Cancer Res. mesothelioma: evaluation with CT, MR imaging, and PET.
2005;65 :2602-2609. Radiographies. 2004;24: 1 05-1 1 9 .
30. Gee GV, Stanifer ML, Christensen BC, et al. SV40 associated 5 3 . Nowak AK, Armato SG Ill, Ceresoli GL, et al. Imaging in
miRNAs are not detectable in mesotheliomas. Br J Cancer. pleural mesothelioma: a review of imaging research presented
20 1 O ; 1 03:885-8 88. at the 9th International Meeting of the International Meso
3 1 . Stricl<ler HD, Goeden JJ, Devesa SS, et al. Trends in US. thelioma Interest Group. Lung Cancer. 2 0 1 0;70: 1 -6.
Pleural mesothelioma incidence rates following simian virus 54. Flores RJ, Akhurst T, Gonen M , et al. Positron emission to
40 contamination of early poliovirus vaccines. J Natl Cancer mography defines metastatic disease but not locoregional dis
Inst. 2003;95:38--45. ease in patients with malignant pleural mesothelioma. J lhorac
32. Price B . Analysis of current trends in United States mesothe Cardiovasc Surg. 2003 ; 1 26: 1 1- 1 5 .
lioma incidence. Am ] Epidemiol. 1 997; 1 45 : 2 1 1-2 1 8 . 5 5 . Gottehrer A, Taryle DA, Reed CE, e t al. Pleural fluid analysis
3 3 . Robinson BW, Lake RA. Advances i n malignant mesothe in malignant mesothelioma. Chest. 1 9 9 1 ; 1 00: 1 003- 1 006.
lioma. N Engl] Med. 2005;353 : 1 5 9 1 - 1 603. 56. Klempman S . The exfoliative cytology of diffuse pleural meso
34. Peto J, Decarli A, La Vecchia C, et al. The European mesothe thelioma. Cancer. 1 962; 1 5 :69 1-704.
lioma epidemic. Br J Cancer. 1 999;79:666-672. 57. Paganuzzi M, Onetto M, Marroni P, et al. Diagnostic value of
3 5 . Jarvholm B, Englund A, Albin M. Pleural mesothelioma in CYFRA 2 1 - 1 tumor marker and CEA in pleural effusion due
Sweden: an analysis of the incidence according to the use of to mesothelioma. Chest. 2001 ; 1 1 9 : 1 1 3 8-1 1 42.
asbestos. Occup Environ Med. 1 999;56: 1 1 0-1 1 3 . 5 8 . Scherpereel A, Grigoriu BD, Conti M, et al. Soluble me
3 6 . Branscheid D , Krysa S, Bauer E , e t al. Diagnostic and thera sothelin-related protein in the diagnosis of malignant pleu
peutic strategy in malignant pleural mesothelioma. Br J ral mesothelioma. Am J Respir Crit Care Med. 2006; 1 73 :
Cardiothorac Surg. 1 9 9 1 ; 5 :466--472. 1 1 5 5- 1 1 60.
37. Corson JM. Pathology of diffuse malignant pleural mesothe 5 9 . Creaney J, Robinson BW Serum and pleural fluid bio
lioma. Semin lhorac Cardiovasc Surg. 1 997;9:347-3 5 5 . markers for mesothelioma. Curr Opin Pulm Med. 2009; 1 5 :
3 8 . Cagle PT, Allen TC. Pathology o f the pleura: What the 366-370.
pulmonologists need to know. Respirology. 20 1 1 ; 1 6:430-438. 60. Davies HE, Sadler RS , Bielsa S, et al. The clinical impact and
3 9 . Corson JM. Pathology of malignant mesothelioma. In: reliability of pleural fluid mesothelin in undiagnosed pleural
Antman K, Aisner J, eds. Asbestos-Related Malignancy. effusions. Am J Respir Crit Care Med. 2009; 1 80:437--444.
Orlando, FL: Grune & Stratton; 1 9 87: 1 79- 1 99 . 6 1 . Creaney J, Yeoman D , Naumoff L, et al. Soluble mesothelin
40. Nash G, Otis C N . Cancer Committee, College o f American in effusions-a useful cool for the diagnosis of malignant me
Pathologists. Protocol for the examination of specimens sothelioma. Thorax. 2007;62 : 569-576.
from patients with malignant pleural mesothelioma: a 62. Nurminen M, Dejmek A, Martensson G, et al. Clinical utility
basis for checklists. Arch Pathol Lab Med. 1 99 9 ; 1 23 : of liquid-chromatographic analysis of effusions for hyaluro
39-44 . nate content. Clin Chem. 1 994;40:777-780.
206 PLE U RAL D I S EASES
63. Hillerdal G, Lindquist U, Engstrom-Laurent A. Hyaluronan in 82. Cagle PT, Churg A. Differential diagnosis o f benign and ma
pleural effusions and in serum. Cancer. 1 99 1 ;67:24 1 0-24 14. lignant mesothelial proliferations on pleural biopsies. Arch
64. Martensson G, Thylen A , Lindquist U, e t al. The sensitivity Pathol Lab Med. 2005 ; 1 29: 1 42 1 - 1 427.
of hyaluronan analysis of pleural fluid from patients with ma 83. Warnock ML, Stoloff A, Thor A. Differentiation of adeno
lignant mesothelioma and a comparison of different methods. carcinoma of the lung from mesothelioma: periodic acid
Cancer. 1 994;73: 1406- 1 4 1 0 . Schiff, monoclonal antibodies B72.3, and Leu M l . Am J
65. Law M R , Hodson M E , Turner-Warwick M. Malignant meso Pathol. 1 98 8 ; 1 33:30-38 .
thelioma of the pleura: clinical aspects and symptomatic treat 8 4 . Jandik WR, Landas S K , Bray CK, e t al. Scanning electron
ment. Eur J Respir Dis. 1 984;65 : 1 62-1 68. microscopic distinction of pleural mesotheliomas from ad
66. Stevens M W, Leong AS , Fazzalari N L , e t a l . Cytopathology o f enocarcinomas. Mod Pathol. 1993;6:76 1-764.
malignant mesothelioma: a stepwise logistic regression analy 85. Burmer GC, Rabinovitch PS, Kulander BG, et al. Flow cy
sis. Diagn Cytopathol. 1 992;8:333-342. romecric analysis of malignant pleural mesotheliomas. Hum
67. Wirth PR, Legier J, Wright GL Jr. Immunohistochemical Pathol. 1 989;20:777-783.
evaluation of seven monoclonal antibodies for differentiation 86. Greillier L, Baas P, Welch JJ, et al. Biomarkers for malig
of pleural mesothelioma from lung adenocarcinoma. Cancer. nant pleural mesorhelioma: current status. Mo! Diagn Ther.
1 9 9 1 ;67:655-662. 2008 ; 1 2:375-390.
68. Frisman DM, McCarthy WF, Schleiff P, et al. lmmunocyto 87. Scherpereel A. Mesothelin as a mesothelioma marker. Int
chemistry in the differential diagnosis of effusions: use of lo Pleural News!. 2006;4: 1 8 .
gistic regression to select a panel of antibodies to distinguish 8 8 . Creaney J, Yeoman D , Demelker Y, e t al. Comparison of
adenocarcinomas from mesothelial proliferations. Mod Pathol. osteopontin, megakaryocyte potentiating factor, and me
1 993;6: 1 79-1 84. sothelin proteins as markers in the serum of patients with
69. Brown RW, Clark GM, Tandon AK, et al. Multiple-marker malignant mesochelioma. J Thorac Oncol. 2008;3:85 1-857.
immunohisrochemical phenotypes distinguishing malignant 89. Pass HI, Lott D, Lonardo F, et al. Asbestos exposure, pleural
pleural mesothelioma from pulmonary adenocarcinoma. Hum mesochelioma, and serum osceopontin levels. N Engl J Med.
Pathol. 1 993;24:347-354. 2005;353: 1 5 64-1 573.
70. Andrion A, Magnani C, Betta PG, et al. Malignant mesothe 90. Scherpereel A, Grigoriu BD, Conti M, et al. Soluble
lioma of the pleura: interobserver variability. J Clin Pathol. mesochelin-relaced protein in the diagnosis of malig
l 995;48:856-860. nant pleural mesothelioma. Am J Respir Crit Care Med.
7 1 . Maskell NA, Gleeson FV, Davies RJ. Standard pleural biopsy 2006; 1 73: 1 1 55-1 1 60.
versus CT-guided cutting-needle biopsy for diagnosis of ma 9 1 . Hassan R, Remaley AT, Sampson ML, et al. Detection and
lignant disease in pleural effusions: a randomised controlled quantitation of serum mesothelin, a tumor marker for pa
trial. Lancet. 2003;36 1 : 1 326-1 330. tients with mesochelioma and ovarian cancer. Clin Cancer
72. Mecintas M, Ozdemir N, lsiksoy S, et al. CT-guided pleural Res. 2006; 1 2:447-453.
needle biopsy in the diagnosis of malignant mesothelioma. 92. Shiomi K, Miyamoto H , Segawa T, et al. Novel ELISA sys
J ComputAssist Tomogr. 1 995; 1 9:370-374. tem for detection ofN-ERC/mesochelin in the sera of meso
73. Agarwal PP, Seely JM, Matzinger FR, et al. Pleural meso chelioma patients. Cancer Sci. 2006;97:923-932.
thelioma: sensitivity and incidence of needle track seeding 93. Creaney J, Olsen NJ, Brims F, et al. Serum mesothelin
after image-guided biopsy versus surgical biopsy. Radiology. for early detection of the asbestos-induced cancer malig
2006;57: 1 20-123. nant mesothelioma. Cancer Epidemiol Biomarkers Prev.
74. Menzies R , Charbonneau M . Thoracoscopy for the diagnosis 20 1 O ; 1 9:2238-2246.
of pleural disease. Ann Intern Med. 1 99 1 ; 1 1 4:27 1-276. 94. Scherpereel A, Astoul P, Baas P, et al. Guidelines of the Eu
75. Hucker J, Bhatnagar NK, Al-Jilaihawi AN, et al. Thoracos ropean Respiratory Society and the European Society of
copy in the diagnosis and management of recurrent pleural Thoracic Surgeons for management of malignant pleural
effusions. Ann Thorac Surg. 1 9 9 1 ;52: 1 1 45-1 1 47. mesochelioma. Eur Respir]. 2009;35 :479-495.
76. Greillier L, Cavailles A, Fraticelli A, et al. Accuracy of pleural 95. Schneider J, Hoffmann H, Dienemann H, et al. Diagnostic
biopsy using thoracoscopy for the diagnosis of hisrologic sub and prognostic value of soluble mesorhelin-related proteins
type in patients with malignant pleural mesorhelioma. Cancer. in patients with malignant pleural mesothelioma in compari
2007; 1 1 0:2248-2252. son with benign asbestosis and lung cancer. J Thorac Oncol.
77. Nagendran M, Pallis T, Patel K, et al. Should all patients who 2008;3 : 1 3 1 7- 1 324.
have mesothelioma diagnosed by video-assisted thoracoscopic 96. Cappia S, Righi L, Mirabelli D, et al. Prognostic role of os
surgery have their intervention sites irradiated? Interact Car teopontin expression in malignant pleural mesothelioma.
diovasc Thorac Surg. 20 1 1 ; 1 3:66-69. Am J Clin Pathol. 2008; 1 30:58-64.
78. Ordonez NG. The immunohistochemical diagnosis of epithe 97. Wheatley-Price P, Yang B , et al. Soluble mesothelin-related
lial mesorhelioma. Hum Pathol. 1 999;30:3 1 3-323. peptide and osteopontin as markers of response in malignant
79. Ordonez NG. What are the current best immunohistochemi mesochelioma. J Clin Oncol. 201 0;28:33 1 6-3322.
cal markers for the diagnosis of epithelioid mesothelioma? 98. Van Gelder T, Damhuis RA, Hoogsteden HC. Prognostic
A review and update. Hum Pathol. 2007;3 8 : 1 - 1 6 . factors and survival in malignant pleural mesothelioma. Eur
80. Granville LA , Younes M , Churg A , e t al. Comparison o f Respir]. 1 994;7: 1 035-1038.
monoclonal versus polyclonal calretinin antibodies fo r 99. Curran D, Sahmoud T, Therasse P, et al. Prognostic factors in
immunohistochemical diagnosis of malignant mesothelioma. patients with pleural mesothelioma: the European Organiza
Appl Immunohistochem Mo! Morphol. 2005 ; 1 3:75-79. tion for Research and Treatment of Cancer experience. J Clin
8 1 . Barberis MC, Faleri M, Veronese S, et al. Calretinin. A se Oncol. 1 998; 1 6 : 1 45-152.
lective marker of normal and neoplastic mesothelial cells in 1 00. West SD, Lee YC. Management of malignant pleural meso
serous effusions. Acta Cytol. 1 997;4 1 : 1 757- 1 76 1 . chelioma. Clin Chest Med. 2006;27:335-354.
CHAPT E R 1 1 I P R I MARY TU M O RS O F TH E PLE U RA 207
1 0 1 . Fennell DA, Parmar A, Shamash J, er al. Statistical validation 1 2 1 . Goudar RK. Management options for malignant pleu
of the EORTC prognostic model for malignant pleural me ral mesothelioma : clinical and cost considerations. Drugs.
sorhelioma based on three consecutive phase II trials. J Clin 2007;67: 1 1 49- 1 1 6 5 .
Oneal. 2005;23 : 1 84- 1 8 9. 1 22 . Treasure T, Sedrakyan A. Pleural mesothelioma: lirrle evi
1 02. Johansson L, Linden CJ. Aspects of hisroparhologic subtype dence, still time to do trials. Lancet. 2004;364: 1 1 83-1 1 8 5 .
as a prognostic factor in 85 pleural mesorheliomas. Chest. 1 2 3 . Van Rurh S, Baas P, Zoermulder FA. Surgical treat
1 996; 1 09 : 1 09- 1 1 4 . ment of malignant pleural mesothelioma: a review. Chest.
1 03 . Pass H I , Temeck B K , Kranda K, er a l . Preoperative 2003; 1 2 3 : 5 5 1-56 1 .
rumor volume is associated with outcome in malig 1 24 . Treasure T, Lang-Lazdunski L, Waller D , er al. Extra-pleu
nant pleural mesorhelioma. ] Thorac Cardiovasc Surg. ral pneumonecromy versus no extra-pleural pneumonec
1 9 9 8 ; 1 1 5 : 3 l 0-3 1 7 . romy for patients with malignant pleural mesorhelioma:
1 04. Flores RM . The role of PET i n the surgical manage clinical outcomes of the Mesorhelioma and Radical Sur
ment of malignant pleural mesorhelioma. Lung Cancer. gery (MARS) randomised feasibility study. Lancet Oneal.
2005 ;49 (suppl l ) : S27-S32. 20 1 1 ; 1 2:763-772.
105. Thylen A, Hjerpe A, Marrensson G . Hyaluronan content in 1 2 5 . Martin-Vear AE, Edwards JG, Rengarajan A, er al. Palliative
pleural fluid as a prognostic factor in patients with malignant surgical debulking in malignant mesothelioma. Predictors
pleural mesorhelioma. Cancer. 200 1 ; 92: 1 224- 1 230. of survival and symptom control. Eur J Cardiothorac Surg.
1 06. Aelony Y, Yao JF, King RR. Prognostic value of pleural fluid 200 1 ;20: 1 1 1 7- 1 1 2 1 .
pH in malignant epithelial mesorhelioma after talc pou 1 26. Halstead JC, Lim E , Venkateswaran RM , e t al. Improved sur
drage. Respiration. 2006;73 :334-339. vival with VATS pleurectomy-decortication in advanced ma
1 07. Demirag F, Unsal E, Yilmaz A, er al. Prognostic significance lignant mesothelioma. EurJ Surg Oneal. 2005;3 1 :3 1 4-320 .
of vascular endothelial growth factor, rumor necrosis, and 1 27. Vogelzang NJ, Rusthoven JJ, Symanowski J, et a l . Phase I I I
mitotic activity index in malignant pleural mesorhelioma. study of pemetrexed i n combination with cisplatin versus
Chest. 2005; 1 2 8 :3382-3387. cisplatin alone in patients with malignant pleural mesorhe
108. Edwards JG, Cox G, Andi A, et al. Angiogenesis is an in lioma. J Clin Oneal. 2003;2 1 :2636-2644.
dependent prognostic factor in malignant mesorhelioma. Br 1 2 8 . Van Meerbeeck JP, Gaafar R, Manegold C, et al. Random
J Cancer. 200 1 ; 8 5 : 863-868. ized phase III study of cisplatin with or without raltitrexed in
1 09 . Gordon GJ, Dong L, Yeap BY, et al. Four-gene expression patients with malignant pleural mesothelioma: an intergroup
ratio rest for survival in patients undergoing surgery for me study of the European Organisation for Research and Treat
sorhelioma. J Natl Cancer Inst. 2009; 1 0 1 : 678-686. ment of Cancer Lung Cancer Group and the National Can
1 1 0. Musk AW, Olsen N, Alfonso H, er al. Predicting sur cer Institute of Canada. ] Clin Oneal. 2005;23 :688 1-6889.
vival for malignant mesothelioma. Eur Respir ]. 20 1 1 ;3 8 : 1 2 9 . O'Brien ME, Watkins D, Ryan C, er al. A randomised trial
1 420- 1 424. in malignant mesorhelioma (M) of early (E) versus delayed
1 1 1 . Butchart EG, Ashcroft T, Barnsley WC, et al. The role of sur (D) chemotherapy in symptomatically stable patients: the
gery in diffuse malignant mesothelioma of the pleura. Semin MED trial. Ann Oneal. 2006; 1 7:270-275.
Oneal. 1 9 8 1 ; 8 : 3 2 1 -328. 1 30 . Ong ST, Vogelzang NJ . Chemotherapy in malignant pleural
1 1 2. Rusch VW, Venkatraman ES. Important prognostic factors mesothelioma. A review. ] Clin Oneal. 1 996; 1 4 : 1 007- 1 0 1 7 .
in patients with malignant pleural mesorhelioma, managed 1 3 1 . Lee JD, Perez S, Wang HJ, e r a l . Intrapleural chemotherapy
surgically. Ann Thorac Surg. 1 999;68 : 1 799- 1 804. for patients with incompletely resected malignant mesotheli
1 1 3 . Jett JR. Malignant pleural mesorhelioma. A proposed new oma: the UCLA experience. ] Surg Oneal. 1 995;60 :262-267.
staging system. Chest. 1 9 9 5 ; 1 0 8 : 89 5-897. 1 32 . de Perrot M, Kurt AM, Robert JH, er al. Clinical behavior
1 1 4 . Law MR, Gregor A, Hodson ME, et al. Malignant mesothe of solitary fibrous tumors of the pleura. Ann Thorac Surg.
lioma of the pleura: a study of 52 treated and 64 untreated 1 999;67: 1456-1459.
patients. Thorax. 1 984;39:25 5-259. 1 33 . Mattson K, Holsri LR, Tammilehro L, er al. Mulrimodality
1 1 5 . Tremblay A, Patel M, Michaud G . Use of tunneled pleu treatment programs for malignant pleural mesothelioma us
ral catheters in malignant mesorhelioma. J Bronchol. ing high-dose hemirhorax irradiation. IntJ Radiat Oneal Biol
2005 ; 1 2 :302-306. Phys. 1 992;24:643-650.
1 1 6. Elmes PC, Simpson MJC. The clinical aspects of mesothe 1 34 . McCormack PM, Nagasaki F, Hilaris BS, et al. Surgical
lioma. QJ Med. 1 976; 1 79:427-449. treatment of pleural mesorhelioma. J Thorac Cardiovasc Surg.
1 1 7. Jackson MB, Pounder D, Price C, er al. Percutaneous cervi 1 982;84: 834-842.
cal cordoromy for the control of pain in patients with pleural 1 3 5 . Asroul P, Picat-Joossen D, Viallat JR, et al. Intrapleural ad
mesorhelioma. Thorax. 1 999;54:238-24 1 . ministration of interleukin-2 for the treatment of patients
1 1 8 . Sugarbaker DJ, Flores RM , Jaklitsch MT, et al. Resection with malignant pleural mesothelioma: a phase II study. Can
margins, extrapleural nodal status, and cell type determine cer. 1 998;83:2099-2 1 04.
postoperative long-term survival in trimodality therapy of 1 36. Marro AL, Fitzpatrick DR, Robinson BW, er al. Antisense
malignant pleural mesothelioma: results in 1 83 patients. oligonucleotides specific for transforming growth factor beta
] Thorac Cardiovasc Surg. 1 999; 1 1 7: 54-6 5 . 2 inhibit the growth of malignant mesorhelioma both in vi
1 1 9. Sugarbaker DJ, Jaklirsch MT, Bueno R , e t a l . Prevention, tro and in vivo. Cancer Res. 1 997;57:3200-3207.
early detection, and management of complications after 328 1 37. Goudar RK. Management options for malignant pleu
consecutive exrrapleural pneumonecromies. J Thorac Cardio ral mesorhelioma: clinical and cost considerations. Drugs.
vasc Surg. 2004; 1 2 8 : 1 38-146. 2007;67: 1 1 49- 1 1 6 5 .
1 20. Aziz T, Jilaihawi A, Prakash D . The management of ma 1 3 8 . Sterman DH. Gene therapy for malignant pleural me
lignant pleural mesorhelioma; single centre experience in sorhelioma. Hematol Oneal Clin North Am. 200 5 ; 1 9 :
10 years. Eur J Cardiothorac Surg. 2002;22:298-305. 1 1 47- 1 1 73 .
208 PLE U RAL D I S EASES
1 39. Sterman DH, Haas A, Moon E, et al. A trial of intrapleural 1 5 8 . Nador RG, Cesarman E, Chadburn A , e t al. Primary effu
adenoviral-mediated interferon-{alpha}2b gene transfer for sion lymphoma: a distinct clinicopathologic entity associated
malignant pleural mesothelioma. Am J Respir Crit Care Med. with the Kaposi's sarcoma-associated herpes virus. Blood.
20 1 1 ; 1 84 : 1 395-1399. 1 996;88 :645-656.
1 40. Cardillo G, Lococo F, Carleo F, et al. Solitary fibrous tumor 1 5 9. Ascoli V, Scalzo CC, Danese C, et al. Human herpes virus-8
of the pleura. Curr Opin Pulm Med. 20 1 2; 1 8 :339-336. associated primary effusion lymphoma of the pleural cavity in
1 4 1 . Fraser RS , Muller NL, Colman N, et al. eds. Pleural neo HIV-negative elderly men. Eur Respir]. 1 999; 1 4 : 1 2 3 1 - 1 234.
plasms. In: Diagnosis ofDiseases ofthe Chest, 4th ed. Philadel 1 60. Uphoff CC, Carbone A, Gaidano G, et al. HHV-8 infection
phia, PA: WB Saunders; 1 999:2807-2847. is specific for cell lines derived from primary effusion (body
1 42. Sandvliet RH, Heysteeg M , Paul MA. A large thoracic mass cavity-based) lymphomas. Leukemia. 1 998; 1 2 : 1 806- 1 809.
in a 57-year-old patient. Chest. 2000; 1 1 7:897-900. 1 6 1 . Castillo JJ, Shum H, Lahijani M, et al. Prognosis in pri
1 43 . Cardillo G, Carbone L, Carleo F, et al. Solitary fibrous tu mary effusion lymphoma is associated with the num
mors of the pleura: an analysis of 1 1 0 patients treated in a ber of body cavities involved. Leuk Lymphoma. 20 1 2; 5 3 :
single institution. Ann 1horac Surg. 2009; 8 8 : 1 632- 1 637. 2378-2382.
1 44. Hutchinson WB, Friedenberg MJ. Intrathoracic mesothe 1 62. Cesarman E, Godwin JH. HIV/AIDS case histories: diag
lioma. Radiology. 1 963;80:937-945 . nostic problems. Primary effusion lymphoma. AIDS Patient
1 4 5 . England DM, Hochholzer L , McCarthy MJ. Localized Care STDS. 1 998; 1 2:403-404.
benign and malignant fibrous tumors of the pleura. A 1 63 . Mbulaiteye SM, Biggar RJ, Goederr JJ, et al. Pleural and peri
clinicopathologic review of 223 cases. Am J Surg Pathol. roneal lymphoma among people with AIDS in the United
1 989; 1 3:640-6 5 8 . States. JAcquir Immune Defic Syndr. 2002;29 : 4 1 8-42 1 .
1 46. Okike N, Bernatz P E , Woolner LB. Localized mesothelioma 1 64. Hisamoro A, Yamane H , Hiraki A, e t al. Human herpes
of the pleura: benign and malignant variants. J 1horac Car virus-8-negative primary effusion lymphoma in a patient
diovasc Surg. 1 978;75 :363-372 . with common variable immunodeficiency. Leuk Lymphoma.
1 47. de Perrot M, Fischer S, Brundler MA, et al. Solitary fibrous 2003;44:201 9-2022.
tumors of the pleura. Ann 1horac Surg. 2002;74:285-293. 1 6 5 . Carbone A, Gloghini A. KSHV/HHV8-associated lympho
1 4 8 . Sung SH, Chang JW, Kim J, et al. Solitary fibrous tumors of mas. Br J Haematol. 2008; 140: 1 3-24.
the pleura: surgical outcome and clinical course. Ann 1horac 1 66. Wakely PE Jr, Menezes G Jr, Nuovo GJ . Primary effusion
Surg. 2005;79:303-307. lymphoma: cyropathologic diagnosis using in situ molecu
1 49. Briselli M, Mark EJ, Dickerson GR. Solitary fibrous tumors lar genetic analysis for human herpesvirus 8. Mod Pathol
of the pleura: eight new cases and review of 360 cases in the 2002; 1 5 : 944-950.
literature. Cancer. 1 9 8 1 ;47:2678-2689. 1 67. Light RW, Hamm H . Pleural disease and the acquired immune
1 5 0. Le Roith D. Tumor induced hypoglycemia. N Engl J Med. deficiency syndrome. Eur Respir J. 1 997; 1 0 :2638-2643.
1 999;34 1 :757-758. 1 68 . Perez CL, Rudoy S. Anti-CD20 monoclonal antibody treat
1 5 1 . Ulrik CS, Viskum K. Fibrous pleural tumour producing 1 7 1 ment of human herpesvirus 8-associated, body cavity-based
litres o f transudate. Eur Respir J. 1 998; 1 2 : 1 230- 1 232. lymphoma with an unusual phenotype in a human immuno
1 52. Rena 0, Filosso PL, Papalia E, et al. Solitary fibrous tumour deficiency virus-negative patient. Clin Diagn Lab Immunol.
of the pleura: surgical treatment. Eur J Cardiothorac Surg. 200 1 ; 8 :993-996.
200 1 ; 1 9 : 1 8 5- 1 89. 1 69. Nakatsuka S, Yao M, Hoshida Y, et al. Pyothorax-asso
1 53 . Cortes J, Rodriguez J, Garcia-Velloso MJ, et al. ( ( 1 8) ciated lymphoma: a review of 1 06 cases. J Clin Oncol.
F]-FDG PET and localized fibrous mesothelioma. Lung. 2002;20:425 5-4260.
2003; 1 8 1 :49-54. 1 70. Petitjean B, Jardin F, Joly B, et al. Pyothorax-associated
1 54. Weynand B, Noel H, Goncette L, et al. Solitary fibrous tu lymphoma: a peculiar clinicopathologic entity derived from
mor of the pleura: a report of five cases diagnosed by trans B cells at late stage of differentiation and with occasional
thoracic cutting needle biopsy. Chest. 1 997; 1 1 2 : 1 424- 1 428. aberrant dual B- and T-cell phenotype. Am J Surg Pathol.
1 5 5 . Takahama M, Kushibe K, Kawaguchi T, et al. Video-assisted 2002;26:724-732.
thoracoscopic surgery is a promising treatment for solitary 1 7 1 . Cheung C, Schonell M, Manoharan A. A variant of pyotho
fibrous tumor of the pleura. Chest. 2004; 1 25 : 1 1 44- 1 1 47. rax-associated lymphoma. PostgradMedf. 1 999;75 : 6 1 3-6 1 4 .
1 56. Magdeleinat P, AJifano M , Perino A, et al. Solitary fibrous tu 1 72. Aruga T, Itami J, Nakajima K , et al. Treatment for pyotho
mors of the pleura: clinical characteristics, surgical treatment rax-associated lymphoma. Radiother Oncol. 2000; 56:59-63.
and outcome. EurJ Cardiothorac Surg. 2002;2 1 : 1 087- 1 093. 1 73 . Ueda T, Andreas C, ltami J, et al. Pyothorax-associated
1 57. Ibrahimbacha A, Farah M, Saluj a ] . An HIV-infected patient lymphoma: imaging findings. AJR Am J Roentgenol.
with pleural effusion. Chest. 1 999; 1 1 6: 1 1 1 3- 1 1 1 5 . 20 1 0 ; 1 94:76-84.
Parapneumon i c E ffusi ons and E mpyema
Despite the advent of potent antibiotics, bacterial all age groups worldwide, confirmed by reports from
pneumonia still results in significant morbidity and the United States, Canada, Europe, and Asia (8) . The
mortality in the American population. The annual mortality rate of empyema has risen alarmingly. In
incidence of bacterial pneumonia is estimated to Utah, death rates from empyema were sixfold higher
be 4 million, with approximately 25% of patients in 2000-2004 compared to 1 950-1 975 (9) . Overall
requiring hospitalization ( 1 ) . Because as many as in the United States, the incidence of empyema per
40% of hospitalized patients with bacterial pneumo 1 00,000 persons had roughly doubled between 1 996
nia have an accompanying pleural effusion (2) , effu and 2008 with roughly equal increases occurring in
sions associated with pneumonia, parapneumonic all age groups ( 1 0) . In this study, the increase was
effusions (PPE) , account for a large percentage of largely due to increases in nonpneumococcal empy
pleural effusions. The morbidity and mortality rates ema and staph empyema ( 1 0) . The explanation for
in patients with pneumonia and pleural effusions are the increase in empyema incidence is not clear but
higher than those in patients with pneumonia alone. has been attributed at least in part to the induction
In one study of 1 ,424 patients hospitalized with com of the heptavalent pneumococcal conjugate vaccine
munity-acquired pneumonia, patients with pleural (PCV7) in 2000. After the introduction of this vac
effusions were 2.7 times more likely to be treatment cine, there was a reduction in invasive pneumococcal
failures than were those without pleural effusions disease, but the incidences of pneumococcal empy
(3) . In another study, the relative risk of mortality ema in children and adults have both increased (8) .
in patients with community-acquired pneumonia The decrease in incidence of empyema from serotypes
was 7.0 times higher for patients with bilateral pleu covered by the vaccine was overcompensated by an
ral effusions and 3.4 times higher for patients with emergence of disease caused by nonvaccine serotypes
unilateral pleural effusion of moderate or greater size (particularly serotype 1 ) (8) .
as compared with other patients with community Most pleural effusions associated with pneumo
acquired pneumonia alone (4) . In assessing risks of nia resolve without any specific therapy directed
patients with community-acquired pneumonia, the toward the pleural fluid (2) , but approximately 1 0%
presence of a pleural effusion is given the same weight of patients require operative intervention. Delay
as a Po 2 less than 60 mm Hg (5) . Espana et al. (6) in instituting proper therapy for these effusions is
recommend that any patient with pneumonia and a responsible for some of the morbidity associated with
loculated effusion or an effusion greater than 2 cm parapneumonic effusions. In one series of 39 patients
in thickness on the decubitus be hospitalized. Some from San Francisco General Hospital selected on the
of the increased morbidity and mortality in patients basis of pus in the pleural space, a positive Gram's
with parapneumonic effusions are due to mismanage stain or culture, a pH of less than 7.0, or a glucose
ment of the pleural effusion (7) . level of less than 40 mg/dL, the mean duration of
Pleural infection (complicated parapneumonic pleural drainage was 2 1 ± 1 8 days, with a mortality
effusion and empyema) is rising in incidence across rate of 1 0% ( 1 1 ) .
209
210 PLE U RAL D I S EASES
pus in the pleural space, but how many white blood PAT H O P H Y S I O LO G I C F EAT U R E S
cells (WBCs) need be present in pleural fluid to make
it pus? Weese et al. (32) defined an empyema as pleu Th e evolution o f a parapneumonic pleural effusion
ral fluid with a specific gravity greater than 1 .0 1 8, a can be divided into three stages, which are not sharply
defined but gradually merge together (34) . First is the
WBC count greater than 500 cells/mm3, or a protein
level greater than 2.5 g/dL. Vianna (33) defined an exudative stage, characterized by the rapid outpour
empyema as pleural fluid on which the bacterial cul ing of sterile pleural fluid into the pleural space. The
tures are positive or the WBC is greater than 1 5 ,000/ origin of this fluid is not definitely known, but it is
mm3 and the protein level is above 3.0 g/dL. Because probably the interstitial spaces of the lung. The origin
many parapneumonic pleural effusions meeting these of the pleural fluid in sheep with Pseudomonas aeru
criteria resolve without operative intervention (2) , I ginosa pneumonia is the interstitial spaces of the lung
prefer to reserve the term empyema for those pleu (35). It is possible that some of the pleural fluid origi
ral effusions with thick, purulent appearing pleu nates in the capillaries in the visceral pleura owing
ral fluid. Of course, some patients with empyema to their increased permeability secondary to the con
have no associated pneumonic process, as shown in tiguous pneumonitis. The pleural fluid in this stage is
Table 1 2. 1 . characterized by a low WBC count, a low LDH level,
The main decision in managing a patient with a and a normal glucose level and pH (36) . If appropri
parapneumonic effusion is whether to insert chest ate antibiotic therapy is instituted at this stage, the
tubes. Therefore, I use the term complicated para pleural effusion progresses no further, and the inser
pneumonic effusion to refer to those effusions that tion of chest tubes is not necessary.
do not resolve without therapeutic thoracentesis If appropriate antibiotic therapy is not instituted,
or tube thoracostomy. Many complicated parap in some instances, bacteria invade the pleural fluid
neumonic effusions are empyemas, but some para from the contiguous pneumonic process, and the
pneumonic effusions with nonpurulent appearing second, fibropurulent stage evolves. This stage is
pleural fluid are also complicated parapneumonic characterized by the accumulation of large amounts
effusions. of pleural fluid with many polymorphonuclear leu
kocytes, bacteria, and cellular debris. Fibrin is depos
ited in a continuous sheet covering both the visceral
and parietal pleura in the involved area. As this stage
TABLE 1 2 . 1 • Eve nt or State Prec i p itat i n g
progresses, there is a tendency toward loculation and
E m pye m a i n 3 1 9 Pat i e nts
the formation of limiting membranes. These loculi
prevent extension of the empyema but make drain
I • • '"' • .. I • age of the pleural space with chest tubes increasingly
difficult. As this stage progresses, the pleural fluid pH
P u l m o n a ry i n fection 1 77 55
S u rg ical p roce d u re 66 21
and glucose levels become progressively lower and the
Tra u m a 18 6 LDH level becomes progressively higher.
Eso p h a g e a l perforation 15 5 The last stage is the organization stage, in which
Sponta n eo u s 7 2 fibroblasts grow into the exudate from both the
p n e u m othorax visceral and parietal pleural surfaces and produce
Thoracentesis 6 2 an inelastic membrane called the pleural peel. This
S u b d i a p h ra g matic 4 inelastic pleural peel over the visceral pleura encases
i nfection the lung and renders it virtually functionless. At this
Septice m i a 4
stage, the exudate is thick, and if the patient remains
M isce l l a neous or 22 7
u n k n own
untreated, the fluid may drain spontaneously through
Tota l 319 1 00 the chest wall (empyema necessitatis) or into the
lung, producing a bronchopleural fistula.
Data from Yeh TJ , H a l l D P, E l l ison R G . E m pye m a Empyemas may arise without an associated pneu
t h o r a c i s : a review o f 1 1 0 cases. A m R e v Respir Dis. monic process. When three series (37-39) totaling
1 9 6 3 ; 8 8 : 7 8 5-790; S n i de r G L. S a i e h S S . E m pyema
3 1 9 cases of empyema are combined (Table 1 2 . 1 ) ,
of t h e t h o rax i n a d u lts: review of 1 0 5 cases. Chest.
1 9 6 8 ; 5 4 : 1 2- 1 7 ; a n d S m ith JA, M u l l e rworth M H ,
most patients had pulmonary infections, but post
West l a k e G W, e t a l . E m py e m a t h o r a c i s : 1 4-yea rs surgical empyemas were also important. A small
experience in a teach i n g center. Ann Thorac Surg. percentage of empyemas complicate thoracentesis
1 99 1 ; 5 1 : 3 9-42 , with p e r m i ss i o n . or tube thoracostomy for pneumothorax, hence the
212 PLE U RAL D I S EASES
necessity for maintaining sterile techniques dur Several conclusions can be made from Table 1 2.2.
ing these procedures. The pleural effusions associ First, aerobic organisms are isolated slightly more fre
ated with esophageal perforation are almost always quently than anaerobic organisms. Second, S. aureus
infected (see Chapter 1 8) . Patients with rheumatoid and S. pneumoniae account for approximately 70% of
pleural effusions frequently develop empyema (see all aerobic gram-positive isolates. Third, when there is
Chapter 2 1 ) ; the genesis of the empyema in this situa a single aerobic gram-positive organism in the pleu
tion is thought to be the formation of a bronchopleu ral fluid, it almost always is S. aureus, S. pneumoniae,
ral fistula through necrotic subpleural nodules (40) . or Streptococcus pyogenes. Fourth, gram-positive aero
bic organisms are isolated approximately twice as
frequently as are gram-negative aerobic organisms.
E X P E R I M E N TA L E M PY E M A
Fifth, although Escherichia coli is the most commonly
There has been surprisingly little work done with isolated gram-negative aerobic organism, it is rarely
experimental empyema. This work is summarized in the lone pathogen isolated from pleural fluid. Sixth,
Chapter 4. Klebsiella sp, Pseudomonas sp, and Hemophilus influen
zae are the next three most commonly isolated aerobic
gram-negative organisms, and these three organisms
BACTE R I O LO G I C F E AT U R E S
account for approximately 75% of all aerobic gram
Th e bacteriologic features o f culture-positive para negative empyemas with a single organism. Seventh,
pneumonic effusions have changed since the intro Bacteroides sp and Peptostreptococcus are the two most
duction of antibiotics. Before the antibiotic era, commonly isolated anaerobic organisms from infected
most empyema fluids grew Streptococcus pneumoniae pleural fluid. Eighth, it is uncommon for a single
or Streptococcus hemolyticus (4 1 ) . Then between anaerobic organism to be isolated from pleural fluid.
1 9 5 5 and 1 9 6 5, Staphylococcus aureus was the bacte The most recent comprehensive report on the bac
ria most commonly isolated from pleural fluid (4 1 ) . teriology of complicated parapneumonic effusions
I n the early 1 970s, anaerobic organisms were most comes from the large trial of intrapleural streptoki
commonly isolated (42) . However, in the 1 9 80s nase in the United Kingdom (46) . In this study of 434
and 1 990s, it appeared that the aerobic organisms patients, the Gram's stain was positive in 250 (58%)
were again responsible for most empyema. Brook patients, the cultures grew a single aerobic growth
and Frazier (43) in 1 993 reviewed the microbiology in 1 5 1 (3 5%) patients, a single anaerobic growth
of 1 97 patients whose pleural fluid was culture posi in 29 (7%) patients, and a polymicrobial growth in
tive for bacteria in two military hospitals. In 64% 52 ( 1 2%) patients (46) . Pleural fluid was available
of patients, only aerobic bacteria were isolated, for molecular microbiologic analysis in 404 of the sub
whereas in 1 3 % of patients, only anaerobic organ jects. For 70 of the culture-negative cases, bacteria were
isms were isolated and, in 23% of patients, both identified by subsequent nucleic acid amplification.
aerobic and anaerobic organisms were isolated. Overall, a micro biologic diagnosis was obtained in 320
Alfrageme et al. (44) again in 1 993 reviewed the patients (74%) . In patients with community-acquired
microbiology of 82 patients treated for empyema pneumonia, the organisms most commonly responsi
at a respiratory unit in Spain and reported results ble were Streptococcus intermedius-anginosus-constellatus
similar to those of Brook and Frazier (43) . Of their (milleri) group in 80, S. pneumoniae 7 1 , and other
76 patients with positive cultures, 62% had exclu streptococcus species 25, S. aureus 34 (7 methicillin
sively aerobic bacteria, whereas 1 6% had exclu resistant Staphylococcus aureus [MRSA]) , gram nega
sively anaerobic bacteria, 1 7% had both aerobic tives 29, and anaerobes 67. In patients with hospital
and anaerobic organisms, and 5% Mycobacterium acquired parapneumonic effusions, the most common
tuberculosis or fungi. organism was S. aureus 2 1 , of which 1 5 were MRSA
The organisms isolated from positive pleural fluid (46) . In a second study (47), Streptococcus intermedius
cultures in three separate series (42,43,45) are tabu anginosus-constellatus (milleri) was also the most com
lated in Table 1 2.2. These series represent 342 patients, mon organism isolated in culture positive complicated
from whom 580 organisms were isolated. Aerobic parapneumonic effusions.
organisms alone were isolated from 1 8 1 patients The bacteriology of complicated parapneumonic
(53%), anaerobic organisms only were isolated from effusions seems to be different in Taiwan. One series
76 patients (22%), and both aerobic and anaerobic in the late 1 990s from Taiwan reported that Klebsiella
organisms were isolated from 85 patients (25%) . pneumonia was isolated from 34 of 1 3 9 patients
C H APTE R 1 2 I PARAPN E U M O N I C E F F U S I O N S A N D E M PY E M A 213
(24.4%) with positive pleural fluid cultures (48) . If population studied. If aspiration is responsible for
the patient is in the intensive care unit, gram-negative the underlying pneumonia, anaerobic organisms are
aerobic organisms are most likely to be responsible, more likely to be responsible (43) . This also explains
with K pneumonia being the most common organ somewhat the high incidence of anaerobes in Bartlett's
ism (49) . The microbiology of empyema in elderly series because their patient population was made up
patients and young adults is similar (50) . of elderly veterans. In contrast, S. pneumoniae is
Several other points should be made concern more likely to be the causative factor in young ambu
ing the bacteriology of infected pleural fluid. First, latory patients whereas in postthoracotomy patients,
to a large part, the incidence of anaerobic isolates is S. aureus is most likely to be responsible.
dependent on the care with which the pleural fluid The bacteriology of infected pleural fluid in
is cultured for anaerobes. The relatively high inci children varies somewhat from that in adults in
dence of anaerobes in the series of Bartlett et al. (42) that H. influenzae is more common and anaerobic
is partially explained by the intense interest these organisms are less common. In one study of 72 cul
investigators had in culturing anaerobes. Second, ture-positive pleural fluids, aerobic organisms were
the organisms cultured depend somewhat on the found in 48 (67%) , anaerobic organisms were found
214 PLE U RAL D I S EASES
in 17 (24%) , and mixed aerobes and anaerobes fluid cultures were positive for anaerobic organisms.
were found in 7 ( 1 0%) (5 1 ) . The most commonly Aerobic organisms were also cultured from the pleu
isolated organisms in this series were H. influenzae ral fluid in 1 8 (40%) of the patients with positive
( 1 5 isolates) , Bacteroides sp ( 1 5) , S. pneumoniae ( 1 3), anaerobic pleural fluid cultures. Some patients with
S. aureus ( I O) , and anaerobic cocci (9) . In another anaerobic pleural infection have no concomitant
series of 1 73 culture-positive pleural fluids in children parenchymal disease.
younger than 1 5 years, 38% were due to S. aureus,
28% were due to S. pneumoniae, 23% were due to G RA M - P O S IT I V E BACT E R I A
H. injluenzae, and 1 1 % were due to other organisms.
In this series, anaerobic isolates were rare (52) . S. pneumoniae is still responsible for many bacterial
pneumonias, and many patients have an associated
pleural effusion. Taryle et al. (54) studied 53 patients
I N C I D E N C E O F P L E U RA L E F F U S I O N S
with pneumococcal pneumonia and found that 57%
W I T H VA R I O U S BACT E R I A L
had an associated parapneumonic effusion, whereas my
PN E U M O N IAS
colleagues and I found that 40% of 1 53 patients with
Once a patient has a bacterial pneumonia, the inci pneumococcal pneumonia had an associated pleural
dence of associated pleural effusion and the frequency effusion (2) . In a more recent paper, 40% of 52 patients
with which the pleural fluid becomes infected largely with bacteremic pneumococcal pneumonia had pleu
depend on the infecting organism (Table 1 2.3) . ral effusions as compared with 2 1 % of patients with
Infected pleural fluid is most common in anaerobic pneumococcal pneumonia without bacteremia (5 5).
pneumonia. In one series of 1 43 patients with anaer Pleural fluid cultures are usually negative in patients
obic infections of the lung (53) , 50 (3 5%) had pleu with pneumococcal parapneumonic effusions. Of the
ral effusions, and in 47 (94%) of these, the pleural 8 1 patients with pleural effusions in the foregoing two
Anaerobic (53) 35 90
G ra m - positive
Streptococcus (2, 54) 40-60 1 -5
pneumonia e
Staphylococcus aureus
Adu lts ( 5 8 , 5 9) 50 20
C h i l d ren (54, 57) 70 80
Streptococcus pyogenes (63,63) 5 5-9 5 3 0-40
Bacillus an thracis (7 5 , 7 6) 90-1 00 2 0- 1 00
Aero b i c g ra m -negative
Escherichia coli (65) 40 80
Pseudomonas (66) 2 5-50 40-50
Klebsiel/a pneumonia e (69) 50 50
Haemophilus influenza
Adu lts (64, 7 2 , 7 3 ) 1 0-45 20
C h i l d ren ( 5 9 , 6 3 , 7 0, 7 1 ) 75 80
Proteus species (74) 20 50
Legionella species (79-82) 2 5-60 ?
series, only 3 (4%) had pleural fluid cultures that were G RA M - N E G AT I V E BACT E R I A
positive for S. pneumoniae. Nevertheless, as shown
in Table 1 2.2. S. pneumoniae is responsible for many Of pneumonias due to gram-negative aerobic organ
positive pleural fluid cultures. The explanation for isms, those caused by E. coli are most likely to have
this apparent paradox is the fact that such a large per complicated parapneumonic effusions. In one series
centage of pneumonias is due to S. pneumoniae. The of 20 patients (65), 40% had pleural effusion, and
incidence of parapneumonic effusions is higher when in 6 of these 8 patients, pleural fluid cultures were
patients wait 48 hours or more after the development positive. All eight patients with pleural effusion in
of symptoms before seeking medical attention (54) . this series had to be treated by tube thoracostomy or
Pneumonia secondary to S. aureus is likely to have open thoracotomy. Rarely, however, is E. coli the sole
an accompanying culture-positive pleural effusion. isolate from pleural fluid (Table 1 2.2) . Patients with
Pseudomonas pneumonia are also likely to have pleural
Indeed, in one study of the causes of pleural effusion
in children, staphylococcal empyema was the most effusions. In one series of 56 patients with ventilator
frequent cause (56) . Wolfe et al. (56) reviewed 98 associated P. aeruginosa pneumonia, 1 3 (23%) had a
children with pleural effusions seen at Duke Uni pleural effusion and 7 ( 1 2.5%) developed empyema
versity between 1 9 52 and 1 967 and reported that (66) . In another series of 28 patients with nosoco
S. aureuswas responsible for 35 (36%) of the effusions.
mial pneumonia due to P. aeruginosa, 1 3 (46%) had
In a series of 75 cases of staphylococcal pneumonia bilateral effusions and an additional 5 ( 1 8%) had a
in infants and young children (57) , more than 70% unilateral effusion (67) . As evident in Table 1 2.2, Pseu
had pleural effusions, and the pleural fluid cultures domonas sp and E. coli account for more than 50% of
were positive in approximately 80%. In adults, more all aerobic gram-negative isolations from pleural fluid.
than 50% of patients with staphylococcal pneumonia In Taiwan, K pneumonia is the most frequent cause
will have an accompanying pleural effusion (58), and of community-acquired thoracic empyema or compli
effusions are more common with MRSA than with cated parapneumonic effusion accounting for 40 of
MSSA. Pleural fluid cultures are positive in at least 1 69 positive cultures in one study (68) . Pleural effu
20% of adults with pleural effusions (59) . In one sion occurs in about 50% of patients with Klebsiella
pneumoniae (69) . The mortality rate is significantly
series (60) of 14 patients with community acquired
MRSA pneumonia, pleural effusions were present in higher in patients with pleural fluid cultures positive
nine patients (64%) and pleural fluid cultures were for K pneumoniae than it is for patients with cultures
positive in 5 of the 9 patients (55%) with pleural positive for other organisms (68) .
effusion. Patients who have right-sided endocarditis In recent years, H. influenzae has been respon
from S. aureus frequently have pleural effusions, but sible for an increasing number of pneumonias in
the cultures are positive in only a small percentage both children (52,70,7 1 ) and adults (72) . With
H. influenzae pneumonia, the pleura is frequently
(6 1 ) . In this situation, the effusions are exudates with
a very high pleural fluid LDH. involved, particularly in children (7 1 ) . In a series of
Pneumonias due to S. pyogenes are uncommon, but 65 cases of pneumonia in children, 49 (75%) had
they are associated with parapneumonic effusion in pleural effusions, and the cultures were positive in
most cases. Welch et al. (62) reported that 95% of20 36 of 46 patients (78%) (7 1 ) . In one large series
patients had an associated pleural effusion, whereas (73) of 2 1 1 patients with H. influenzae pneumo
Basiliere et al. (63) reported that 57% of 95 patients nia, only 22 patients ( 1 0 .4%) had a pleural effusion.
with streptococcal pneumonia had a pleural effusion. Proteus sp causes a substantial proportion of gram
The pleural fluid cultures are positive in 30% to 40% negative pneumonias, but associated pleural effusions
of those with pleural effusion (62,63) . The pleural are uncommon, and when they are present, they are
effusions secondary to streptococcal pneumonia are usually small and uncomplicated (74) .
located more commonly on the left side. Of the 73
pleural effusions in the foregoing series, nearly two
A N T H RAX
thirds were on the left side. Streptococcal pneumo
nia occurs in epidemics, particularly among military Bacillus anthracis is a large gram-positive, spore
recruits (63) . In some patients, the development of forming, rod-shaped organism that may contami
the pleuritis is explosive with this organism. Patients nate goat hair, wool, or animal hides (75 ) . Although
may develop large pleural effusions with low glucose only one case of anthrax was reported in the United
levels and pH in less than 1 2 hours (64) . States between 1 980 and 2000, interest has been
216 PLE U RAL D I S EASES
rekindled in this organism with its use as a bioter started. The pleural fluid cytologic specimen should
rorism agent in Washington, D.C. in 200 1 (76) . be stained immunohistochemically for anthrax.
This virulent organism causes pulmonary disease
when the spores are inhaled into the alveoli, are M I S C E L LA N E O U S PAT H O G E N I C
engulfed by alveolar macrophages, and are carried to O RGAN I S M S
the hilar lymph nodes, where they multiply in their
vegetative state. After causing flu-like symptoms for Pleural effusions may also occur in 30% to 65%
several days, the bacteria are disseminated hematog of patients with pneumonias due to Legionella sp
enously. This dissemination is marked by the acute (75 ,76,79) . In one series (80) of 43 patients with
onset of dyspnea, cyanosis, tachycardia, fever, and Legionnaire's disease, pleural effusions were present
shock. In the cases in Washington, the medium time in 1 0 patients (23%) on admission, whereas another
from exposure to symptoms was 4 days and then 14 developed effusions during the first week after
patients did not seek medical attention for another admission, and 3 new effusions were discovered after
3 . 5 days (76) . The characteristic radiologic findings the first week. In some cases, the organisms can be
are mediastinal widening, patchy nonsegmental pul demonstrated by direct immunofluorescence or cul
monary infiltrates, and unilateral or bilateral pleural ture of the pleural fluid (8 1 ) . Usually, the pleural effu
effusions. sions are small and clinically unimportant, but one
Pleural effusions were present in all 1 0 of these patient had a multiloculated pleural effusion due to
patients treated in Washington, D.C. with the bio Legionella sp and required a decortication (82) .
terrorism attack (76) and 7 patients required drain Several unusual organisms should be considered
age. Indeed, when Kyriacou et al. (77) compared the in patients with pneumonia and pleural effusions.
clinical characteristics of 47 cases of anthrax with Tularemia may be manifested as a pneumonia, and,
those of 376 patients with community-acquired if so, there is frequently an accompanying pleural
pneumonia or influenza, they found that the most effusion (83) . Interestingly, the pleural fluid in asso
accurate prediction was the presence of a pleural ciation with tularemia is a lymphocyte predominant
effusion or mediastinal widening on the chest radio exudate with a high adenosine deaminase (ADA)
graph. All 47 patients with anthrax had a pleural level (83) . Pleural effusions occurred in 5 to 33% of
effusion and/or mediastinal widening (77) . In the the cases of acute melioidosis (84) . As with tularemia,
recent outbreak, the pleural fluid was a bloody exu the pleural fluid with melioidosis can be lympho
date (the pleural fluid red blood cell [RBC] count cyte predominant (84) . Clostridial pleuropulmonary
was above 70,000 cells/mm3 in all) , with a relatively infections are uncommon; by 1 970, only 1 7 cases
low WBC count (200-3 ,000 cells/mm3) and an had been reported (85). Almost all patients with
LDH that varied from 282 to 1 ,762 IU/L (no upper clostridial pulmonary infections have a pleural effu
limit of normal for serum provided) (76) . lmmu sion that is culture positive (85) . Complicated para
nohistochemical stains for B. anthracis were posi pneumonic pleural effusions have also been reported
tive on all pleural fluid cytologic and pleural tissue with pneumonias due to brucellosis (86) , Hemophilus
specimens (76) . parainfluenzae (87) , Bacillus cereus (88) , Citrobacter
It is important to make the diagnosis of anthrax diversus (89) , and Listeria monocytogenes (90) and can
early. In the outbreak in Washington, 6 of the 1 0 probably occur with any bacterium that is a pathogen
patients received antibiotics before they entered the in humans.
fulminant stage and all survived. In contrast, none
of the four patients who received antibiotics after
they entered the fulminant stage survived (76) . In C L I N I CA L M A N I F E STAT I O N S
a review of the world literature, Holty et al. (78) The clinical manifestations of parapneumonic effu
reported that the mortality rate was 97% in patients sions and empyema depend to a large part on whether
who reached the fulminant stage. The only patient the patient has an aerobic or anaerobic infection.
who survived was a veterinarian who might have
had partial immunity (78) . This diagnosis should
Aerobic Bacterial Infections
be considered in all patients with an acute illness
and with mediastinal widening or pleural effusions The clinical presentation of patients with aerobic bac
(77) . After blood cultures are obtained, appropriate terial pneumonia and a pleural effusion is no differ
antibiotics, for example, ciprofloxacin, should be ent from that of patients with bacterial pneumonia
C H APTE R 1 2 I PARAPN E U M O N I C E F F U S I O N S A N D E M PY E M A 217
scans of 50 patients with parapneumonic effusions bacterial cultures, higher yields will be obtained if the
and reported that 1 8 (36%) had lymph nodes in their pleural fluid is directly inoculated into blood culture
mediastinum that were greater than 1 cm in diameter. bottles at the time of thoracentesis ( 1 00, 1 0 1 ) .
The possibility of a pleural effusion can also be Th e pleural fluid cultures i n patients with para
evaluated by ultrasound. Ultrasound has two advan pneumonic effusions are frequently negative, even
tages: first, it is portable and can be performed eas when the fluid is pus. To identify the organism
ily in the intensive care unit and second, it also will responsible for the pneumonia, nuclei acid amplifica
delineate whether the pleural fluid is septated. How tion has been used to identify the bacteria responsible
ever, in one study of 50 patients with parapneumonic for a complicated parapneumonic effusion. Maskell
effusions, there was no relationship between the et al. ( 1 02) performed this procedure on 404 pleu
appearance on ultrasound and whether the patient ral fluid specimens obtained during the First Multi
would need surgical treatment (96) . In contrast, a center lntrapleural Sepsis Trial ( 1 03) . They reported
second study (97) of 1 4 1 patients with complicated that the nucleic acid amplification technique identi
parapneumonic effusions found that patients with a fied bacteria in 70 samples which were negative on
complex nonseptated sonographic pattern were more culture ( 1 02) .
frequently successfully treated with a small-bore cath Not all patients with an acute illness, parenchymal
eter 48/60 (80%) than were patients with a complex infiltrates, and pleural effusion have an acute bacte
septated sonographic pattern 4 1 1 8 1 (5 1 %) . rial pneumonia; pulmonary embolization, acute pan
The amount of free pleural fluid can be semiquan creatitis, tuberculosis, Dressler's syndrome, and other
titated by measuring the distance between the inside diseases can produce identical pictures. The possi
of the chest wall and the bottom of the lung on either bility of pulmonary embolization should always be
the decubitus radiograph or the CT scan of the chest. considered if the patient does not have purulent spu
This distance can also be measured with ultrasound. tum or a peripheral leukocytosis above 1 5 ,000/mm3•
If this distance measures less than 10 mm, one can Most patients with acute tuberculous pleuritis have
assume that the effusion is not clinically significant no infiltrate on the decubitus film with the involved
and, therefore, a thoracentesis is not indicated. My side superior or on the chest CT scan.
colleagues and I reported that 53 patients with acute The pleural fluid with parapneumonic effusions
bacterial pneumonia had such small effusions, and in varies from a clear, yellow exudate to thick, foul
each of the patients, the pneumonia and the pleu smelling pus. If the odor of the pleural fluid is feculent,
ral effusion cleared with only antibiotics and left no the patient is likely to have an anaerobic pleural infec
residual pleural disease (2) . Moffet et al. (98) have tion (53, 1 04) . Although Sullivan et al. ( 1 04) reported
demonstrated that a pleural fluid thickness of 2.5 cm that 1 1 % of aerobic empyemas were described as foul
on a CT scan correlates to a thickness of 1 .0 cm on smelling, it is probable that these represented mixed
the decubitus radiograph. Moreover, a recent article aerobic and anaerobic pleural infections in that sophis
(99) suggested that a thoracentesis is indicated only ticated anaerobic culture techniques were not used
if the thickness of the pleural fluid is greater than in this study. Only approximately 60% of anaerobic
20 mm on the CT scan because effusions smaller empyemas have a foul odor (53, 1 04) . If frank pus is
than this are rarely complicated. obtained with diagnostic thoracentesis, a pleural fluid
If the thickness of the fluid is greater than pH determination should not be done. When thick,
1 0-20 mm, a therapeutic thoracentesis should be purulent material is processed through blood gas
performed immediately because it is impossible to machines, it is likely to plug up the machine or dam
separate complicated from uncomplicated effusions age the membranes. Once laboratory personnel have
without a thoracentesis. The pleural fluid is exam this experience with one such pleural fluid, they are
ined grossly for color, turbidity, and odor. Aliquots hesitant to process additional pleural fluids. The dif
are sent for determination of the pleural fluid glu ferential WBC on the pleural fluid usually reveals pre
cose, LDH, and protein levels, pH (must be analyzed dominantly polymorphonuclear leukocytes. If many
with a blood gas machine) , and differential and total small lymphocytes, mesothelial cells, or macrophages
WBC counts. Samples of pleural fluid are also sent are seen, alternate diagnoses should be considered. If
for bacterial cultures, both aerobic and anaerobic, food particles are seen in the pleural fluid, the patient
and for Gram's stain, as well as for cytologic stud has an esophageal-pleural fistula ( 1 05).
ies and mycobacterial and fungal smears and cultures, Not all patients with parapneumonic effu
if clinically indicated. When pleural fluid is sent for sions have an acute illness, so the possibility of a
C H APTE R 1 2 I PARAPN E U M O N I C E F F U S I O N S A N D E M PY E M A 219
Complicated
4,500 •
Event o r State N u m be r
L D H , l a ct i c d e h y d ro g e n a s e .
rheumatoid disease, malignant tumors, and tubercu The one situation in which the pleural fluid pH
losis (30) , may also have a low pleural fluid pH or is not reduced in complicated parapneumonic effu
glucose level and usually do not need to be treated by sion is when the offending organism is of the Pro
tube thoracostomy. teus sp. These organisms produce ammonia by their
If one uses the pleural fluid pH as a guide for the urea-splitting ability, which can lead to an elevated
placement of chest tubes, it must be measured with pleural fluid pH. Pine and Hollman ( 1 09) reported
a blood gas machine. If the pleural fluid pH is mea three cases of complicated parapneumonic effusions
sured with a pH meter or with a pH indicator strip due to Proteus organisms in which the pleural fluid
paper, the results are not sufficiently accurate ( 1 08) . pH exceeded 7.8.
Moreover, because the pleural fluid pH is influenced In the natural evolution of a parapneumonic
by the arterial pH (30) , the arterial pH should be effusion, the pleural fluid pH falls before the glu
measured before the pleural fluid is drained, if it is cose level falls (30, 1 1 0) and, therefore, the pH is a
to be drained solely on the basis of the pleural fluid more sensitive indicator of complicated parapneu
pH. To serve as a definite indication for tube thora monic effusion than the pleural fluid glucose level.
costomy, the pleural fluid pH should be at least 0.30 The lowered pH with complicated parapneumonic
units less than the arterial pH. effusions appears to be caused by the metabolism of
C H APTE R 1 2 I PARAPN E U M O N I C E F F U S I O N S A N D E M PY E M A 22 1
P l e u ra l Risk of
P l e u ra l Space Pleural F l u i d Fluid Poor
Anatomy Bacteriology C h e m i stry Category O utco m e D ra i n a g e
glucose by the leukocytes in the pleural fluid, result reviewed the clinical courses of 26 patients who had
ing in increased levels of lactate and carbon diox a pH of less than 7.20, a positive Gram's stain, or a
ide in the pleural fluid ( 1 1 0) . When some loculi positive culture. Sixteen patients were initially treated
are infected and others are sterile in patients with with intravenous antibiotics alone without tube tho
loculated pleural effusions, the carbon dioxide prob racostomy, whereas the remaining 1 0 were treated
ably equilibrates across the fibrin membranes, sepa with tube thoracostomy plus intravenous antibiot
rating the different loculi, more readily than does ics. Only 2 of the 16 patients treated with antibiot
the glucose. Accordingly, pleural fluid acidosis is ics alone subsequently required tube thoracostomy.
present in most loculi, although some loculi may Three of the four patients with pleural fluid pH less
contain nearly normal glucose levels ( 1 1 1 ) . It should than 7.00 never required tube thoracostomy. The
be noted that if the pleural fluid is loculated, there mean duration of hospitalization was longer in the
might be significant differences in the pleural fluid group that received a chest tube immediately. Poe et
pH from one locule to another. Maskell et al. ( 1 1 2) al. ( 1 1 4) reviewed 9 1 patients with parapneumonic
measured the pleural fluid pH from two or more effusions and concluded that measurement of the
locules in seven patients with parapneumonic effu pleural fluid glucose, pH, and LDH has limited
sions and reported that the pleural fluid pH varied usefulness in predicting the need for eventual chest
markedly from locule to locule. Three of the seven tube drainage or decortication, or both. However, if
patients had pHs both above and below 7.20 in dif their data are examined closely ( 1 1 5) , this conclusion
ferent locules ( 1 1 2) . The pleural fluid glucose levels is not supported. When patients with frank empy
were markedly reduced in all pleural fluids in which emas are excluded, 1 0 of 1 8 patients (56%) who had
it was measured ( 1 1 2) . a pleural fluid pH value below 7.00 or a pleural fluid
It should b e mentioned that there is not univer glucose below 40 mg/dL received chest tube drain
sal agreement concerning the usefulness of using the age. In contrast, only 8 of 52 patients ( 1 5%) who did
pleural fluid glucose and pH as indicators for pleu not meet these criteria underwent tube thoracostomy
ral fluid drainage. Berger and Morganroth ( 1 1 3) (p < 0.005).
222 PLE U RAL D I S EASES
Heffner et al. ( 1 1 6) performed a meta-analysis severe morbidity associated with delayed tube thora
regarding the ability of the pleural fluid levels costomy justifies the placement of a few extra chest
of pH ( n = 2 5 1 ) , LDH ( n = 1 1 4) , and glucose tubes.
( n = 1 3 5) to identify those parapneumonic effu Another possible marker for a complicated para
sions that needed drainage. In general, they found pneumonic effusion is polymorphonuclear elastase
that those effusions that were drained had a lower (PMN-E) . Aleman et al. ( 1 1 9) measured the levels
glucose, a lower pH, and a higher LDH than did of PMN-E in 1 2 5 patients with parapneumonic
those that were not drained, but there was much effusions including 42 typical parapneumonic effu
overlap ( 1 1 6) . In general, the pleural fluid pH was sions, 1 7 borderline complicated parapneumonic
a little better at making the differentiation than was effusions, and 66 complicated parapneumonic effu
the pleural fluid glucose or LDH, but there was sions or empyemas. They reported that a pleural
much overlap in the values for all three measure fluid level of PMN-E greater than 3,500 ,ug/mL was
ments between those patients who received drain more sensitive and specific than the pleural fluid
age and those who did not. Overall, one encounters pH or glucose in identifying parapneumonic effu
several problems when trying to assess the value sion in which the cultures were positive ( 1 1 9) . This
of these measurements. First, the individual who same group has stated that a PMN-E greater than
made the decision to institute pleural drainage 3,000 ,ug/mL is useful in separating complicated and
knew the results of the biochemical tests. Second, noncomplicated parapneumonic effusions ( 1 20) . To
the upper limit of LDH may have varied from my knowledge, no other group has confirmed these
institution to institution. Third, some of the pleu findings.
ral fluid pHs may have not been measured with a There have been several articles assessing the util
blood gas machine (substantial number of pleural ity of other markers in pleural fluid to distinguish
fluid pHs above 7 . 5 0 ) . complicated from uncomplicated parapneumonic
Jimenez Castro e t al. ( 1 1 7) evaluated the utility effusions. Markers evaluated have included interleu
of the pleural fluid glucose, pH, LDH, and volume kin-8 ( 1 2 1 ) , C-reactive protein ( 1 2 1- 1 24) , trigger
in identifying patients who required tube thoracos ing receptor expressed on myeloid cells (sTREM- 1 )
tomy among 238 patients admitted to one hospi ( 1 2 1 ) , procalcitonin ( 1 2 1 ) , lipopolysaccharide
tal. They found that the pleural fluid pH had the binding protein ( 1 24), matrix metalloproteinase
highest diagnostic accuracy, followed by the pleu (MMP)-2 ( 1 25), MMP-8 ( 1 25), MMP-9 ( 1 25),
ral fluid glucose, LDH, and pleural fluid volume 8-isoprostane ( 1 26), and Cu/Zn superoxide dis
( 1 1 7) . They reported that the optimal binary deci mutase ( 1 26) . None of these have proven to be
sion threshold was less than or equal to 7. 1 5 for the superior to the pleural fluid pH, glucose, or LDH
pleural fluid pH, less than 72 mg/dL for the pleural in differentiating complicated from uncomplicated
fluid glucose, and greater than 865 IU/L for the parapneumonic effusions ( 1 27) .
pleural fluid LDH (upper normal limit for serum
300 IU/L) ( 1 1 7) .
LO C U LAT E D P L E U RA L E F F U S I O N S
In view o f the factors mentioned i n the preceding
text, there is no doubt that some patients with para Pleural effusions are already loculated when some
pneumonic effusions that have a pleural fluid pH patients with pneumonia and pleural effusion are
below 7.00, a pleural fluid glucose below 40 mg/ first evaluated. Although small amounts of freely
dL, a positive Gram's stain or a positive pleural fluid moving fluid can be demonstrated in most patients
culture can be cured with antibiotics alone. Never with loculated pleural effusion, such is invariably
theless, it is recommended that the pleural fluid be not the case. Loculated pleural effusions manifest as
drained in patients with parapneumonic effusions, pleural-based masses without air bronchograms on
who have a pleural fluid pH below 7.00, a pleural the standard chest radiograph (Fig. 6.5). Frequently,
fluid glucose below 40 mg/dL, or a positive pleural it is difficult to distinguish pleural fluid loculi from
fluid Gram's stain ( 1 1 5) because these are indicators peripheral parenchymal infiltrates on standard chest
that it is likely that the parapneumonic process will radiographs. Ultrasonic techniques are effective in
not resolve with antibiotics alone. It should be noted distinguishing pleural fluid loculi from parenchy
that the lower the pleural fluid pH or the pleural mal infiltrates ( 1 28, 1 29). As little as 5 mL of locu
fluid glucose, the less likely the effusion will resolve lated pleural fluid can be identified by ultrasound.
without tube thoracostomy ( 1 1 8) . The risk of more Therefore, if a loculated pleural effusion is suspected,
C H APTE R 1 2 I PARAPN E U M O N I C E F F U S I O N S A N D E M PY E M A 223
parapneumonic effusions require no invasive pro or thoracotomy with decortication are necessary
cedure other than the initial thoracentesis to delin in the majority.
eate the characteristics of the pleural effusion (4).
If a Class 2 effusion rapidly enlarges in size or if the
Classification of the American College of
patient remains toxic with significant pleural fluid,
Chest Physicians
then a repeat thoracentesis should be performed.
Class 3: Borderline Complicated Parapneumonic In 2000, the ACCP developed a classification of para
Effusion. Patients with Class 3 parapneumonic pneumonic effusions on the basis of the anatomical
effusions have negative bacterial smears and cul characteristics of the fluid (A) , the bacteriology of
tures and a glucose level above 40 mg/dL, but the the pleural fluid (B) and the chemistries (C) of the
pH is between 7.00 and 7.20, the LDH is above pleural fluid ( 1 07) . This classification is somewhat
3 times the upper limit of normal, or the pleu analogous to the TMN classification used to classify
ral fluid is loculated. The relatively low pH, the tumors. The classification is shown in Table 1 2.6.
relatively high LDH, and the loculated effusion all The anatomy (A) of the pleural effusion is based
indicate a high level of inflammation in the pleural on the size of the effusion, whether it is free flowing
space. Some Class 3 pleural effusions resolve with and whether the parietal pleural is thickened. � effu
no invasive procedure, whereas others do not. sions are small effusions ( < 1 0 mm in thickness on
Class 4: Simple Complicated Parapneumonic Effu the decubitus radiographs, ultrasound examination,
sion. Patients with Class 4 parapneumonic effu or CT scans) and are free flowing. A1 effusions are
sions have a pleural fluid pH less than 7.00, a pleural greater than 1 0 mm in thickness but occupy less than
fluid glucose level less than 40 mg/L, or a positive 50% of the hemithorax, are free flowing, and are not
Gram's stain or culture. The pleural fluid does not associated with parietal pleural thickening. A2 effu
look like pus and it is not loculated. Patients with sions occupy more than 50% of the hemithorax or
Class 4 parapneumonic effusions should be treated are loculated and/or are associated with thickening of
with some form of invasive therapy because many the parietal pleura. I tend to ignore the thickening
will not resolve solely with antibiotics. of the parietal pleura in this classification because it
Class 5: Complex Complicated Parapneumonic Effu has been shown that thickening of the parietal pleura
sion. Patients with Class 5 parapneumonic effu on CT scan is not related to the outcome with a
sions meet the criteria for Class 4 parapneumonic parapneumonic effusion (96) .
effusions, but, in addition, the fluid is loculated. The bacteriology (B) of the effusion is based on
These patients require fibrinolytics or thoracos whether smears or cultures are positive. Bx effusions
copy to break down the adhesions, and some of the are those in which the culture and Gram's stain results
patients require thoracotomy with decortication. are unknown, presumably because the effusion was
Class 6: Simple Empyema. Patients with Class 6 small and a thoracentesis was not done. B0 effusions
parapneumonic effusions have pleural fluid that have negative Gram's stains and cultures of the pleural
is frank pus, which is either free flowing or con fluid. B1 effusions are those in which the Gram's stain
fined to a single loculus. These patients should or culture are positive, but the pleural fluid is not pus.
be treated with a chest tube. Patients who have B2 effusions are those where the pleural fluid is pus.
Class 6 parapneumonic effusions frequently have The chemistry (C) of the effusion is based on the
a thick peel over the visceral pleura that prevents pH of the pleural fluid. Cx effusions are those on
the underlying lung from expanding. If a sizable which the pleural fluid pH is unknown, presumably
empyema cavity remains after several days of chest because a thoracentesis was not done. C0 effusions
tube drainage, consideration should be given to are those with a pleural fluid pH greater than 7.20.
performing a decortication to eradicate the empy C1 effusions are those with a pleural fluid pH less
ema cavity. than 7.20. To obtain an accurate pleural fluid pH,
Class 7: Complex Empyema. Patients with Class 7 the pleural fluid must be measured with a blood gas
parapneumonic effusions have frank pus in their machine ( 1 08). If a pleural fluid pH measurement
pleural space that is multiloculated. Although with a blood gas machine is not available, an alter
these patients should initially be treated with chest native measurement is the pleural fluid glucose level
tubes and attempts can be made to facilitate drain with a cutoff level of 60 mg/dL.
age with fibrinolytics, more invasive measures such On the basis of the A, B, and C classification, the
as thoracoscopy with the breakdown of adhesions effusion is categorized. Category 1 effusion is a small
C HAPT E R 1 2 I PARAPN E U M O N I C E F F U S I O N S A N D E M PY E M A 22 5
( < 1 0 mm thickness on decubitus, CT scan or ultra Metronidazole penetrated most easily, followed by
sound studies) and free-flowing effusion. Because penicillin, clindamycin, vancomycin, ceftriaxone,
the effusion is small, no thoracentesis is performed and gentamicin (Fig. 1 2 .2) ( 1 35). Subsequent studies
and the bacteriology and chemistry of the fluid are demonstrated that the quinolones, clarithromycin,
unknown. The risk of a poor outcome with a category azithromycin, linezolid, and ertapenem penetrate the
1 effusion is very low. infected pleural space well ( 1 36-1 39). This variance
Category 2 effusion is small to moderate in size in the penetrance of antibiotics into the pleural fluid
(> 1 0 mm thickness and < 1 12 the size of the hemi should be considered when an antibiotic is selected
thorax) and is free flowing. The Gram's stain and cul for the treatment of patients with parapneumonic
ture of the pleural fluid are negative and the pleural effusions. No reason exists to increase the dose of anti
fluid pH is more than 7.20. The risk of a poor out biotics merely because a pleural effusion is present.
come with a category 2 effusion is low. For patients hospitalized with community-acquired
Category 3 effusion meets at least one of the fol pneumonias that are not severe, the recommended
lowing criteria: (a) the effusion occupies more than agents are a fluoroquinolone alone, such as levofloxa
1 /2 the hemithorax, is loculated, or is associated with cin, moxifloxacin, gatifloxacin, or gemifloxacin, or a
a thickened parietal pleura; (b) the Gram's stain or /3-lactam (cefotaxime, ceftriaxone, ampicillin-sulbac
culture is positive; or (c) the pleural fluid pH is less tam, or ertapenem) ( 1 40). There is no reason to add
than 7.20 or the pleural fluid glucose is less than 60 a macrolide because atypical pathogens rarely cause a
mg/dL. The risk of a poor outcome with a category 3 pleural effusion ( 1 4 1 ) . For patients with severe com
effusion is moderate. munity-acquired pneumonia in whom pseudomonas
Category 4 effusion is characterized by pleural infection is not an issue, the recommended agents are a
fluid that is pus. The risk of a poor outcome with a /3-lactam plus either an advanced macrolide or a respi
category 4 effusion is high. ratory fluoroquinolone ( 1 40) . If a pseudomonas infec
tion is suspected, an antipseudomonas antibiotic such
as piperacillin, piperacillin-tazobactam, imipenem,
MANAG E M ENT meropenem, or cefepime should be included ( 1 40) .
Th e management o f parapneumonic effusions and Because anaerobic bacteria cause a sizable percentage of
empyemas involves rwo separate areas-selection of parapneumonic effusions, anaerobic coverage is recom
an appropriate antibiotic and management of the mended for all patients with parapneumonic effusions
pleural fluid. with either clindamycin or metronidazole ( 1 42) . In
patients with health care-associated pleural infection,
coverage should be provided for gram-negative enteric
Anti biotic Selection
bacteria and MRSA. A reasonable antibiotic selection
All patients with parapneumonic effusions or empy in such patients is a carbapenem such as meropenem
ema should be treated with antibiotics. If the Gram's and vancomycin ( 1 4 1 ) . There are no useful studies
stain of the pleural fluid is positive, it should guide on duration of therapy for bacterial infections of the
the selection of an antibiotic. The initial antibiotic pleural space. The current standard of practice is to
selection is usually based on whether the pneumonia continue antibiotics for several weeks ( 1 43, 1 44) .
is community-acquired or hospital-acquired and on
how sick the patient is. The initial antibiotic selection
lntrapleural Antibiotics
and dose are influenced to some extent by whether or
not a pleural effusion is present because some anti Intrapleural antibiotics were first used to treat an
biotics, for example, aminoglycosides, do not pen infected pneumonectomy space by Clagett and
etrate pleural fluid easily. The ease of penetrance of Geraci ( 1 45) in 1 963. Since that time, there have
various antibiotics into the pleural space was studied been several reports ( 1 46-1 50) regarding the use of
in our rabbit model of empyema ( 1 08). Antibiotic intrapleural antibiotics in the treatment of empyema
levels in samples of pleural fluid and serum were complicating pneumonia. All of these reports have
collected serially for up to 8 hours after penicillin, indicated positive results, but none were randomized
clindamycin, gentamicin, metronidazole, vancomy controlled studies. Until such controlled studies doc
cin, or ceftriaxone were administered intravenously. umenting the efficacy of intrapleural antibiotics are
The degree to which the different antibiotics pene completed, they are not recommended for patients
trated the infected pleural space was highly variable. with parapneumonic effusions.
226 PLE U RAL D I S EASES
M etronidazole Penicillin
1 00 - 1 00
0.1
��
0 60 1 20 1 80 240 300 360 420 480 0 60 1 20 1 80 240 300 360 420 480
Time (min utes) Time (minutes)
Clindamycin Ceftriaxone
1 00
1 000
--- Serum ----- Serum
:::::;- -o--- Pleural fluid :::::;- --0- Pleural fluid
-§, 10 -§,
_§, 1 00
_§, Q)
c: c:
·c:; � 0
>- x
1
E ro
ro
"t) 1 �Q)
.£ 1 0 -,
(.)
(3
0.1 1
0 60 1 20 1 80 240 300 360 420 480 0 60 1 20 1 80 240 300 360 420 480
Time (minutes) Time (minutes)
Vancomycin Gentamicin
1 00
Serum 1 00
----- Serum
:::::;- Pleural fluid Pleural fluid
.__
E
:::::;-
-<>--
Ci
10
_§, 10 -§,
Qi
r-
> � _§,
� c:
c: ·c:;
·c:; E
�·�
>- ro
0
u
E
c:
c
Q)
CJ
i ;�---<>-- 0------
--------..::::--.
ro
0.1 � ·C
> I
i
Options for Management of Pleural Fluid Storm et al. ( 1 50) reported that 48 of 5 1 patients
(94%) with empyema (e.g. , purulent pleural fluid
There are several treatment options available for the
or positive microbiologic studies on the pleural
management of the pleural fluid in patients with
fluid) were successfully treated with daily thoracen
parapneumonic effusion and these include observa
tesis. Simmers et al. ( 1 53) treated 29 patients with
tion, therapeutic thoracentesis, tube thoracostomy,
complicated parapneumonic effusions with daily
intrapleural instillation of fibrinolytics, VATS with
ultrasound-guided thoracenteses and reported that
the breakdown of adhesions and possible decortica
24 (83%) were successfully treated. The drawback
tion, thoracotomy with decortication and the break
to this latter study was that the patients underwent
down of adhesions, and open drainage.
an average of 7.7 ± 3.5 thoracenteses and the aver
Observation age hospitalization was for 3 1 days ( 1 53). Ferguson
et al. ( 1 54) reported that 1 9 of 46 patients (4 1 %)
In general, observation is not an acceptable option with empyema (e.g. , opaque fluid in the pleural space
because the pleural fluid from patients with para with the cloudiness due to neutrophils or organisms)
pneumonic effusions should be sampled as soon as were treated successfully with repeated thoracentesis.
it is identified. This sampling is important because Ozol et al. ( 1 55) recently reported that therapeutic
examination of the fluid is necessary to determine thoracentesis cured 42 of 44 patients (95 .4%) who
if drainage of the fluid is indicated (2) (Table 1 2.4) . had a mean pleural fluid pH of 6.8 and a mean pleu
Although only approximately 1 0 % of patients with ral fluid LDH of 1 ,8 1 8 IU/L. There have been no
parapneumonic effusions require drainage, it is controlled studies comparing therapeutic thoracen
important not to delay drainage in those who require tesis with small tube thoracostomy in the treatment
it because an effusion that is free flowing and easy of patients with complicated nonloculated parapneu
to drain can become loculated and difficult to drain monic effusions.
over a period of 1 2 to 24 hours (53, 1 5 1 ) . Observa
tion is the appropriate course if the patient has a Tube Th oracostomy
Class 1 parapneumonic effusion, that is, the effusion For the past several decades, the initial drainage
is less than 1 0 mm in thickness on the decubitus chest modality for most patients with complicated para
radiograph, ultrasonography, or chest CT scan. pneumonic effusions has been tube thoracostomy.
The chest tube should be positioned in a dependent
Therapeutic Thoracen tesis
part of the pleural effusion. Initially, the chest tube
Therapeutic thoracentesis was first proposed as a should be connected to an underwater-seal drainage
treatment modality for parapneumonic effusions in system. If the visceral pleura is covered with a fibrin
the middle of the 1 9th century ( 1 2, 1 3) . In 1 962, ous peel, the application of negative pressure to the
the American Thoracic Society (ATS) recommended chest tube may help expand the underlying lung and
repeated thoracentesis for nontuberculous empyemas hasten the obliteration of the empyema cavity. The
that were in the early exudative phase (34) . In 1 968, management of patients with chest tubes is discussed
Snider and Saleh (37) recommended that patients in Chapter 29.
with empyema could be managed with two therapeu What is the size of chest tubes that should be used
tic thoracenteses, but if fluid accumulated after that to treat complicated parapneumonic effusions? In the
time, then tube thoracostomy should be performed. past, relatively large (28 to 36 F) tubes have been rec
Recently, however, therapeutic thoracentesis as a ommended because of the belief that smaller tubes
treatment for parapneumonic effusions has received would become obstructed with the thick fluid. There
relatively little consideration. is, however, some data to suggest that such large tubes
As discussed in the chapter on experimental ani are unnecessary ( 1 44, 1 56) . The British Thoracic Soci
mals and pleural disease (see Chapter 4) , studies in ety (BTS) guidelines ( 1 44) state that a small bore
our rabbit model of empyema have shown that daily catheter 1 0 to 1 4 F will be adequate for most cases of
therapeutic thoracentesis starting 48 hours after complicated parapneumonic pleural infection. How
empyema induction is at least as effective as tube tho ever, there is no consensus on the optimal size of the
racostomy initiated at the same time ( 1 52) . In the chest tube for drainage ( 1 44) . The guidelines recom
last 20 years, there have been studies that suggest that mended regular flushing if a small bore flexible cath
some patients with complicated parapneumonic effu eter is used. The flushing technique recommended is
sions can be cured with therapeutic thoracentesis. the instillation of 20 to 30 ml saline every 6 hours
228 PLE U RAL D I S EASES
via a three-way stopcock ( 1 44) . It should be noted location ( 1 6 1 ) . Failure can also be due to loculi of the
that flushing larger bore drains is technically more pleural fluid that prevent complete pleural drainage,
difficult as these do not routinely have three-way taps or the failure may be due to fibrinous tissues coating
and disconnection for irrigation might encourage the the visceral pleura that prevent the underlying lung
introduction of secondary infection ( 1 44) . from expanding. If drainage is inadequate, a chest CT
A recent study ( 1 57) reviewed data on 405 scan should be obtained to delineate which of the fac
patients who participated in the Multi-center Intra tors mentioned in the preceding text is responsible.
pleural Streptokinase Trial (MIST I ) ( 1 42) that was If multiple loculi of pleural fluid are demonstrated,
designed to assess the efficacy and safety of strepto consideration should be given to performing VATS
kinase in patients with complicated parapneumonic with the lysis of adhesions.
effusions. As part of the study, they collected data on If the patient responds clinically and radiologically
the size and the complications of chest tube. They to closed-tube drainage of the pleural space, how long
reported that there was no significant difference in should the chest tubes be left in place? In general,
the frequency with which patients either died or chest tubes should be left in place until the volume of
required thoracic surgery in patients receiving chest the pleural drainage is less than 50 mL for 24 hours
tube of varying sizes (size < 1 0 F, 2 1 /58 [36%] ; size and until the draining fluid becomes clear yellow. The
1 0- 1 4 F, 75/208 [36%] ; size 1 5-20 F, 28/70 [40%] ; amount of sediment (representing WBCs and debris)
size >20 F, 30/69 [44%] [p = 0.27] ) ( 1 57). More in the collection system should be quantitated daily
over, patients who received the larger tubes reported and the chest tube should not be removed if more
significantly more pain ( 1 57) . It should be noted that than 5 mL sediment collects daily. If the chest tube
the patients were not randomized to the four differ ceases to function (no spontaneous fluctuation with
ent chest tube sizes but the size of the tube was deter respiratory efforts) , it should be removed because it
mined by the attending physician. The patients who serves no useful purpose and can be a conduit for
received the large bore chest tubes (> 20 F) tended pleural superinfection.
to have a significantly higher percentage of visibly At times, a patient responds clinically and radio
purulent fluid and a significantly higher pleural fluid logically to closed-tube drainage, but purulent
lactate dehydrogenase level ( 1 57) . drainage continues from the chest tube. In this situa
I n another study, 8 t o 1 2 F pigtail o r 1 0 to 1 4 tion, the decision to take a more aggressive approach,
F Malecot catheters were placed using the Seldinger for example, thoracoscopy or thoracotomy, can be
technique to treat 1 03 patients with empyema ( 1 58). aided by the injection of contrast material through
These small catheters served as the definitive treat the chest tube into the pleural space ( 1 62) . When
ment in 80 of the patients (78%) . These results are only a tube tract remains, the chest tube is gradu
certainly as good as those reported in recent surgi ally withdrawn over a few days, and the cavity is
cal series ( 1 59, 1 60) in which much larger tubes were allowed to fill in with granulation tissue. When a
used, but the parapneumonic effusions in the surgi larger cavity (greater than 50 mL) is demonstrated,
cal series may have been in a higher class. The advan empyemectomy with decortication or an open drain
tage of the smaller tube is that it is easier to insert age procedure should be performed.
and is less painful to the patient. The percutaneous How effective is tube thoracostomy alone in treat
catheters in the two studies referenced were placed ing patients with complicated parapneumonic effu
by interventional radiologists, and it is quite likely sions? The ACCP consensus statement concluded
that the excellent results are due to accurate catheter that therapeutic thoracentesis or tube thoracos
placement. tomy alone appears to be insufficient treatment for
Successful closed-tube drainage of complicated managing most patients with category 3 or 4 para
parapneumonic effusions is evidenced by improve pneumonic effusion ( 1 07) . Although I was on the
ment in the clinical and radiologic status within 24 committee that produced this consensus statement, I
hours. If the patient has not demonstrated signifi disagree with it. Different series have reported widely
cant improvement within 24 hours of initiating tube varying success rates with tube thoracostomy, but all
thoracostomy, either the pleural drainage is unsat are above 25%. Earlier in this section, a success rate
isfactory or the patient is receiving the wrong anti of more than 75% was reported with the use of small
biotic. In such patients, the culture results should chest tubes ( 1 58). Lim and Chin ( 1 63) reported a
be reviewed. Unsatisfactory pleural drainage is fre 28% success rate, whereas Athanassiadi et al. ( 1 64)
quently because the tube is positioned in the wrong reported that 75% of patients with complicated
C H APTE R 1 2 I PARAPN E U M O N I C E F F U S I O N S A N D E M PY E M A 229
parapneumonic effusions were managed successfully pleural drainage in patients with empyemas. Sub
with tube thoracostomy. In a recent study, 67% of70 sequently, the use of intrapleural streptokinase and
patients with complicated parapneumonic effusions streptodornase was largely abandoned because its
were managed successfully with tube thoracostomy intrapleural injection was associated with systemic
(28 to 32 F) ( 1 65). side effects, including febrile reactions, general mal
aise, and leukocytosis ( 1 67) . In the late 1 970s, Bergh
lntrapleural Fibrinolytics
et al. ( 1 68) reported the results with the intrapleural
Difficulties arise in the drainage of complicated injection of streptokinase alone in 1 2 patients with
parapneumonic effusions as a result of pleural fluid empyema. They reported radiologic improvement in
loculation. The pleural fluid loculations are produced 1 0 of their 1 2 patients.
by fibrin membranes that prevent the spread of the The rightful place for fibrinolytics in the manage
infected pleural fluid throughout the body, but which ment of loculated parapneumonic effusions remains
make drainage of the pleural space difficult. The the to be determined. In a landmark study ( 1 03) on the
ory behind the use of intrapleural fibrinolytics is that use of intrapleural fibrinolytics for the treatment of
they will destroy the fibrin membranes and facilitate complicated parapneumonic effusion, the adminis
drainage of the pleural fluid ( 1 66) . Many years ago, tration of streptokinase had no effect on the need for
Tillett et al. (28) reported that the intrapleural injec surgery or the duration of hospitalization (Fig. 1 2.3)
tion of streptokinase and streptodornase facilitated ( 1 03) . In this multicenter, randomized, controlled,
1 .00
Ol
t _g
8. :t:
e :s: 0.25
0..
0.00
0 3 6 9 12
Months
No. at Risk
P lacebo 222 1 61 1 50 1 45 1 20
A Streptoki nase 208 1 40 1 29 1 25 1 01
1 .00 "
\
Ol \
c \
·c: 0.75 \
'(ii 'O \
E � \
Q) = \
� al \
c :!:::: 0.50 FIGURE 1 2 .3 • A: Propor-
\
0 a. "
(/)
·-
t:: 0 'l
t i o n of patie nts s u rvivi n g
8. ..c 0.25 P l acebo
without req u i ri n g p l e u r a l
e d r a i n a g e surgery. B: Propor-
0..
Streptokinase t i o n re m a i n i n g hosp ita l i zed
0.00 d u ri n g fo l l ow-u p . (From
0 2 3 Maskell NA, Da vies CW, Nunn
AJ, et al. U K. con trolled trial
Months
of intrapleural streptokinase
No. a t Risk for pleural infection. N E n g l
P lacebo 222 28 6 1 J M e d . 2005; 352:865-874,
8 Streptoki nase 208 27 7 3
with permission.)
23 0 PLE U RAL D I S EASES
double-blind study, 454 patients were randomized to however, that the intrapleural injection of a fibrino
receive 25 0,000 IU of streptokinase or saline, twice lytic in rabbits with empyemas increases the drain
daily for 3 days, both with a total volume of 30 mL. age by a large amount, but does not improve the
To be eligible for the study, patients had to have pleu empyema score ( 1 79) . A meta-analysis including the
ral fluid that was macroscopically purulent, that was UK randomized study with streptokinase concluded
positive for bacteria on Gram's stain, or that had a that there was no support for the routine use of fibri
pH below 7.20. As can be seen from Figure 1 2.3, the nolytic therapy for patients who require chest tube
patients in the placebo group actually had a slightly drainage ( 1 8 0) .
higher survival rate without surgery than those in The original articles on enzymatic debridement
the streptokinase group (I 03) . One criticism of this for loculated parapneumonic effusions used Vari
study is that loculation of the pleural fluid was not dase, which consists of a fibrinolytic (streptoki
required and the fibrinolytics theoretically would be nase) and a DNase (streptodornase) . It is unclear
most useful if the pleural fluid were loculated. How how much the DNase contributed to the efficacy
ever, when the subgroups with loculation were ana of the preparation. We have shown that when thick
lyzed, there was no benefit from streptokinase in this empyemic material from rabbits is incubated with
group. either streptokinase or urokinase, there is no sig
In a second study, Diacon et al. ( 1 69), in a single nificant liquefaction of the fluid ( 1 67) . In contrast,
center, randomized, double-blind study, assigned when the fluid is incubated with Varidase, the
44 patients to receive daily pleural washes with fluid becomes completely liquefied over 4 hours.
streptokinase or saline for 4 or 5 days. Eligibil Although Varidase is currently not available in the
ity criteria for this study included frank pus in United States, recombinant human DNase (Pul
the pleural space or a complicated parapneu mozyme, Genentech, San Francisco, CA) is avail
monic effusion (pH <7.00 or pH <7.20 and able. Simpson et al. ( 1 8 1 ) have demonstrated that
evidence of pleural fluid loculation on chest x-ray recombinant DNase by itself is very effective at
or ultrasound) ( 1 69) . After 3 days, there was no reducing the viscosity of human empyema fluid.
significant differences in the groups, but after They also reported that the intrapleural admin
7 days, streptokinase-treated patients had a higher istration of 5 mg recombinant DNase for 3 days
clinical success rate (82% vs. 48%,p < 0 . 0 1 ) and fewer resulted in the resolution of an empyema in one
referrals for surgery (9% vs. 43%, p < 0.02) ( 1 69 ) . patient who refused surgery ( 1 82) . We demon
Th e patient characteristics were somewhat different strated that in our rabbit model of empyema the
in the two double-blind studies. The patients in the intrapleural injection of recombinant DNase in
Diacon study were younger (38 vs. 66 years old) combination with tissue plasminogen activator
and had lower values for pleural fluid pH (6.6 vs. (tPA) , but not by itself, resulted in a significant
6.8). In addition the patients in the Diacon study improvement in empyema score ( 1 79 ) . It has been
received antibiotics longer. hypothesized that streptokinase may be ineffective
There have been five other controlled studies because plasminogen levels in the pleural fluid are
in which a fibrinolytic was compared with a saline probably low and so may have been excessively
control ( 1 65 , 1 70-1 73). None of these studies was bound by the streptokinase into activator complex.
double blind. Four of the five studies concluded that Direct plasminogen activators, such as tPA, may
fibrinolytics were effective. In addition, there have overcome this problem as they do not require the
been at least five uncontrolled studies ( 1 7 4- 1 78), generation of an activator complex.
each with more than 20 patients, which concluded The recently reported MIST2 study ( 1 83) evalu
that fibrinolytics are useful in the management of ated the efficacy of tPA, DNase, tPA plus Dnase,
patients with loculated parapneumonic effusions. and placebo in the treatment of complicated parap
Success has been reported with both streptokinase neumonic effusions. In this double dummy, double
( 1 74-1 77) and urokinase ( 1 74, 1 77, 1 78). Each placebo randomized study 2 1 0 patients were ran
agent is administered intrapleurally in a total vol domized to one of the above four regimens. The
ume of 50 to 1 00 mL as long as it appears to be interventions were given twice daily for 3 days (tPA
facilitating pleural drainage. Some of the positive 1 0 mg, DNase 5 mg) . The primary outcome measure
results were based on the observation that there is was the absolute change in the pleural opacity on the
more pleural fluid drainage after the intrapleural chest radiograph from the day of randomization until
administration of a fibrinolytic. It should be noted, 7 days post randomization ( 1 83) .
C H APTE R 1 2 I PARAPN E U M O N I C E F F U S I O N S A N D E M PYE M A 23 1
The administration of the combination of tPA be completed with VATS in 66 (38%) ( 1 88) . If with
and DNase resulted in a significantly greater reduc VATS, the lung cannot be mobilized sufficiently to
tion in the pleural opacity ( - 29.5%) than did the reach the chest wall and diaphragm, the VATS inci
administration of tPA ( - 1 7.2%) , DNase ( - 1 4.7%) , sion can be enlarged so that the decortication can be
or placebo ( - 1 7.2%) ( 1 83) . There were no signifi completed with a full thoracotomy ( 1 86) .
cant differences between tPA, DNase, and placebo VATS i s very effective at treating incompletely
( 1 83) . When the secondary endpoints were exam drained parapneumonic effusions. Between the third
ined, the patients who received the combination had and fourth editions of this book, there were four arti
a significant decrease in surgical referrals and a signifi cles that reported the results ofVATS in this situation
cantly shorter hospital stay than did the patients who ( 1 89-1 92) . When these four studies with a total of
received placebo ( 1 83). tPA alone did not differ sig 232 patients are combined, VATS was the definitive
nificantly from placebo on either of these secondary procedure in 1 78 of the patients (77%) . The overall
endpoints while DNase alone was actually associated mortality rate was 3%, and the median time for chest
with more surgical referrals than was placebo ( 1 83) . tube drainage after the procedure ranged from 3.3 to
In view of the discussion in the preceding text, 7 . 1 days. The median hospital stay after VATS ranged
what is the rightful place of fibrinolytics in the from 5.3 to 1 2.3 days ( 1 89-1 92) . Since the fourth
therapy of loculated parapneumonic effusions? The edition of this book there has been one large study
results of the two multicenter studies ( 1 03 , 1 83) from with 234 patients reported from Taiwan ( 1 93) . In
the United Kingdom cast doubt on the effectiveness this study, 1 94 of the 234 patients (83%) had satisfac
of intrapleural fibrinolytics alone for the therapy of tory results with the first VATS procedure, although
complicated parapneumonic effusions. However, the 24 ( 1 0%) required open thoracotomy and 1 6 (7%)
second study ( 1 83) demonstrates that the combina required reoperation ( 1 83). In general, patients who
tion of tPA plus DNase is significantly better than require open thoracotomy had their complicated
placebo, while either agent by itself is not more effec parapneumonic effusion for a longer period before
tive than placebo. It is recommended that the combi thoracoscopy was attempted ( 1 93, 1 94) .
nation of tPA and DNase be used if one elects to use There was one small study that randomized 20
fibrinolytic therapy. Randomized studies are needed patients with either a loculated pleural effusion or a
to compare VATS with the intrapleural administra pleural Buid pH less than 7.20 to receive either chest
tion of tPA and DNase. It should be noted that a rare tube drainage plus streptokinase or VATS ( 1 95).
patient who receives a fibrinolytic intrapleurally will In this study, VATS was the definitive procedure in
develop a hemothorax ( 1 84, 1 85). 1 0 of 1 1 patients (9 1 %), whereas streptokinase was
definitive in 4 of 9 patients (44%) ( 1 95). The authors
Video -Assisted Thoracoscopy with Lysis o f
of this study concluded that in patients with locu
A dhesions and/or Decortication
lated parapneumonic effusions, a primary treatment
One option for the patient with an incompletely strategy of VATS is associated with a higher efficacy,
drained parapneumonic effusion is VATS. Although shorter hospital duration, and less cost than a treat
medical thoracoscopy is occasionally used in this situ ment strategy that uses catheter-directed fibrinolytic
ation, VATS is usually preferred because if the lung therapy ( 1 95).
cannot be expanded, the VATS can be converted An alternate use of VATS in the management of
to a full thoracotomy. A chest CT scan should be parapneumonic effusions is to use a VATS procedure
obtained before VATS to provide anatomic informa to insert the initial chest tube. At the time that the
tion about the size and extent of the empyema cavity chest tube is inserted, VATS is used to irrigate the
and whether the pleural surfaces are thickened ( 1 86) . pleural space and break down all the fibrous strands
However, it should be noted that some patients will ( 1 96) . One randomized study of 70 patients com
need a decortication even when no pleural thickening pared the results when the chest tube was inserted
was evident on the CT scan ( 1 87) . With VATS, the in the standard manner and when it was inserted in
loculi in the pleural space can be disrupted, the pleu conjunction with VATS ( 1 96). In this study, patients
ral space can be completely drained, and the chest with chest tubes inserted through VATS had a shorter
tube can be optimally placed ( 1 86) . In addition, if hospital stay (8.3 vs 1 2 . 8 days) and required less
the lung is trapped, an attempt can be made to per decortication ( 1 7% vs 3 7%) ( 1 96) . It appears to me
form a decortication. In one study of 1 72 patients that this approach to complicated parapneumonic
undergoing decortication, the decortication could effusions is advantageous.
232 PLE U RAL D I S EASES
Decortica tion and inserting one or more short, large-bore tubes into
the empyema cavity. Following this procedure, the
With decortication, all the fibrous tissue is removed
tubes are irrigated daily with a mild antiseptic solu
from the visceral and parietal pleura, and all pus is
tion. The drainage from the tubes can be collected in
evacuated from the pleural space ( 1 97) . Decortica
a colostomy bag placed over the tubes. The advan
tion eliminates the pleural sepsis and allows the
tage of this method over closed-tube drainage is that
underlying lung to expand.
drainage is more complete and the patient is freed
Decortication can be performed with VATS or with
from attachment to the chest tube bottles.
a full thoracotomy. In one study ( 1 7 1 ) , decortication
A similar but more complicated procedure is
was performed on 308 patients including 1 23 who
? pen- � �p drainage, in which a skin and muscle flap
underwent open thoracotomy and 1 85 who under
1s positioned so that it lines the tract between the
went VATS. The patients who underwent open tho
pleural space and the surface of the chest (203) after
racotomy had their surgery between 1 996 and 200 1
two or more overlying ribs are resected. The advan
while those undergoing VATS were treated between
tage of this open flap (Eloesser flap) is that it creates
2000 and 2006. Only 1 1 of the 1 85 patients (5.9%)
a skin-lined fistula that provides drainage without
who underwent VATS needed to be converted to an
tubes. Therefore, it can be more easily managed by
open th� racotomy ( 1 98). In this study, the group
the patient at home and permits gradual obliteration
who received VATS had significantly less pain, shorter
of the empyema space.
operative time, shorter hospital stay, and a more rapid
It is important not to convert to an open drainage
r� turn to work ( 1 98). The postoperative hospitaliza
procedure too early in the course of a complicated
tion was 3.9 days in open thoracotomy group and
parapneumonic effusion. With an open drainage pro
2.8 days in the VATS group ( 1 98). The results of
cedure, the pleural space is exposed to atmospheric
this and other studies suggest that when decortica
pressure. If the visceral and parietal pleura adjacent
tion is necessary, VATS is the procedure of choice
to the empyema cavity have not been fused by the
( 1 99) . VATS decortication can be performed with the
inflammatory process, exposure of the pleural space
patient awake using epidural anesthesia (200) .
to atmospheric pressure will result in a pneumotho
When managing patients with pleural infec
rax. Before open drainage procedures, this possibility
tions in the acute stages, decortication should only
can be evaluated by leaving the chest tube exposed
be considered for the control of pleural infection.
to atmospheric pressure for a short period and deter
Decortication should not be performed just to
mining radiologically whether the lung has collapsed.
remove thickened pleura because such thickening
If t? e lung does collapse in this situation, an open
usually resolves spontaneously over several months
dramage procedure can still be performed by creat
(20 1 ,202) . If after 6 months the pleura remains
ing an airtight seal and connecting the large tube to
thickened and the patient's pulmonary function is
a water-seal drainage apparatus (204) . The high mor
sufficiently reduced to limit activities, decortication
tality rate in patients with parapneumonic effusions
should be considered. However, this is an uncommon
during World War I has been attributed to perform
event. The prevalence of residual pleural thickening
ing open drainage procedures too early (26) .
greater than 1 0 mm 6 months after 348 patients
A patient treated with an open procedure can
were hospitalized with parapneumonic effusions was
expect to have an open chest wound for a prolonged
1 3.8% (202) . Residual pleural thickening was more
period. I � one series (53) of 33 patients treated using
common when patients had pus in their pleural space
open dramage procedures, the median time for the
or when there was delayed resolution of the parap
drainage site to heal was 1 42 days. With decortication,
neumonic effusion (202) . The results of pulmonary
the period of convalescence is much shorter (205),
function in patients with and without residual pleural
but decortication is a major surgical procedure that
thickening did not differ significantly (202) .
cannot be tolerated by markedly debilitated patients.
Open Drainage
Recommended Management of
Chronic drainage of the pleural space can be achieved
Parapneumonic Effusions
with open drainage procedures. Two different types
of procedures can be performed (203) . The simplest When a patient with pneumonia is initially evaluated,
procedure involves resecting segments of one to three one should ask if the patient has a parapneumonic
ribs overlying the lower part of the empyema cavity effusion. If the diaphragms are not visible throughout
C H APTE R 1 2 I PARAPN E U M O N I C E F F U S I O N S A N D E M PY E M A 23 3
the entire length on the lateral radiographs, decubitus thoracentesis, one need not worry about the parap
radiographs, ultrasonic examination, or chest CT scan neumonic effusion. If the pleural fluid reaccumulates
should be obtained to determine whether free pleu and there were no bad prognostic factors at the time
ral fluid is present. If free pleural fluid is present and of the initial thoracentesis, no additional therapy is
the distance between the inside of the chest wall and indicated as long as the patient is doing well. If the
the outside of the lung is more than 1 0 to 20 mm, the fluid reaccumulates and there were bad prognostic
pleural fluid needs to be sampled. If there is doubt as factors present at the time of the initial thoracente
to how much of the density in a hemithorax is paren sis, a second therapeutic thoracentesis should be per
chymal and how much is pleural, a CT scan of the formed. If the fluid reaccumulates a second time, a
chest should also be obtained. If more than minimal tube thoracostomy should be performed if any of the
fluid is demonstrated on the CT scan, the pleural fluid bad prognostic factors were present at the time of the
should be sampled. The reason for sampling the pleu second therapeutic thoracentesis.
ral fluid in these situations is to determine whether Performance of the therapeutic thoracentesis also
any bad prognostic factors are present (Table 1 2.4) . delineates whether the pleural fluid is loculated. If
The options for the invasive treatment of com the pleural fluid is loculated, and if any of the other
plicated parapneumonic effusions are listed in bad prognostic factors listed in Table 1 2.4 are present,
Table 1 2.7. In general, one moves from the less inva then more aggressive therapy should be initiated. The
sive treatments to the more invasive treatments. It is patient should be subjected to VATS if it is available
important to abandon a treatment within 1 or 2 days and the patient is an operative candidate. Otherwise,
if it is ineffective. Not every treatment needs to be a chest tube should be inserted and 1 0 mg tPA plus
used. If a patient is going to need a decortication, 5 mg DNase should be administered intrapleurally
it should be performed within 1 0 to 14 days of the twice a day for 3 days (MIST2) . If the lung does not
initial identification of the parapneumonic effusion. expand, then decortication should be performed. If
Even in the 2 1 st century, there are frequent delays in the patient was not subjected to VATS originally and
performing the definitive procedures in patients with the pleural fluid remains incompletely drained after
complicated parapneumonic effusions. For example, 1 0 or more days, consideration should be given to
Chu et al. (206) reported that the mean time from performing a VATS if the patient is an operative can
admission until a surgical consult was requested for didate or an open drainage procedure if the patient is
26 patients with empyema in Regina, Saskatchewan, not an operative candidate.
Canada, was 44 days. The ACCP ( 1 07) and the BTS ( 1 44) have both
An outline of the recommended treatment for a published guidelines for the management of compli
patient with a parapneumonic effusion is presented cated parapneumonic effusions. The guidance given
in Figure 1 2.4. If a patient has sufficient pleural fluid by the ACCP guidelines ( 1 07) is rather vague in that
to warrant a thoracentesis, it is recommended that they say most category 3 and category 4 effusions
a therapeutic rather than a diagnostic thoracentesis (Table 1 2.6) are not cured with tube thoracostomy
should be performed initially. Although there are no or therapeutic thoracentesis. They recommend the
controlled studies validating this approach, the rea intrapleural instillation of fibrinolytics, VATS, or
soning behind this recommendation is as follows. If open thoracotomy rather than a stepwise progression
no fluid reaccumulates after the initial therapeutic as I recommend. The BTS recommends a diagnostic
thoracentesis rather than a therapeutic thoracentesis
and then the insertion of a chest tube if any bad prog
nostic factors are present. If the fluid is not drained
adequately after 5 to 7 days, they recommend consul
tation with a thoracic surgeon ( 1 43) .
Tu be thoracosto my
Tu be thoracosto my with the i ntra p l e u ra l a d m i n is S P E C I A L S IT U AT I O N S W I T H E M PY E M A
tratio n of fi b r i n o lytics
Empyema in Children
Thoracoscopy with the b reakdow n of ad hesions
Thoracotomy with decortication
Parapneumonic effusions are the most common cause
of pleural effusion in children (207) . As mentioned
Listed i n order of i n crea s i n g i nvasiveness. earlier in this chapter, the bacteriology of empyema
234 PLE U RAL D I S EASES
Parapneumonic effusion
Therepeutic thoracentesis
Continue
antibiotics
Decortication
antibiotics
Open d rainage
procedure
Tube thoracostomy
in children varies somewhat from that in adults. In However, the application of real-time polymerase
children the incidence of anaerobic infection is lower, chain reaction (PCR) for bacterial DNA can improve
whereas that of H. influenzae is higher. In Texas, the the positive results to more than 80% (2 1 2) .
most common bacterial cause of empyema in chil Since the introduction o f the 7 -valent PCV7 in
dren now is MRSA (208). However, other recent the United States, the pneumococcal pneumonia hos
studies have reported that S. pneumoniae serotype 1 pitalization rates decreased 61 %, but the pneumonia
is the most common pathogen isolated (209,2 1 0) . hospitalizations complicated by empyema increased
Th e bedside diagnosis o f S. pneumoniae infection can 2.0 1 -fold (2 1 3) . Rates of pneumococcal and strep
be made with the Binax NOW S. pneumoniae kit tococcal empyema remained stable, whereas the rates
which had a sensitivity of 83.3% and a specificity of for staphylococcal and other empyema increased by
93.5% in one study of 1 30 patients with pediatric 4.08- and 1 .89-fold, respectively (2 1 3) . A second
empyema (2 1 1 ) . It is noteworthy that the vaccines study (2 1 4) reported similar findings. In a study from
available in the United States and the United King Australia, nonvaccine serotypes accounted for 5 1 of
dom are not directed against this serotype (209,2 1 0) . 53 pneumococcal isolates (2 1 5) . It is possible that
Th e majority o f children with complicated parapneu the vaccine results in the selection of pneumococci
monic effusions have negative pleural fluid cultures. which are more likely to cause empyema. There have
CHAPTER 12 I PARAP N E U M O N I C E F F U S I O N S A N D E M PY E M A 23 5
also been increases i n the incidence o f empyema in significant scoliosis, evidence of parenchymal entrap
Taiwan, a country without a pneumococcal vaccina ment, and anaerobic infection. Tube thoracostomy,
tion program (2 1 6) . The reason for the increased inci VATS , or decortication should be instituted if the
dence of empyema is unknown (2 1 7) . patient has rwo or more of these indicators of a poor
Another difference berween children and adults prognosis.
is that the complicated parapneumonic effusions In general, the management of children with
can lead to scoliosis in the children. Mukherjee et al. empyema is very similar to that for adults. The chil
(2 1 8) reported that scoliosis of 10 degrees or more dren are usually healthy, so there is very little, if any,
was noted in 7 1 of 1 22 pediatric patients (7 1 %) . role for open drainage procedures. If the fluid is not
Patients with scoliosis did not require more decortica drained adequately with tube thoracostomy the rwo
tions. The scoliosis resolved in all patients (2 1 8) . The main options are intrapleural fibrinolytics and VATS .
percentage of children with pneumonia who require Gates et al. (223) reviewed the literature in 2004 and
surgery for pleural complications has ranged from found that patients treated with VATS or thoracot
3% in a study in Brazil (2 1 9) to 28% in a study from omy had significantly shorter hospitalizations (9 .9
Salt Lake Ciry, Utah, USA (209) . and 10.5 days, respectively) than did patients treated
The treatment for children with complicated with chest tube only or chest tube plus fibrinolyt
parapneumonic effusions is similar to that for ics ( 1 6.4 and 1 8 .9 days, respectively) . Early VATS
adults with parapneumonic effusions. One differ (within 48 hours of admission) leads to significantly
ence berween the management of parapneumonic shorter hospitalizations that if the VATS is performed
effusions in children and adults is that a diagnostic later (225) .
thoracentesis is frequently not performed in children However, some series show that patients treated
because this requires sedation (2 1 9) . Another big dif with fibrinolytics usually do not require VATS or
ference berween children and adults is that children thoracotomy. Wells and Havens (226) treated 7 1
are almost always in good general health and the final patients with 25,000 to 1 00,000 I U urokinase or
results are usually excellent (220) . 0 . 1 mg/kg alteplase and reported that the treatment
In 2005, the BTS has published guidelines on the was successful in 70 patients (99%) . Randomized
management of pleural infection in children (22 1 ) . controlled studies evaluating the effectiveness of
Highlights o f the guidelines include the following rec intrapleural fibrinolytics have produced conflicting
ommendations: (a) ultrasound must be used to confirm results. In one study, 40 patients with Light's class
the presence of a pleural fluid collection; (b) ultrasound 5, 6, or 7 parapneumonic effusions were random
should be used to guide thoracentesis; (c) biochemical ized to receive saline or streptokinase 1 5,000 IU/kg/
analysis of pleural fluid is unnecessary in the man dose intrapleurally for 3 consecutive days. There was
agement of parapneumonic effusion; (d) effusions no difference in the duration of fever or the median
that are enlarging and/or compromising respiratory duration of drainage (227) . In a second study that
function should not be managed by antibiotics alone; was double blind and multicenter, 60 children with
(e) if a child has significant pleural infection, a chest a persistent fever of greater than 38°C after more
tube should be inserted; (f) small rather than large than 24 hours of parenteral antibiotic therapy or the
chest tubes should be used to minimize discomfort; presence of a pleural collection causing respiratory
(g) when the effusion consists of thick pus or is locu distress were randomized to receive intrapleural uro
lated, intrapleural fibrinolytics should be given; and kinase 40,000 IU in 40 mL of saline or saline only
(h) patients should be considered for surgical treat 1 2 hourly for 3 days (228) . In this study, the mean
ment if they have persisting sepsis in association with hospital stay was significantly shorter with urokinase
a persistent pleural fluid collection, despite a chest (7.4 days) than it was with saline (9 . 5 days) but the
tube and antibiotics (22 1 ) . need for surgery in the urokinase group ( n 2) and
=
Although the BTS guidelines state that the pleu the saline group ( n = 3) did not differ significantly
ral fluid pH and glucose levels should not be used (228) . Since the MIST l ( 1 03) and MIST2 ( 1 83)
in selecting which pediatric patients with parapneu double-blind randomized placebo controlled studies
monic effusions should undergo tube thoracostomy, in adults have failed to demonstrate any benefit from
a low pleural fluid pH or a low pleural fluid glucose a thrombolytic itself, one should be cautious in using
appears to be important in indicating which children fibrinolytics in children. To my knowledge, the util
will have a poor outcome (209,222-224) . Addi ity of the combination of a fibrinolytic and DNase
tional indicators of a poor outcome in children are has not been assessed in children.
236 PLE U RAL D I S EASES
Several authors have recommended that VATS be the tree and an overwhelming pneumonia can result. It
primary means of therapy for pediatric patients (229) . should be emphasized that the presence of a bron
In one series of 4 1 patients, all were cured with VATS chopleural fistula in conj unction with infected pleu
and the median length of stay was only 7 days (230) . ral fluid is a medical emergency. Drainage should be
In a second series, 49 children had VATS performed instituted immediately to prevent the possibility of
within 48 hours of admission and the median total contaminating the entire respiratory system by the
length ofstay was 9 . 9 days (208) . In a third study (23 1 ) , infected pleural fluid (Fig. 1 2 . 5 ) .
1 1 4 children had VATS for empyema. Th e median Th e presence o f a bronchopleural fistula should
hospital stay postoperatively was 7 days, and the be suspected when a patient with a pleural fluid col
procedure was successful in 96 patients (93 % ) . Five lection raises more sputum than would be expected
patients needed to be converted to thoracotomy and from the associated pulmonary disease. If the patient
three patients developed a recurrent empyema (23 1 ) . raises large amounts of sputum only when lying in
I n the one randomized study comparing fibrinolytics one position, a bronchopleural fistula is strongly
and VATS, 60 patients were randomized to receive suggested. Radiologically, a bronchopleural fistula is
tube thoracostomy with urokinase or VATS if they had manifested by the presence of an air-fluid level in the
a pleural effusion and remained febrile for more than pleural space when the radiograph is obtained with
24 hours after parenteral antibiotics were initiated or if the patient in the upright position (Fig. 1 2 . 5 ) . It is
they had respiratory distress (232) . In this study, there sometimes difficult to determine whether the air
was no difference in the median hospital stay (VATS 8 fluid levels are in the lung parenchyma or in the pleu
days, urokinase 7 days) , the need for additional therapy ral space. The utility of ultrasound and CT studies in
(VATS 4, urokinase 5), or radiologic outcome after making this differentiation is discussed in Chapter 6.
6 months (232) . In a second randomized study (233)
with 36 patients comparing VATS with 4 mg tPA, there
E m pyema D i sta l t o a n O bstru cted B ro n c h u s
was no difference in days of hospitalization after inter
vention, days of oxygen requirement but VATS was O n e contraindication t o the placement o f chest tubes
associated with higher charges. If VATS is performed in patients with complicated parapneumonic effu
and the sepsis is still not controlled, then the patient sions is the presence of a malignant tumor obstruct
should be subjected to decortication (222,230) , but ing a lobar or main stem bronchus. If chest tubes are
this is an unusual event in children (208,234) . It placed in such patients, the bronchial obstruction
should be emphasized that decortication is indicated will prevent expansion of the lung underlying the
only for pleural sepsis. If the patient has extensive pleural effusion and the unfortunate patient will be
pleural thickening without pleural fluid, the thicken saddled with a chest tube or an open chest wound for
ing can be expected to resolve completely over the next the remainder of his life. When a patient is discovered
6 months without surgery (235 ) . to have a complicated parapneumonic effusion distal
Two recent reviews (236,237) of randomized con to an obstructed bronchus, appropriate antibiotics
trolled studies comparing VATS with chest drain and should be administered in conjunction with therapy
fibrinolytics concluded that the best available evi for the obstructed bronchus, which could include
dence does not support the contention that VATS is radiotherapy, an endobronchial stent, or laser therapy.
superior to chest drain with fibrinolytics. Clinicians Tube thoracostomy can be instituted if the obstruc
should consider local expertise in the interventions, tion is relieved with the therapy. If the obstruction
potential adverse events, and patient preference when persists, the patient can be sent home with a prescrip
deciding between these two treatment options. tion of appropriate oral antibiotics. In my experi
ence, continuous administration of oral antibiotics to
patients with pleural sepsis and bronchial obstruction
E m pye m a Associated w i t h
allows them to live in symbiosis with their pleural
B ronchop l e u ra l Fist u l a
infection without excessive systemic toxicity.
When a n empyema i s complicated b y the presence o f
a bronchopleural fistula, adequate pleural drainage is
Postp n e u m o n ecto my E m pyema
crucial (Fig. 1 2 . 5 ) . Pleural fluid that is not drained
exteriorly with chest tubes is likely to drain interi Empyemas following thoracic surgical proce
orly into the lung through the fistula. The bacteria dures account for approximately 25% of all empy
are then spread throughout the bronchopulmonary emas (32,37,45 ) , and the procedure is usually a
CHAPTER 12 I PARAP N E U M O N I C E F F U S I O N S A N D E M PY E M A 237
c D
FIGURE 1 2 .5 • A: Poste roanterior (PA) chest rad i o g r a p h of a 44-ye a r-old man who h a d p l e u ritic ch est
pain, feve r, and l e u kocytosis for 1 week. B: Co m p uted tomography (CT) scan dem onstrating that most of the
opacity in the rig ht l ung fie ld in the PA chest radiograph is p l e u ra l fluid. C: Ante roposterio r (AP) ch est rad i o
graph 2 days l ater dem onstrating extensive infiltrates through out both l ungs in addition to p l e u r a l effu sion.
N ote l ucent a rea (arrow) in rig ht m i d l ung fie l d ind icating an a rea in the p l e u ra l space that had d ra ine d . The
infiltrates a re a resu lt of t h e p l e u r a l p u s ente ring the tracheobronchia l tree. D: CT scan 1 day after the AP
ch est rad i ogra p h dem onstrating extensive b i latera l p u l m ona ry infiltrate. N ote that there is no chest t u b e in
p l ace so that the pus in the p l e u ra l space continues to d rain t h ro u g h the tracheob ronch i a l tree.
pneumonectomy. After a pneumonectomy, there is becomes 80% to 90% filled with fluid within 2 weeks
a characteristic evolution of radiologic findings, and and completely filled within 2 to 4 months (238) . Dur
deviations from this pattern suggest the possibiliry of ing this period, the mediastinum progressively shifts
postpneumonectomy empyema. Immediately after ipsilaterally (Fig. 1 2.6B) . Failure of the mediastinum to
pneumonectomy, the ipsilateral pleural space contains shift in the postoperative period indicates an abnormal
air, the mediastinum is shifted to the ipsilateral side, iry in the postpneumonectomy space (238). Similarly,
and the hemidiaphragm is elevated (Fig. 1 2.6A) . The the most sensitive indicator of late complications in
postpneumonectomy space then begins to fill with the pneumonectomy space is the return to the midline
serosanguineous fluid at a rate of approximately two of a previously shifted mediastinum or a shift of the
rib spaces a day. In most patients, the pleural space mediastinum to the contralateral side (Fig. 1 2.7) (238).
23 8 PLE U RAL D I S EASES
A 8
FIGURE 1 2 .6 • Appea ra nce of chest ra d i o g r a p h afte r p n e u m o n ecto my. A: Posteroa nte r i o r chest rad io
g ra p h from a patient 1 week after p n e u m o n ecto my. N ote that the postp n e u m o n ectomy space conta i n s an
a i r-fl u i d l eve l and that the med iasti n u m i s sh ifted towa rd the side of the p n e u m o nectomy. B: Poste roa n
terior chest rad i o g r a p h from the s a m e patient 1 yea r after p n e u m o nectomy. The m e d i a sti n u m h a s sh ifted
m o re toward the s i d e with the p n e u m o n ecto my, a n d the h e m ithorax is com p l etely opacifi e d .
The postoperative occurrence of empyema is a and, consequently, it is difficult to sterilize the space.
dreaded complication of pneumonectomy. The com Approximately 80% of patients with postpneu
plication is particularly serious because it is impos monectomy empyema have a bronchopleural or an
sible to eliminate the space containing the infection, esophagopleural fistula as a complication (239,240) .
A 8
Therefore, all patients with this type o f empyema the anterior axillary line is usually appropriate (240) .
should have a barium swallow and a bronchoscopic In the first few weeks following pneumonectomy, the
examination (240) . mediastinum is not stable. Therefore, suction should
The incidence of empyema following pneumonec not be applied to the chest tube and an open drainage
tomy is approximately 5% (239 ,240) , and the mor procedure should not be performed.
tality rate from postpneumonectomy empyema is After the initial drainage, the treatment is largely
approximately 1 2% (239) . Berween 1 9 8 5 and 1 998, dependent on whether the patient has a broncho
713 patients underwent pneumonectomy at the pleural fistula. If the patient does not have a bron
Mayo Clinic and empyema developed in 53 of the chopleural fistula, then the treatment of choice is
patients (7. 5%), and 32 of these had a bronchopleu closed-tube drainage of the pleural space in con
ral fistula (24 1 ) . The administration of antibiotics j unction with antibiotic irrigation ( 1 47,239,246) .
prophylactically starting before surgery and continu A chest tube is inserted into the most dependent
ing until the pleural drains are removed resulted in a portion of the patient's empyema cavity and is
significant reduction in postpneumonectomy empy connected to an underwater-seal drainage appara
ema in one study (242) . An infected pneumonectomy tus. The antibiotics to which the offending organ
space usually becomes manifest in one of four ways: isms are susceptible can be instilled by this tube,
(a) a febrile illness with signs of systemic toxicity; if a double-lumen tube is used ( 1 47) , or through
(b) expectoration of large amounts of pleural fluid; a separate, smaller tube inserted into the second or
(c) an air-fluid level in the pneumonectomy space; or third intercostal space at the midclavicular line. The
(d) the drainage of purulent material from the surgi antibiotics can be infused through the pleural space
cal incision. The time from pneumonectomy to the continuously. Alternatively, with the drainage tube
development of an empyema ranges from 2 days to clamped, several hundred milliliters of antibiotic
7 years, with most infections evident within 4 weeks solution can be instilled into the pleural space and
(243,244) . The diagnosis should be suspected in allowed to remain for several hours. The pleural fluid
any patient who, following this operation, becomes is then drained, and the sequence is repeated. When
febrile, starts expectorating large amounts of pleural the drained fluid becomes clear, the irrigation fluid
fluid, has purulent drainage from his thoracotomy is changed to normal saline solution for 24 hours. If
wound, or has a mediastinum that is at midline or the culture of this drainage is sterile, 1 00 mL of con
is shifted to the contralateral side (Fig. 1 2.7) . Other centrated irrigation fluid is left in the pleural space,
causes of a contralateral mediastinal shift post pneu and the chest tubes are removed. Gossot et al. (248)
monectomy are hemothorax and chylothorax (245) . recommend thoracoscopic debridement shortly after
Diagnosis is established by a thoracentesis, demon the initial chest tube is placed. The debridement
strating bacteria on the Gram's stain of the pleural facilitates sterilization of the postpneumonectomy
fluid. If more than several weeks have passed since space.
the patient's pneumonectomy, ultrasound exami Antibiotic irrigation of the pneumonectomy space
nation of the pleural space should be performed to is also effective in patients in whom the broncho
identify the appropriate location for the thoracente pleural fistula is repaired surgically. Gharagozloo et
sis. S. aureus is the bacterium responsible for most al. (239) reported that this approach was successful
postpneumonectomy empyemas (243,246,247) , but in all 22 patients who had their bronchopleural fis
gram-negative organisms such as E. coli, Pseudomo tula repaired primarily. At the time that the bron
nas sp, and Proteus sp, as well as fungi are at times chopleural fistula was repaired, the pleural space was
responsible. meticulously debrided. In 20 patients, the Gram's
All patients with postpneumonectomy empyema stain pleural fluid was negative on day 8 of the irriga
should be treated with a chest tube and appropriate tion, and in the remaining 2 patients, Gram's stain
antibiotics. Placement of a chest tube more than 1 of the pleural fluid was negative on day 1 6 after the
week following the pneumonectomy may be dif operation.
ficult, because the evacuated hemithorax rapidly An accelerated treatment program for post
loses volume by apposition of ribs and shift of the pneumonectomy empyema has been described by
mediastinum. Therefore, a chest CT scan is useful to Schneiter et al. (249) . With their procedure, they
assess the size and location of the residual space and perform repeated surgical debridement of the chest
to determine the optimal line of approach. Placement cavity under general anesthesia every 48 hours with
of the tube in the fourth or fifth intercostal space in temporary closure of the chest which is filled with
240 PLE U RAL D I S EASES
povidone-iodine-soaked towels and an antibiotic the first few weeks of surgery, attempts can be made
solution based on the infecting organism (249) . In a to close the bronchopleural fistula directly. Recently,
series of 75 patients, they reported that 97.3% were Gharagozloo et al. (239) reported the successful pri
successfully, the mean hospitalization was 1 8 days mary repair of all 22 bronchopleural fistulas. None
and the mortality rate was only 4% within the first of the patients had a bronchial stump opening that
90 days (249) . was more than 25% of the diameter of the bronchus.
An alternate approach to postpneumonectomy Attempts can be made to close the bronchopleural
empyema involves the creation of a large opening in fistula with fibrin glue endoscopically. In one study,
the chest by resecting several inches of the rib infe 8 of 36 patients had their bronchopleural fistula suc
rior to the thoracotomy incision and one or more ribs cessfully treated using fibrin sealant and decalcified
superior to it. The procedure is called the Clagett pro spongy calf bone (25 1 ) . Pairolero et al. (253) advo
cedure after the surgeon who first described it ( 1 45 ) . cate the intrathoracic transposition of extrathoracic
Th e superficial fascia i s sutured down t o the peri skeletal muscle to facilitate closure of the fistula.
osteum of the resected ribs to leave a large window They attempted this procedure in 28 patients follow
in the thoracic wall. Each day, the empyema cavity ing open drainage and reported success in 24. The
is irrigated with a mildly antiseptic solution such median number of operations was 5, with a range of
as half-strength Dakin's solution or chlorhexidine 1 to 1 9 . The median hospitalization time was 34 days,
(Hibitane) (247) . These irrigations are continued for with a range of 4 to 1 37. A related method closes the
several weeks until the drainage is no longer purulent fistula with an omentopexy (254,2 5 5 ) .
and the empyema cavity appears to be well debrided.
At this time, the opening in the chest wall is closed,
Posttra u m atic E m pye m a
and a 0.25% solution of neomycin is placed in the
cavity. It has been recommended by some (250), the Empyema remains a distressing complication after
that Clagett procedure be followed by a thoracoplasty thoracic injury. In one study published in 1 997,
but I see no reason for this as a thoracoplasty is a dis the incidence of empyema requiring decortication
figuring procedure. was 4% in 584 patients who were treated with tube
Goldstraw (247) treated 29 patients with post thoracostomy (256), whereas in another series, the
pneumonectomy empyemas in the foregoing man incidence was 1 . 8% of 5 ,474 patients (25 7) . Factors
ner and attempted closure in 22 of the patients . In that predicted the development of an empyema were
17 of these 22 patients (77% ) , closure was success retained hemothorax (odds ratio, 1 2. 5 ) , pulmonary
ful in that no evidence was seen of a recurrence of contusion (odds ratio, 6.3), and multiple chest tube
the empyema from 5 weeks to 9 years after closure. placement (odds ratio, 2 . 5 ) . Factors that did not pre
The five patients in whom closure failed initially dict empyema were severity of injury, mechanism
were subj ected to a second fenestration, and a suc of injury, the setting in which tube thoracostomy
cessful closure was eventually obtained in two of was performed, number of days chest tubes were
these patients . Other workers have reported much in place, and antibiotics at the time of tube thora
poorer results with the Clagett procedure. Shamj i costomy (256) . Aguilar et al. (256) recommend the
et al. (25 1 ) achieved successful closure in only 2 of early drainage of the pleural space with the VATS
3 1 patients (6%) managed in this manner, whereas technique in posttrauma patients with fluid collec
Bayes et al. (252) reported success in 1 0 of 28 tions in the pleural space. In general, the manage
patients (36%) . This treatment is time consuming, ment of posttraumatic empyema is the same as that
with a mean interval between fenestration and clo of parapneumonic empyema.
sure of 40 days (range 2 1 -74 days), and patients
usually have to remain hospitalized for the entire
period (202) . The closed irrigation method is the
REFERENCES
procedure of choice in my opinion.
If a bronchopleural fistula is present, several differ I. Halm EA, Teirstein AS . Management of community acquired
ent procedures can be used in an attempt to close the pneumonia. N Engl] Med. 2002;347:2039-2045.
fistula. On occasion, the bronchopleural fistula will 2. Light RW, Girard WM, Jenkinson SG, et al. Parapneumonic
effusions. Am J Med. 1 980;69:507-5 1 1 .
close with continuous irrigation of the infected pleu 3. Menendez R, Torres A, Zalacain R, et al. Risk factors of treat
ral space, but usually more extensive procedures must ment failure in community acquired pneumonia: implications
be done. If the bronchopleural fistula develops within for disease outcome. Thorax. 2004;59:960-965.
CHAPTER 12 I PARAP N E U M O N I C E F F U S I O N S A N D E M PY E M A 24 1
4. Hasley PB, Albaum MN, Li Y-H, et al. Do pulmonary radio 3 1 . Ferguson MK. Thoracoscopy for empyema, bronchopleural
graphic findings at presentation predict mortality in patients fistula, and chylothorax. Ann Thorac Surg. 1 993;56:644-645 .
with community-acquired pneumonia? Arch Intern Med. 3 2 . Weese WC, Shindler ER, Smith I M , e t al. Empyema o f the
1 996; 1 56:2206-22 1 2 . thorax then and now. Arch Intern Med. 1 973; 1 3 1 : 5 1 6-520.
5 . Fine MJ, Auble TE, Yealy O M , e t al. A prediction rule c o iden 33. Vianna NJ. Nontuberculous bacterial empyema in
tify low-risk patients with community-acquired pneumonia. patients with and without underlying diseases. JAMA.
JAMA. 1 996;275 : 1 34-1 4 1 . 1 97 1 ;2 1 5 :69-75 .
6 . Espana P P, Capelastegui A, Quintana JM, e t al. A prediction 34. Andrews NC, Parker EF, Shaw RR, et al. Management of non
rule co identify allocation of inpatient care in community cuberculous empyema. Am Rev Respir Dis. 1 962;8 5 : 93 5-936.
acquired pneumonia. Eur Respir]. 2003;2 1 :695-70 1 . 35. Wiener-Kronish JP, Sakuma T, Kudoh I, et al. Alveolar epithe
7 . Light RW. Parapneumonic effusions and empyema. Proc Am lial injury and pleural empyema in acute P aeruginosa pneumo
Thorac Soc. 2006;3:75-80. nia in anesthetized rabbits . ] Appl PhysioL 1 993;75 : 1 66 1 - 1 669.
8 . Tobin CL, Lee YC (Gary) . Pleural infection-What we need 36. Light RW. Management of parapneumonic effusions. Arch
co know but don't. Curr Opin Pu/ Dis. 20 1 2; 1 8 :32 1-32 5 . Intern Med. 1 9 8 1 ; 1 4 1 : 1 339- 1 34 1 .
9 . Bender J, Ampofo K , Sheng X , e t al. Parapneumonic empy 37. Snider GL, Saleh S S . Empyema o f the thorax i n adults: review
ema deaths during past century, Utah. Emerg Infect Dis. of 1 05 cases. Chest. 1 968;54: 1 2- 1 7.
2009; 1 5 :44-48. 3 8 . Yeh TJ, Hall DP, Ellison RG . Empyema thoracis: a review of
1 0 . Grijalva CG, Zhu Y, Nuorti JP, et al. Emergence of parapneu 1 1 0 cases. Am Rev Respir Dis. 1 963;88 :78 5-790.
monic empyema in the USA. Thorax. 20 1 ! ;66:663-668. 3 9 . Smith JA, Mullerworth MH, Westlake GW, et al. Empyema
1 1 . Broaddus VC. Infections in the pleural space. An update on thoracis: 1 4-year experience in a teaching center. Ann Thorac
pathogenesis and management. Semin Respir Crit Care Med. Surg. 1 9 9 1 ; 5 1 :39-42.
1 99 5 ; 1 6:303-3 1 4 . 40. Jones FL, Blodgett RC. Empyema in rheumatoid pleuropul
1 2 . Adams F. The Genuine Works of Hippocrates. New York, NY: monary disease. Ann Intern Med. 1 97 1 ;74:665-67 1 .
William Wood; 1 948:266. 4 1 . Finland M , Barnes MW. Changing ecology of acute bacterial
13. Bowditch HI. Paracentesis thoracic: an analysis of 25 cases of empyema: occurrence and mortality at Boston City Hospi
pleuritic effusion. Am Med Monthly. 18 53;3-4 5. tal during 12 selected years from 1 93 5 co 1 972. J Infect Dis.
1 4 . Trousseau A. Lectures on clinical medicine delivered at the Hotel 1 978; 1 37:27 4-29 1 .
Dieu Paris, Vol 3. JR McCormick (trans) . London, England: 42. Bartlett JG, Gorbach SL, Thadepalli H, et al. Bacteriology of
New Sydenham Soc; 1 987;3 : 1 98 . empyema. Lancet. 1 974; 1 :33 8-340 .
1 5 . Hewitt C. Drainage for empyema. B r Med]. 1 876; 1 : 3 1 7. 43. Brook I , Frazier EH. Aerobic and anaerobic microbiology o f
16. Estlander JA. Sur le resection des coot/: clans l'empyi:me chro empyema. A retrospective review i n two military hospitals.
nique. Rev Mens Med Chir. 1 897; 8 : 8 8 5 . Chest. 1 993; 1 03 : 1 502- 1 507.
1 7 . Schede M . Die Behandlung der empyeme. Verh Innere Med 44. Alfageme I, Munoz F, Pena N, et al. Empyema of the thorax
Weisbaden. 1 8 90;9:4 l. in adults. Etiology, microbiologic findings, and management.
18. Fowler GR. A case of thoracoplasty for the removal of a large Chest. 1 993; 1 03:839-843 .
cicatricial fibrous growth from the interior of the chest, the 4 5 . Varkey B, Rose HD, Kutty CPK, et al. Empyema thoracis dur
result of an old empyema. Med Rec. 1 8 93;44:938. ing a ten-year period. Arch Intern Med. 1 98 1 ; 1 4 1 : 1 77 1 - 1 776.
19. Beck C. Thoracoplasty in America and visceral pleureccomy 46. Maskell NA, Batt S, Hedley EL, et al. The bacteriology of pleu
with report of a case. JAMA. 1 897;28 : 5 8 . ral infection by genetic and standard methods and its mortality
2 0 . Eggers C. Radical operation for empyema. A n n Surg. significance. Am J Respir Crit Care Med. 2006; 17 4: 8 1 7-823.
1 923;77:327. 47. El Solh AA, Alhajjhasan A, Ramadan FH, et al. A comparative
2 1 . Paget S . The Surgery ofthe Chest. Bristol, England: John Wright study of community- and nursing home-acquired empyema
Co; 1 896:204-206. thoracis. J Am Geriatr Soc. 2007; 5 5 : 1 847- 1 852.
22. Graham EA. Some Fundamental Considerations in the Treatment 4 8 . Chen KY, Hsueh PR, Liaw YS , et al. A I O-year experience
ofEmpyema Thoracis. St. Louis, MO: CV Mosby; 1 92 5 : 1 4. with bacteriology of acute thoracic empyema : emphasis on
23. Olch PD, Evarts A. Graham in World War I: the Empyema Klebsiella pneumoniae in patients with diabetes mellicus. Chest.
Commission and Service in the American Expeditionary 2000; 1 1 7: 1 685-1 689.
Forces. J Hist Med Allied Sci. 1 989;44:430-446. 49. Tu C-Y, Hsu W-H, Hsia T-C, et al. The changing pathogens of
24. Scone WJ . The management of postpneumonic empyema complicated parapneumonic effusions or empyemas in a med
based upon 3 1 0 cases. Am J Med Sci. 1 9 1 9; 1 5 8 : 1 -29. ical intensive care unit. Intensive Care Med. 2006;32:570-576.
2 5 . Graham EA, Bell RD . Open pneumothorax: its relations co 5 0 . Tsai TH, Jerng JS, Chen KY, et al. Community-acquired
the treatment of empyema. Am] Med Sci. 1 9 1 8; 1 56:839-87 1 . thoracic empyema in older people. J Am Geriatr Soc.
26. Empyema Commission. Cases of empyema at Camp Lee, Vir 2005 ; 5 3 : 1 203- 1 209.
ginia. JAMA. 1 9 1 8 ;71 :366-373. 5 1 . Brook I. Microbiology of empyema in children and adoles
27. Scone WJ . The management of postpneumonic empyema cents. Pediatrics. 1 990;85 :722-726.
based on 3 1 0 cases. Am ] Med Sci. 1 9 1 9; 1 5 8 : 1 -29. 52. Freij BJ, Kusmiesz H, Nelson JD, et al. Parapneumonic effu
2 8 . Tillett WS, Sherry S, Read CT. The use of strepcokinase sions and empyema in hospitalized children: a retrospective
streptodornase in the treatment of postpneumonic empyema. review of 227 cases. Pediatr Infect Dis. 1 984;3 : 5 78-5 9 1 .
] Thorac Surg. 1 95 1 ;2 1 :275-297. 5 3 . Bartlett JG, Finegold SM. Anaerobic infections o f the lung
29. Glenert J. Sugar levels in pleural effusions of different etiolo and pleural space. Am Rev Respir Dis. 1 974; 1 1 0: 5 6-77.
gies. Acta Tuberc Scand. 1 962;42:222-227. 54. Taryle DA, Potts DE, Sahn SA. The incidence and clinical
30. Light RW, MacGregor Ml, Ball WC Jr, et al. Diagnostic signi correlates of parapneumonic effusions in pneumococcal pneu
ficance of pleural fluid pH and Pco 2• Chest. 1 973;64 : 5 9 1-596. monia. Chest. 1 978;74: 1 70- 1 73 .
242 PLE U RAL D I S EASES
5 5. Musher DM, Alexandraki I, Graviss EA, et al. Bacceremic 77. Kyriacou DN, Stein AC, Yarnold PR, et al. Clinical predic
and nonbacteremic pneumococcal pneumonia. A prospective tors of biotercorism-related inhalational anthrax. Lancet.
study. Medicine (Baltimore). 2000;79:21 0-22 1. 2004;364:449-452.
56. Wolfe WG, Spock A , Bradford WD. Pleural fluid in infants 78. Hairy JE, Bravata DM, Liu H, et al. Systematic review: a cen
and children. Am Rev Respir Dis. 1 968;98:1 027- 1 032. tury of inhalational anthrax cases from 1 900 co 2005. Ann
57. Hendren WH III, Haggerry RJ. Scaphylococcic pneumonia in Intern Med. 2006; 1 44:270-280.
infancy and childhood. JAMA. 1 9 5 8; 1 68:6- 1 6. 79. Evans AF, Oakley RH, Whitehouse GH. Analysis of the
58. Morikawa K, Okada F, Ando Y, et al. Methicillin-resistant chest radiograph in Legionnaires' disease. Clin Radio!.
Staphylococcus aureus and mechicillin-suscepcible S. aureus 1 9 8 1 ;32:361-365.
pneumonia: comparison of clinical and chin-section CT find 80. Tan MJ, Tan JS, Hamor RH, et al. The radiologic manifesta
ings. Br J Radio!. 20 1 2;8 5:el68-el75. tions of Legionnaire's disease. Chest. 2000; 1 1 7:398-403.
59. Kaye MG, Fox MJ, Barclecc JG, et al. The clinical spec 8 1 . Krobath FJ, Yu VL, Reddy SC, et al. Clinicoradiographic cor
trum of Staphylococcus aureus pulmonary infection. Chest. relation with extent of Legionnaire disease. AJR Am J Roent
1 990;97:788-792. genol. 1 983; 1 4 1 :263-268.
60. Lobo LJ, Reed KD, Wunderink RG. Expanded clinical pre 82. Randolph KA, Beekman JF. Legionnaires' disease presenting
sentation of communiry-acquiced MRSA pneumonia. Chest. with empyema. Chest. 1 979;75:404-406.
20 1 0; 1 38:1 30- 1 3 6. 83. Pettersson T, Nyberg P, Nordstrom D, et al. Similar pleural
6 1 . Sexauer WP, Quezada Z, Lippmann ML, et al. Pleural effu fluid findings in pleuropulmonary tularemia and tuberculous
sions in right-sided endocarditis: characteristics and patho pleurisy. Chest. 1 996; 1 09:572-575.
physiology. South Med] 1 992;85: 1 1 76-1 1 80. 84. Chung KM, Chou DW, Chen CH, et al. Lymphocytic
62. Welch CC, Tombridge TL, Baker WJ, et al. Bera-hemolytic pleural effusion in acute melioidosis. J Formos Med Assoc.
streptococcal pneumonia: report of an outbreak in a military 2007; 1 06:874-877.
population. Am ] Med Sci. 1 96 1 ;242: 1 57-1 65. 85. Patel SB, Mahler R. Clostridial pleuropulmonary infec
63. Basiliere JL, Bistrong HW, Spence WF. Streptococcal pneu tions: case report and review of the literature. J Infect.
monia: recent outbreaks in military recruit populations. Am J 1 990;2 1 : 8 1 -85.
Med. 1 968;44:5 80-5 89. 86. Al-Anazi AR, Aziz S, Fouda MA. Brucellosis: haemorrhagic
64. Braman SS, Donat WE. Explosive pleuritis. Manifestation of pleural effusion. Med Prine Pract. 2005;14: 1 1 8- 1 20.
group A beta-hemolytic streptococcal infection. Am J Med. 87. Cooney TG, Harwood BR, Meisner DJ. Haemophi
1 986;8 1:723-726. lus parainjluenzae thoracic empyema. Arch Intern Med.
65. Tillotson JR, Lerner AM Characteristics of pneumonias
. 1 9 8 1 ; 1 4 1 :940-94 1 .
caused by Escherichia coli. N Engl] Med. 1 967;277: 1 1 5-1 22. 88. Bekemeyer WB, Zimmerman GA. Life-threatening complica
66. Winer-Muram HT, Jennings SG, Wunderink RG, et al. Ven tions associated with Bacillus cereus pneumonia. Am Rev Respir
tilator-associated Pseudomonas aeruginosa pneumonia: radio Dis. 1 985; 1 3 1 :466-469.
graphic findings. Radiology. 1 995; 1 95:247-252. 89. Madrazo A, Henderson MD, Baker L, et al. Massive empyema
67. Shah RM, Wechsler R, Salazar AM, et al. Spectrum of CT due co Citrobacter diversus. Chest. 1 975;68: 1 04-1 06.
findings in nosocomial Pseudomonas aeruginosa pneumonia. 90. Mazzulli T, Salit IE. Pleural fluid infection caused by Lis
J 7horac Imaging. 2002; 1 7:53-57. teria monocytogenes: case report and review. Rev Infect Dis.
68. Lin YT, ChenTL, Siu LK, et al. Clinical and microbiological char 1 9 9 1 ; 1 3:564-570.
acteristics of communiry-acquired rhoracic empyema or compli 9 1 . Van De Water JM. The treatment of pleural effusion compli
cated parapneumonic effusion caused by Klebsiella pneumoniae cating pneumonia. Chest. 1 970;57:259-262.
in Taiwan. Eur J Clin Microbio! Infect Dis. 20 1 0;29: 1 003-1 0 1 0. 92. Sahn SA, Lakshminarayan S, Char DC. "Silent" empyema
69. Okada F, Ando Y, Tanoue S, et al. Radiological findings in in patients receiving corticosteroids. Am Rev Respir Dis.
acute Haemophilus injluenzae pulmonary infection. Br J 1 973; 1 07:873-876.
Radio!. 20 1 2;85: 1 2 1 - 1 26. 93. Arancibia F, Ewig S, Martinez JA, et al. Antimicrobial treat
70. Asmar Bl, Slavis TL, Reed JO, et al. Hemophilus influenzae ment failures in patients with communiry-acquired pneumo
rype b pneumonia in 43 children.] Pediatr. 1 978;93:389-393. nia: causes and prognostic implications. Am J Respir Crit Care
7 1 . Ginsburg CM, Howard JB, Nelson JD. Report of 65 Med. 2000; 1 62: 1 54-1 60.
cases of Haemophilus injluenzae b pneumonia. Pediatrics. 94. Brixey AG, Luo Y, Skouras V, et al. The efficacy of chest
l 979;64:283-286. radiographs co detect parapneumonic effusions. Respirology.
72. Levin DC, Schwarz Ml, Matthay RA, et al. Bacceremic Hae 20 1 1 ; 1 6: 1 000-1 004.
mophilus injluenzae pneumonia in adults: a report of 24 cases 95. Kearney SE, Davies CW, Tattersall DJ, et al. The characteristics
and a review of the literature. Am ] Med. 1 977;62:21 9-223. and significance of thoracic lymphadenopachy in parapneu
73. Okada F, Ando Y, Honda K, et al. Clinical and pulmonary monic effusion and empyema. Br J Radio!. 2000;73:583-587.
rhin-section CT findings in acute Klebsiella Pneumoniae pneu 96. Kearney SE, Davies CW, Davies RJ, et al. Computed tomog
monia. Eur Radio!. 2009; 1 9:809-8 1 5. raphy and ultrasound in parapneumonic effusions and empy
74. Tillotson JR, Lerner AM Characteristics of pneumonias
. ema. Clin Radio!. 2000;5 5:542-547.
caused by Bacillus proteus. Ann Intern Med. l 968;68:287-294. 97. Chen CH, Chen W, Chen HJ, et al. Transrhoracic ultrasonog
75. Shafazand S, Doyle R, Ruoss S, et al. lnhalacional anrhrax: raphy in predicting the outcome of small-bore catheter drain
epidemiology, diagnosis, and management. Chest. age in empyemas or complicated parapneumonic effusions.
1 999; 1 1 6: 1 369-1 376. Ultrasound Med Biol. 2009;35:350-354.
76. Jernigan JA, Stephens DS, Ashford DA, et al. Bioterrorism 98. Moffett BK, Panchabhai TS, Anaya E, et al. Computed
related inhalational anthrax: the first 10 cases reported in the tomography measurements of parapneumonic effusion indica
United States. Emerg Infect Dis. 200 1 ;7:933-944. tive of thoracentesis. Eur Respir J 20 1 1 ;38: 1406-1 4 1 1.
CHAPTER 12 I PARAP N E U M O N I C E F F U S I O N S A N D E M PY E M A 243
99. Skouras V, Awdankiewicz A, Light RW: What size parapneu 1 22. Chen SC, Chen W, Hsu WH, et al. Role of pleural fluid
monic effusions should be sampled? Thorax. 2 0 1 0;65 : 9 1 . C-reactive protein concentration in discriminating uncom
1 00. Xiol X , Castellvi JM, Guardiola J , e t al. Spontaneous bac plicated parapneumonic pleural effusions from com
terial empyema in cirrhotic patients: a prospective study. plicated parapneumonic effusion and empyema. Lung.
Hepatology. 1 9 96;23:71 9-723. 2006; 1 84 : 1 4 1 - 1 45.
101. Menzies SM, Rahman NM, Wrightson JM, et al. Blood 1 23. Skouras V, Boulcadakis E, Nikoulis D, et al. Prognostic value
culture botcle culture of pleural fluid in pleural infection. of C-reactive protein in parapneumonic effusions. Respirol
Thorax. 2 0 1 1 ;66:65 8-662. ogy. 20 1 1 ; 1 7:308-3 1 4.
1 02. Maskell NA, Batt S, Hedley EL, et al. The bacteriology of pleu 1 24. Poree! JM, Galindo C, Esquerda A, et al. Pleural fluid inter
ral infection by genetic and standard methods and its mortality leukin-8 and C-reactive protein for discriminating compli
significance. Am J Respir Crit Care Med. 2006; 1 74 : 8 1 7-823. cated non-purulent from uncomplicated parapneumonic
1 03. Maskell NA, Davies CW, Nunn AJ, et al. U.K. controlled effusions. Respirology. 2008; 1 3 : 5 8-62.
trial of intrapleural streptokinase for pleural infection. 1 2 5 . Oikonomidi S, Kostikas K, Kalomenidis I, et al. Matrix
N Engl] Med. 2005;352:865-874. metalloproteinase levels in the differentiation of parapneu
1 04. Sullivan KM, O'Toole RD, Fisher RH, et al. Anaerobic monic pleural effusions. Respiration. 2 0 1 0;80:285-29 1 .
empyema thoracis. Arch Intern Med. 1 973; 1 3 1 : 5 2 1 -527. 1 26. Tsilioni I, Kostikas K , Kalomenidis I, e t al. Diagnostic accu
1 05. Massard G, Wihlm JM. Early complications. esophagopleu racy of biomarkers of oxidative stress in parapneumonic
ral fistula. Chest Surg Clin N Am. 1 999;9:6 1 7-63 1 . pleural effusions. EurJ Clin Invest. 20 1 0;4 1 :349-3 5 6.
1 06. Landay MJ, Christensen EE, Bynum LJ, e t al. Anaerobic 1 27. Poree! JM. Pleural fluid tests to identify complicated para
pleural and pulmonary infections. AJR Am J Roentgenol. pneumonic effusions. Curr Opin Pulm Med. 20 1 O;1 6:357-36 1 .
1 980; 1 34:233-240. 1 28. McLoud TC, Flower CD. Imaging the pleura: sonog
1 07. Colice GL, Curtis A, Deslauriers J, et al. Medical and sur raphy, CT, and MR imaging. AJR Am J Roentgenol.
gical treatment of parapneumonic effusions : an evidence 1 99 1 ; 1 56: 1 1 4 5- 1 1 5 3.
based guideline. Chest. 2000; 1 1 8 : 1 1 5 8-1 1 7 1 . 1 29. Yang PC, Luh KT, Chang DB, et al. Value of sonography in
1 08. Cheng D-S, Rodriguez RM , Rogers J , e t al. Comparison of determining the nature of pleural effusion: analysis of 320
pleural fluid pH values obtained using blood gas machine, pH cases. AJR Am j Roentgenol. 1 992; 1 5 9 :29-33.
meter, and pH indicator scrip. Chest. 1 998;1 1 4 : 1 368-1 372. 1 30. Himelman RB, Callen PW. The prognostic value of
1 09. Pine JR, Hollman JL. Elevated pleural fluid pH in Proteus loculations in parapneumonic pleural effusions. Chest.
mirabilis empyema. Chest. 1 98 3;84: 1 09- 1 1 1 . 1 986;90 : 8 5 2-856.
1 1 0. Sahn SA, Taryle DA, Good JT Jr. Experimental empyema: 1 3 1 . Lin FC, Chou CW, Chang SC. Usefulness of the sus
time course and pathogenesis of pleural fluid acidosis and low pended microbubble sign in differentiating empyemic and
pleural fluid glucose. Am Rev Respir Dis. 1 979; 1 2 0 : 3 5 5-36 1 . nonempyemic hydropneumothorax. J Ultrasound Med.
1 1 1. Light RW, Moller D J Jr, George RB. Low pleural fluid p H in 200 1 ;20 : 1 3 4 1 - 1 345.
parapneumonic effusion. Chest. l 975;68:273-274. 1 32. Chen HJ, Yu YH, Tu CY, et al. Ultrasound in periph
1 1 2. Maskell NA, Gleeson FY, Darby M, et al. Diagnostically eral pulmonary air-fluid lesions: color Doppler imaging
significant variations in pleural fluid pH in loculated para as an aid in differentiating empyema and abscess. Chest.
pneumonic effusions. Chest. 2004; 1 26:2022-2024. 2009; 1 3 5 : 1 426- 1 432.
1 1 3. Berger HA, Morganroth ML. Immediate drainage is not 1 33. Stark DD, Federle MP, Goodman PC, et al. Differentiat
required for all patients with complicated parapneumonic ing lung abscess and empyema: radiography and computed
effusions. Chest. 1 990;97:73 1 -735. tomography. AJR Am j Roentgenol. 1 983; 1 4 1 : 1 63- 1 67.
1 1 4. Poe RH , Matthew GM, Israel RH, et al. Utility of pleural 1 34. Light RW. A new classification of parapneumonic effusions
fluid analysis in predicting cube thoracostomy/decortication and empyema. Chest. 1 99 5 ; 1 08:299-3 0 1 .
in parapneumonic effusions. Chest. 1 9 9 1 ; 1 00:963-967. 1 3 5 . Teixeira LR, Sasse SA, Villarino MA, e t al. Antibiotic levels
1 1 5. Light RW: Management of parapneumonic effusions. Chest. in empyemic pleural fluid. Chest. 2000; 1 1 7: 1 734- 1 739.
1 9 9 1 ; 1 00:8 92-893. 1 36. Liapakis IE, Kottakis I, Tzatzarakis MN, et al. Penetration
1 1 6. Heffner JE, Brown LK, Barbieri C, et al. Pleural fluid chemi of newer quinolones in the empyema fluid. Eur Respir ].
cal analysis in parapneumonic effusions. A meta-analysis. 2004;24:466-470.
Am ] Respir Crit Care Med. 1 99 5 ; 1 5 1 : 1 700- 1 708. 1 37. Liapakis IE, Light RW, Pitiakoudis MS, et al. Penetration
1 1 7. Jimenez Castro D, Diaz Nuevo G, Sueiro A, et al. Pleural of clarithromycin in experimental pleural empyema model
fluid parameters identifying complicated parapneumonic fluid. Respiration. 2005;72:296-300.
effusions. Respiration. 2005;72:357-364. 1 3 8. Saroglou M, Ismailos G, Tryfon S, et al. Penetration of
1 1 8. Kalomenidis I, Bourns D. Pleural fluid pH in parapneu azithromycin in experimental pleural empyema fluid. Eur J
monic pleural effusions: drawing the line. Respiration. Pharmacol. 2 0 1 0;626;27 1 -275.
2005;72: 345-343. 1 39. Saroglou M , Tryfon S, Ismailos G, et al. Pharmacokinetics
1 1 9. Aleman C, Alegre J, Segura RM , et al. Polymorphonuclear of linezolid and ercapenem in experimental parapneumonic
elastase in the early diagnosis of complicated pyogenic pleu pleural effusion. ] Inflamm (Lond). 2 0 1 0;7:22.
ral effusions. Respiration. 2003;70:462-467. 1 40. Mandell LA, Barltett JG, Dowell SF, et al. Update of prac
1 20. Alegre J, Jufresa J, Segura R, et al. Pleural-fluid myeloper tice guidelines for the management of community-acquired
oxidase in complicated and noncomplicated parapneumonic pneumonia in immunocompetent adulcs. Clin Infect Dis.
pleural effusions. Eur Respir]. 2002; 1 9:320-325. 2003;37: 1405-1 433.
121. Poree! JM, Vives M , Cao G, et al. Biomarkers of infection 1 4 1 . Wrightson JM, Davies RJ. The approach to the patient with
for the differential diagnosis of pleural effusions. Eur Respir a parapneumonic effusion. Semin Respir Crit Care Med.
]. 2009;34 : 1 383- 1 389. 20 1 0;3 1 :706-7 1 5.
244 PLE U RAL D I S EASES
1 42. Rahman NM, Chapman SJ, Davies RJ. The approach to 1 64. Athanassiadi K, Gerazounis M, Kalantzi N. Treatment of
the patient with a parapneumonic effusion. Clin Chest Med. post-pneumonic empyema thoracis. Thorac Cardiovasc Surg.
2006;27:253-266. 2003;5 l :338-34 1 .
143. Davies CW, Gleeson FY, Davies RJ. BTS guidelines for the 1 6 5 . Misthos P, Sepsas E, Konstantinou M, et al. Early use of
management of pleural infection. Thorax. 2003;5 8 (suppl intrapleural fi b rinolytics in the management of postpneu
2) :ii l 8-ii28 . monic empyema. A prospective srudy. Eur J Cardiothorac
1 44 . Davies H E , Davies RJ, Davies C W, e t al . Management Surg. 2005;28 : 5 99-603.
of pleural infection in adults : British Thoracic Soci 1 66. Rahman NM. Intrapleural agents for pleural infec
ety Pleural Disease Guideline 20 1 0 . Thorax. 20 1 0;65 tion: fibrinolytics and beyond. Curr Opin Pulm Med.
(suppl 2): ii4 1 -ii53. 20 1 2; 1 8 :326-332.
145. Clagett OT, Geraci JE. A procedure for the management of 1 67. Light RW, Nguyen T, Mulligan ME, et al. The in vitro
postpneumonecromy empyema. J Thorac Cardiovasc Surg. efficacy of varidase versus streptokinase or urokinase for
1 963;45 : 1 4 1 - 1 4 5 . liquefying thick purulent exudative material from loculated
1 46. Dieter RA Jr, Pifarre R , Neville WE, e t al. Empyema treated empyema. Lung. 2000; 1 78 : 1 3- 1 8 .
with neomycin irrigation and closed-chest drainage. J Thorac 1 6 8 . Bergh NP, Ekroth R , Larsson S, e t al. !ntrapleural screptoki
Cardiovasc Surg. 1 970;59 :496-5 00. nase in the treatment of haemothorax and empyema. Scand
1 47. Rosenfeldt FL, McGibney D, Braimbridge MY, et al. Com J Thorac Cardiovasc Surg. 1 977; 1 1 :265-268 .
parison between irrigation and conventional treatment for 1 69 . Diacon A H , Theron J, Schuurmans M M , e t al. Intrapleural
empyema and pneumonectomy space infection. Thorax. streptokinase for empyema and complicated parapneumonic
1 9 8 1 ;36:272-277. effusions. Am J Respir Crit Care Med. 2004; 1 70:49-5 3 .
1 4 8 . Hutter JA, Harari D, Braimbridge MV The management 1 70. Chin NK, Lim TK. Controlled trial o f intrapleural strepto
of empyema thoracis by thoracoscopy and irrigation. Ann kinase in the treatment of pleural empyema and complicated
Thorac Surg. 1 9 8 5;39 : 5 1 7-520. parapneumonic effusions. Chest. 1 997; 1 1 1 :275-279.
149. Hakim M , Milstein BB. Empyema thoracis and infected 1 7 1 . Davies RJO, Traill ZC, Gleeson FV. Randomised controlled
pneumonectomy space: case for cyclical irrigation. Ann Tho trial of intrapleural streptokinase in community acquired
rac Surg. 1 9 86;4 1 : 8 5-8 8 . pleural infection. Thorax. 1 997;52:4 1 6-42 1 .
1 5 0. Storm HKR, Krasnik M , Bang K , e t a l . Treatment o f pleural 1 72. Bouras D , Schiza S , Tzanakis N , et al. Intrapleural urokinase
empyema secondary to pneumonia: thoracocentesis regimen versus normal saline in the treatment of complicated parap
versus tube drainage. Thorax. 1 9 92;47 : 8 2 1 -824. neumonic effusions and empyema. A randomized, double
1 5 1 . Cham CW, Haq SM, Rahamim J. Empyema thoracis: a blind study. Am ] Respir Crit Care Med. 1 999; 1 5 9 :37-42 .
problem with late referral? Thorax. 1 993;48:925-927. 1 73 . Tuncozgur B, Ustunsoy H, Sivrikoz MC, et al. Intra
1 52. Sasse S, Nguyen T, Teixeira LR, et al. The utility of daily pleural urokinase in the management of parapneumonic
therapeutic thoracentesis for the treatment of early empy empyema: a randomised controlled trial. Int J Clin Pract.
ema. Chest. 1 999;1 1 6 : 1 703- 1 70 8 . 200 1 ; 5 5 : 6 5 8-660.
1 53 . Simmers TA, Jie C, S i e B . Minimally invasive treatment o f 1 74. Bouras D, Schiza S, Patsourakis G, et al. Intrapleural strep
thoracic empyema. Thorac Cardiovasc Surg. 1 999;47:77-8 1 . tokinase versus urokinase in the treatment of complicated
1 54. Ferguson AD, Prescott RJ, Selkon JB, et al. The clini parapneumonic effusions: a prospective, double-blind study.
cal course and management of thoracic empyema. QJM. Am ] Respir Crit Care Med. 1 997; 1 5 5 :29 1 -2 9 5 .
1 996;89:285-28 9 . 1 7 5 . Jerjes-Sanchez C, Ramirez-Rivera A , Elizalde JJ, et al. Intra
1 5 5 . Ozol D, Oktem S, Erdinc E. Complicated parapneumonic pleural fibrinolysis with streptokinase as an adjunctive treat
effusion and empyema thoracis: microbiologic and thera ment in hemothorax and empyema: a multicenter trial.
peutic aspects. Respir Med. 2006; 1 00:286-29 1 . Chest. 1 996; 1 09: 1 5 1 4- 1 5 1 9 .
1 56. Light RW. Pleural controversy: optimal chest tube size for 1 76. Laisaar T, Puttsepp E , Laisaar V Early administration of
drainage. Respirology. 20 1 1 ; 1 6:244-248. intrapleural streptokinase in the treatment of multiloculated
1 57. Rahman NM, Maskell NA, Davies CW, et al. The relation pleural effusions and pleural empyemas. Thorac Cardiovasc
ship between chest tube size and clinical outcome in pleural Surg. 1 996;44:252-256.
infection. Chest. 20 1 0; 1 37:5 36-543. 1 77. Ternes RT, Follis F, Kessler RM , et al. Intrapleural fibri
1 5 8 . Shankar S, Gulati M, Kang M, et al. Image-guided percutan nolytics in management of empyema thoracis. Chest.
eous drainage of thoracic empyema: can sonography predict 1 996; 1 1 0: 1 02- 1 06.
the outcome? Eur Radio!. 2000; 1 0 :49 5-499. 1 7 8 . Moulton JS, Benkert RE, Weisiger KH , et al. Treat
1 5 9 . Ali I, Unruh H . Management of empyema thoracis. Ann ment of complicated pleural fluid collections with
Thorac Surg. 1 990;5 0 : 3 5 5-3 5 9 . image-guided drainage and intracavitary urokinase. Chest.
1 60. Ashbaugh DG. Empyema thoracis. Factors influencing mor 1 99 5 ; 1 0 8 : 1 252- 1 259.
bidity and mortality. Chest. 1 9 9 1 ;99: 1 1 62-1 1 6 5 . 1 79 . Zhu Z, Hawthorne ML, Guo Y, et al. Tissue plasminogen
1 6 1 . Kerr A , Vasudevan VP, Powell S, e t al. Percutaneous catheter activator combined with human recombinant deoxyribo
drainage for acute empyema. Improved cure rate using CAT nuclease is effective therapy for empyema in a rabbit model.
scan, fluoroscopy, and pigtail drainage catheters. N Y State J Chest. 2006; 1 29: 1 577- 1 5 8 3 .
Med. 1 9 9 1 ;9 1 :4-7. 1 8 0. Tokuda Y, Matsushima D, Stein GH, et a l . Intrapleural
1 62. Sherman MM, Subramanian V, Berger RL. Management of fibrinolytic agents for empyema and complicated parapneu
thoracic empyema. Am ] Surg. 1 977; 1 33:474-479. monic effusions: a meta-analysis. Chest. 2006;1 29 :783-790 .
1 63 . Lim TK, Chin NK. Empirical treatment with fibrinolysis 1 8 1 . Simpson G, Roomes D, Heron M . Effects of streptokinase
and early surgery reduces the duration of hospitalization in and deoxyribonuclease on viscosity of human surgical and
pleural sepsis. Eur Respir J 1 999;1 3 : 5 1 4-5 1 8 . empyema pus. Chest. 2000;1 1 7: 1 728- 1 733.
CHAPTER 12 I PARAP N E U M O N I C E F F U S I O N S A N D E M PY E M A 245
1 82. Simpson G, Roomes D, Reeves B. Successful treatment of 202. Jimenez Castro D, Diaz G, Perez-Rodriguez E, et al. Prognos
empyema throacis with human recombinant deoxyribonu tic features of residual pleural thickening in parapneumonic
clease. Thorax. 2003;58 :365-366. pleural effusions. Europ Respir J 2003;2 1 : 952-9 5 5 .
1 83 . Rahman NM, Maskell NA, West A, et al. Intrapleural use of 2 0 3 . Deslauriers ], Jacques LF, Gregoire J . Role of Eloesser flap
tissue plasminogen activator and DNase in pleural infection. and thoracoplasty in the third millennium. Chest Surg Clin
N Engl] Med. 2 0 1 1 ;365 : 5 1 8-526. N Am. 2002; 1 2:605-623.
1 84. Ruiz A, Poree! JM, Madronero AB, et al. Hemothorax fol 204. Samson PC. Empyema thoracis: essentials of present-day
lowing administration of intrapleural alteplase. Respiration. management. Ann Thorac Surg. 1 97 1 ; 1 l : 2 1 0-220.
2006;73:7 1 5 . 205 . Morin JE, Munro DD, MacLean LD. Early thoracotomy for
1 8 5 . Chai FY, Kuan YC. Massive hemothorax following admin empyema. J Thorac Cardiovasc Surg. 1 972;64 : 530-536.
istration of intrapleural streptokinase. Ann Thorac Med. 206. Chu MW, Dewar LR, Burgess JJ, et al. Empyema thoracis:
2 0 1 1 ;6 : 1 49-1 5 1 . lack of awareness results in a prolonged clinical course. Can
1 86. Silen ML, Naunheim KS . Thoracoscopic approach to the J Surg. 200 1 ;44:284-2 8 8 .
management of empyema thoracis. Indications and results. 2 0 7 . Urine G E , Oz�elik U, Kiper N, e t al. Pediatric pleural effu
Chest Surg Clin N Am. 1 996;6:49 1 -499. sions: etiological evaluation in 492 patients over 29 years.
1 87. de Souza A, Offner PJ, Moore EE, et a!. Optimal management Turk] Pediatr. 2009 ;5 1 :2 1 4-2 1 9 .
of complicated empyema. Am J Surg. 2000; 1 8 0 : 5 07-5 1 1 . 208. Schultz KD, Fan LL, Pinsky J, e t al. Th e changing face of
1 8 8 . Roberts JR. Minimally invasive surgery in the treatment of pleural empyemas in children: epidemiology and manage
empyema: intraoperative decision making. Ann Thorac Surg. ment. Pediatrics. 2004; 1 1 3 : 1 735-1740.
2003;76:225-230. 209. Byington CL, Spencer LY, Johnson TA, et al. An epidemi
1 8 9 . Landreneau RJ , Keenan RJ, Hazelrigg SR, et al. Tho ological investigation of a sustained high rate of pediatric
racoscopy fo r empyema and hemothorax. Chest. parapneumonic empyema: risk factors and microbiological
1 9 9 5 ; 1 09 : 1 8-24. associations. Clin Infect Dis. 2002;34:434-440.
1 9 0 . Cassina PC, Hauser M , Hillejan L, et al. Video-assisted 2 1 0 . Eastham KM , Freeman R, Kearns AM, et al. Clinical fea
thoracoscopy in the treatment of pleural empyema: stage tures, aetiology and ourcome of empyema in children in the
based management and outcome. J Thorac Cardiovasc Surg. north east of England. Thorax. 2004; 5 9 : 5 22-52 5 .
1 9 99; 1 1 7:234-23 8 . 2 1 1 . Strachan RE, Cornelius A , Gilbert G L , e t a l . A bedside assay
1 9 1 . Lawrence D R , Ohri SK, Moxon RE, e t a l . Thoracoscopic to detect Streptococcus pneumoniae in children with empy
debridement of empyema thoracis. Ann Thorac Surg. ema. Pediatr Pulmonol. 20 1 1 ;46: 1 79- 1 83.
1 997;64: 1 448- 1 45 0 . 212. Blaschke AJ, Heyrend C, Byington CL, et al. Molecular
1 92. Striffeler H, Gugger M, Im H o f V, et a l . Video-assisted tho analysis improves pathogen identification and epidemiologic
racoscopic surgery for fibrinopurulent pleural empyema in study of pediatric parapneumonic empyema. Pediatr Infect
67 patients. Ann Thorac Surg. 1 998;65 :3 1 9-323. Dis J 2 0 1 0;40:289-294.
1 93 . Luh SP, Chou MC, Wang LS, et al. Video-assisted thora 2 1 3 . Grijalva CG, Pekka Nuorti J, Zhu Y, et al. Increasing inci
coscopic surgery in the treatment of complicated parapneu dence of empyema complicating childhood community
monic effusions or empyemas: outcome of 234 patients. acquired pneumonia in the United States. Clin Infect Dis.
Chest. 2005 ; 1 27: 1 427- 1 432. 2 0 1 0;50: 805-8 1 3 .
1 94 . Waller DA, Rengarajan A, Nicholson FH, et al. Delayed 2 1 4 . L i ST, Tancredi DJ . Empyema hospitalizations increased in
referral reduces the success of video-assisted thoracoscopic US children despite pneumococcal conjugate vaccine. Pedia
debridement for post-pneumonic empyema. Respir Med. trics. 20 1 0 :26-33.
200 1 ; 9 5 : 8 36-840. 2 1 5 . Strachan RE, Cornelius A, Gilbert GL, et al. Bacterial causes
1 9 5 . Wait MA, Sharma S, Hohn J, et al. A randomized trial of of empyema in children, Australia, 2007-2009. Emerg Infect
empyema therapy. Chest. 1 997; 1 1 1 : 1 548-1 5 5 1 . Dis. 20 1 1 ; 1 7: 1 839- 1 845.
1 96. Bilgin M , Akcali Y, Oguzkaya F. Benefits o f early aggres 2 1 6. Wu PS, Huang LM, Chang IS, et al. The epidemiology of
sive management of empyema thoracis. ANZ J Surg. hospitalized children with pneumococcal/lobar pneumonia
2006;76: 1 20- 1 22. and empyema from 1 997 to 2004 in Taiwan. Eur J Pediatr.
1 97. Thurer RJ. Decortication in thoracic empyema. Indica 20 1 0 ; 1 69 : 8 6 1 -866.
tions and surgical technique. Chest Surg Clin N Am. 2 1 7. Tobin CL, Lee YC (Gary) . Pleural Infection-what we need
1 9 96;6:46 1-490. to know but don't. Current Opin Pul Dis. 20 1 2; 1 8 :32 1-325.
198. Cardillo G, Carleo F, Carbone L, et al. Chronic postpneu 2 1 8 . Mukherjee S, Langroudi B , Rosenthal M, et al. Incidence
monic pleural empyema: comparative merits of thoraco and outcome of scoliosis in children with pleural infection.
scopic versus open decortication. Eur J Cardiothorac Surg. Pediatr Pulmonol. 2007;42:22 1 -224.
2009;36:9 1 4-9 1 8 . 2 1 9 . Cirino LM, Gomes FM, Batista BN. The etiology of exten
1 9 9 . Zahid I, Nagendran M , Routledge T, e t al. Comparison of sive pleural effusions with troublesome clinical course among
video-assisted thoracoscopic surgery and open surgery in the children. Sao Paulo Med J 2004; 1 22:269-272 .
management of primary empyema. Curr Opin Pulm Med. 220. Balfour-Lynn IM. Some consensus but little evidence: guide
20 1 1 ; 1 7:255-25 9 . lines on management of pleural infection in children. Tho
2 0 0 . Tacconi F, Pompeo E, Fabbi E, e t a l . Awake video-assisted rax. 2005;60: 94-96.
pleural decortication for empyema thoracis. EurJ Cardiotho 22 1 . Balfour-Lynn IM, Abrahamson E, Cohen G, et al. BTS
rac Surg. 2 0 1 0;37:594-60 1 . guidelines for the management of pleural infection in chil
20 1 . Neff CC, van Sonnenberg E, Lawson DW, et al. CT follow dren. Thorax. 2005;60 (suppl l ) : i l -i2 1 .
up of empyemas: pleural peels resolve after percutaneous 222. Hoff SJ, Neblett WW, Edwards KM , e t al. Parapneumonic
catheter drainage. Radiology. 1 990; 1 76 : 1 95-197. empyema in children: decortication hastens recovery in
246 PLE U RAL D I S EASES
patients with severe pleural infections. Pediatr Infect Dis ]. of early postpneumonectomy empyema. J Thorac Cardiovasc
1 99 1 ; 1 0 : 1 94- 1 99 . Surg. 1 998; 1 1 6: 943-948 .
223. Gates RL, Caniano DA, Hayes J R , e t a l . Does VATS provide 240. Wain JC. Management o f late postpneumonectomy empy
optimal treatment of empyema in children? A systematic ema and bronchopleural fistula. Chest Surg Clin N Am.
review. J Pediatr Surg. 2004;39:38 1-386. 1 996;6: 529-54 1 .
224. Karaman I, Erdogan D, Karaman A, et al. Comparison of 24 1 . Pairolero PC. Empyema and bronchopleural fistula after
closed-tube thoracostomy and open thoracoromy procedures pneumonectomy: factors affecting incidence. Ann Thorac
in the management of thoracic empyema in childhood. EurJ Surg. 200 1 ;72:243-247.
Pediatr Surg. 2004; 1 4:250-254. 242. Ratto GB, Fantino G, Tassara E, et al. Long-term antimicro
225. Padman R, King KA, Iqbal S, et al. Parapneumonic effusion bial prophylaxis in lung cancer surgery: correlation between
and empyema in children: retrospective review of the duPont microbiological findings and empyema development. Lung
experience. Clin Pediatr (Phi/a). 2007;46 : 5 1 8-522. Cancer. 1 994; 1 1 :345-352.
226. Wells RG, Havens PL. lntrapleural fibrinolysis for parap 243 . Virkkula L, Eerola S . Treatment of postpneumonectomy
neumonic effusion and empyema in children. Radiology. empyema. ScandJ Thorac Cardiovasc Surg. 1 974; 8 : 1 33- 1 37.
2003;228: 370-378. 244. Ueda H, Shibata K, Kusano T. Postoperative pyothorax. Surg
227. Singh M, Mathew J, Chandra S, et al. Randomized con Today. 1 992;22: 1 1 5-1 1 9 .
trolled trial of intrapleural streprokinase in empyema thora 245 . Choe d u H, Lee B H , Kim KH , e t al. Volume-expanding
cis in children. Acta Paediatr. 2004;93 : 1 443- 1 44 5 . complications after pneumonectomy: comparison of CT
228. Thomson A H , Hull J, Kumar M R , et a l . Randomised trial o f findings. Clin Imaging. 2006;30: 1 73- 1 76.
intrapleural urokinase i n the treatment of childhood empy 246. Karkola P, Kairaluoma Ml, Larmi TKI. Postpneumonec
ema. Thorax. 2002;57:343-347. tomy empyema in pulmonary carcinoma patients. J Thorac
229. Scarci M, Zahid I, Bille A, et al. Is video-assisted thora Cardiovasc Surg. 1 976;72:3 1 9-322.
coscopic surgery the best treatment for paediatric pleural 247. Goldstraw P. Treatment of postpneumonectomy empyema:
empyema? Interact Cardiovasc Thorac Surg. 20 1 1 ; 1 3:70-76. the case for fenestration. Thorax. 1 979;34:740-745 .
230. Doski JJ, Lou D, Hicks BA, et al. Management of para 248. Gossot D, Stern J B , Galetta D, e t al. Tboracoscopic manage
pneumonic collections in infants and children. J Pediatr ment of postpneumonectomy empyema. Ann Thorac Surg.
Surg. 2000;3 5 : 265-270. 2004;78 :273-276.
23 1 . Bishay M, Short M, Shah K, et al. Efficacy of video-assisted 249. Schneiter D, Grodzki T, Lardinois D , et al. Accelerated treat
thoracoscopic surgery in managing childhood empyema: a ment of postpneumonectomy empyema: a binational long
large single-centre study. J Pediatr Surg. 2009;44:337-342. term study. J Thorac Cardiovasc Surg. 2008; 1 36: 1 79- 1 8 5 .
232. Sonnappa S, Cohen G, Owens CM, et al. Comparison of 250. Hysi I, Rousse N, Claret A , e t al. Open window thoracos
urokinase and video-assisted thoracoscopic surgery for treat tomy and thoracoplasty to manage 90 postpneumonectomy
ment of childhood empyema. Am J Respir Crit Care Med. empyemas. Ann Thorac Surg. 20 1 1 ;92: 1 833- 1 839.
2006; 1 74:22 1-227. 25 1 . Shamji FM, Ginsberg RJ, Cooper JD, et al. Open window
233. St Peter SD, Tsao K, Harrison C, et al. Tboracoscopic decor thoracostomy in the management of postpneumonectomy
tication vs tube thoracostomy with fibrinolysis for empyema empyema with or without bronchopleural fistula. J Thorac
in children: a prospective, randomized trial. J Pediatr Surg. Cardiovasc Surg. l 983;86: 8 1 8-822.
2009;44: 1 06- 1 1 1 . 252. Bayes AJ, Wilson JA, Chiu RC, et al. Clagett open-win
234. Liu HP, Hsieh MJ, Lu HI, et al. Tboracoscopic-assisted dow thoracostomy in patients with empyema who had
management of postpneumonic empyema in children refrac and had not undergone pneumonectomy. Can J Surg.
rory to medical response. Surg Endosc. 2002; 1 6: 1 6 1 2-1 6 1 4. l 987;30:329-33 1 .
235. Satish B, Bunker M, Seddon P. Management of thoracic 253. Pairolero PC, Arnold PG, Trastek VF, et al. Postpneumonec
empyema in childhood: does the pleural thickening matter? tomy empyema. The role of intrathoracic muscle transposi
Arch Dis Child. 2003 ; 8 8 : 9 1 8-92 1 . tion. J Thorac Cardiovasc Surg. l 990;99 : 9 5 8-968.
236. Mahant S , Cohen E , Weinstein M , et al. Video-assisted 254. Saito H, Tatsuzawa T, Kikkawa H , et al. Transpericar
thorascopic surgery vs chest drain with fibrinolyrics for dial bronchial closure with omentopexy fo r postpneu
the treatment of pleural empyema in children: a systematic monectomy bronchopleural fistula. Ann Thorac Surg.
review of randomized controlled trials. Arch Pediatr Adolesc 1 9 89;47:3 1 2-3 1 3 .
Med. 20 1 0; 1 64:20 1 -203. 255. Shirakusa T, Ueda H, Takata S, e t al. Use o f pedicled
23 7. Krenke K, Peradzynska J, Lange J, et al. Local treatment of omental flap in treatment of empyema. Ann Thorac Surg.
empyema in children: a systematic review of randomized l 990;50 :420-424.
controlled trials. Acta Paediatr. 2 0 1 0;99: 1 449- 1 4 5 3 . 256. Aguilar MM, Battistella FD, Owings JT, et al. Post
238. Fraser R S , Muller N L , Colman N, et a l . Diagnosis of Dis traumatic empyema. Risk factor analysis. Arch Surg.
eases of the Chest, 4th ed. Philadelphia, PA: WB Saunders; 1 997; 1 32:647-650.
2000:2659-2695. 257. Manda! AK, Tbadepalli H, Manda! AK, et al. Posmaumatic
239. Gharagozloo F, Trachiotis G, Wolfe A, et al. Pleural space empyema thoracis: a 24-year experience at a major trauma
irrigation and modified Clagett procedure for the treatment center. ] Trauma. 1 997;43 :764-77 1 .
Tu b e rc u l o u s P l e u ra l E ff u s i o n s
The diagnosis of tuberculous pleuritis should be with the diseased pleura in 12 of 1 5 patients with
considered in any patient with an exudative pleural tuberculous pleuritis. The remaining three patients
effusion. A pleural effusion as an isolated manifesta in this series were found to have parenchymal TB,
tion of tuberculosis (TB) has been likened to a pri although these patients did not have caseous foci
mary chancre as a manifestation of syphilis. Both are adjacent to the pleura.
self-limited and of little immediate concern, but both It appears that delayed hypersensitiviry plays a
may lead to serious disease many years later. large role in the pathogenesis of tuberculous pleural
effusion. As mentioned in the previous paragraph, the
hypersensitiviry reaction is initiated when tubercu
PATH O G E N E S I S A N D
lous protein gains access to the pleural space. When
PAT H O P H Y S I O LO G I C F E AT U R E S
guinea pigs are immunized to tuberculous protein by
When a tuberculous pleural effusion occurs i n the injecting Freund's adjuvant containing dead tubercle
absence of radiologically apparent TB, it may be the bacilli into their footpads, an intrapleural injection
sequel to a primary infection 6 to 1 2 weeks previously of purified protein derivative (PPD) of tuberculin
or it may represent reactivation TB ( 1 ) . In industri 3 to 5 weeks later causes the rapid appearance (-within
alized countries, more pleural effusions may be due 1 2 to 48 hours) of an exudative pleural effusion (7) .
to reactivation than are due to postprimary infec The development of the pleural effusion is suppressed
tion ( 1 ) . However, in a relatively recent study from when the animals are given antilymphocyte serum (8) .
San Francisco, pleural TB cases were approximately The neutrophil appears to play a key role in the
two times more likely to be clustered (as assessed by development of experimental tuberculous pleuritis.
genoryping of the mycobacterial organisms) than When bacillus Calmette-Guerin (BCG)-sensitized
were pulmonary TB and three times more likely to rabbits are given BCG intrapleurally, the resulting
be clustered than nonrespiratory TB cases (2) . Thirry pleural fluid contains predominantly neutrophils
five percent of the pleural TB cases were clustered (2) . for the first 24 hours (9) . If the animals are made
These findings suggest that at least in San Francisco, neutropenic, the accumulation of pleural fluid and
primary infection accounts for a large percentage inflammatory cells, particularly macrophages, is
of pleural TB (2) . However, two subsequent studies decreased. The intrapleural injection of neutrophils
from Houston (3) and Baltimore (4) were unable to in the neutropenic animals restores the response to
confirm these findings. control levels. The neutrophils in the pleural space
The tuberculous pleural effusion is thought to appear to secrete a monocyte chemotaxin that
result from rupture of a subpleural caseous focus in recruits monocytes to the pleural space and thereby
the lung into the pleural space ( 5 ) . Supporting evi contributes to the formation of granulomas (9) .
dence comes from the operative findings of Stead et In this BCG model of experimental tuberculous
al. (6) , who reported that they could demonstrate pleuritis, macrophages predominate in the pleural
a caseous tuberculous focus in the lung contiguous fluid from day 2 to day 5 (9) . It has been shown rhat
247
248 PLE U RAL D I S EASES
mesothelial cells stimulated with BCG or interferon response ( 1 9) . Compared with peripheral blood, pleu
gamma produce macrophage inflammatory protein and ral fluid is enriched with T lymphocytes. The CD4
monocyte chemotactic peptide ( 1 0) . These two pro (helper-inducer) to CDS (suppressor/cytotoxic) ratio
teins account for more than 75% of the mononuclear is 3:4 in pleural fluid, compared with 1 :7 in blood
chemotactic factor in tuberculous pleural fluid (1 O) . ( 1 9) . Pleural fluid lymphocytes from patients with
After this period, lymphocytes are the predominant cells tuberculous pleuritis show greater responsiveness to
in the pleural fluid ( 1 1 ) . When the lymphocytes first PPD than do peripheral blood lymphocytes (20) .
appear in the pleural fluid approximately on day 3, they The obvious explanation for the development of
do not respond to PPD. From day 5 onward, however, the tuberculous pleural effusion is that the delayed
reactivity to PPD is found in most cases ( 1 2) . The hypersensitivity reaction increases the permeability
reactivity of the lymphocytes in the peripheral blood of the pleural capillaries to protein, and the increased
parallels that of the pleural lymphocytes ( 1 2) . (See protein levels in the pleural fluid result in a much
Chapter 4 for further discussion of these experimental higher rate of pleural fluid formation and accordingly
models of tuberculous pleuritis.) result in the accumulation of pleural fluid. However,
It is probable that delayed hypersensitivity also this does not appear to be the mechanism for the pleu
plays a large role in the development of tuberculous ral fluid accumulation. Apicella and Allen (2 1 ) were
pleural effusions in humans. The mycobacterial cul unable to demonstrate any striking increase in the
tures of the pleural fluid from most patients with inflow of protein into the pleural space in their experi
tuberculous pleural effusions are negative (2, 1 3, 1 4) . mental model of delayed hypersensitivity tuberculous
T lymphocytes specifically sensitized to tuberculous pleuritis. They did, however, demonstrate a dramatic
protein are present in the pleural fluid ( 1 5) . In one decrease in the clearance of protein from the pleural
report, approximately 1 in 2,000 of the lympho space (2 1 ) . Leckie and Tothill (22) reported that the
cytes in the pleural fluid was specifically sensitized to pleural lymphatic flow from patients with TB was
tuberculous protein ( 1 5) . In the same report, only 1 approximately 50% that of patients with congestive
in 1 5 ,000 of the lymphocytes in the peripheral blood heart failure. It is probable that the intense inflam
was specifically sensitized to the tuberculous protein. matory reaction in the parietal pleura impedes the
It is unknown whether the increased percentage of lymphatic drainage from the pleural space (see Chap
specifically sensitized lymphocytes in the pleural fluid ter 2) and leads to the accumulation of pleural fluid.
is due to their clonal expansion in the pleural fluid or It should be noted, however, that when mesothelial
is due to the migration of PPD-responding T lym cells are cultured in the presence of BCG, vascular
phocytes from the blood to the pleural space. When endothelial growth factor (VEGF) is released from
pleural lymphocytes from patients with tuberculous the mesothelial cells, and the expression of adherens
pleural effusions are cocultured with PPD, lympho junction protein is down regulated (23). However,
kines are produced ( 1 6) . The level of lymphokine extrapolation of these results in monolayers to the in
production is much greater with pleural lymphocytes vivo situation remains to be verified.
than with peripheral blood lymphocytes ( 1 6) .
Although delayed hypersensitivity to tuberculous
INCIDENCE
protein is probably responsible for most clinical man
ifestations of tuberculous pleuritis, many patients In many areas of the world, TB remains the most
when first evaluated have a negative PPD skin test. common cause of pleural effusions in the absence of
The explanation for this paradox may be a combina demonstrable pulmonary disease. For example, in one
tion of two factors. First, in some ( 1 7) , but not in all series of 642 pleural effusions from northern Spain in
( 1 8) patients with tuberculous pleuritis, a circulating the mid- 1 990s, TB was the most common etiology
mononuclear adherent cell suppresses the specifically of pleural effusion, accounting for 25% of all pleural
sensitized circulating T lymphocytes in the peripheral effusions (24) . A study from Saudi Arabia about the
blood. Second, there may be sequestration of PPD same time demonstrated that TB was also the most
reactive T lymphocytes in the pleural space involving common cause of pleural effusions in that country,
both Leu-2 (suppressor/cytotoxic) and Leu-3 (helper) accounting for 37% of all pleural effusions (2 5 ) .
positive T cells ( 1 8) . I n the United States, the annual incidence o f tuber
Tuberculous pleural effusions are enriched with culous pleuritis has been estimated to be approximately
many potentially immunoreactive cells and substances 1 ,000 cases, and it is said that 3% to 5% of patients
that comprise the vigorous local cell-mediated immune with TB will have tuberculous pleuritis (3,26,27) .
C H A PT E R 1 3 / TU B E R C U LO U S P L E U RAL E F F U S I O N S 249
It is likely that both these numbers are low. Patients had initial symptoms of less than 1 week in dura
with tuberculous pleuritis tend to be underreported tion, whereas 50 (70%) had been symptomatic for
because their mycobacterial cultures are frequently less than a month (36) . In another series, 3 1 of 49
negative. Between 1 98 8 and 1 994, there were 2,8 1 7 patients (63%) had an acute illness that most com
cases o f T B i n patients without the acquired immu monly mimicked acute bacterial pneumonia (5).
nodeficiency syndrome (AIDS) who were reported Most patients (-70%) have a cough, usually nonpro
to the South Carolina state TB registry; 6% of these ductive, and most (-75%) have chest pain, usually
patients had pleural effusions (28 ) . However, in non pleuritic in nature ( 1 , 5 ,37) . If both cough and pleu
AIDS patients with new-onset intrathoracic TB, pleu ritic chest pain are present, the pain usually precedes
ral effusions occur in more than 25% of patients in the cough. Most patients are febrile, but a normal
Burundi (29) and 20% in South Africa (30) . temperature does not rule out the diagnosis. In one
Patients who are immunocompromised are more series, 7 of 49 patients ( 1 4%) were afebrile (5). Occa
likely to develop TB than nonimmunocompromised sionally, the onset of TB is less acute, with only mild
individuals. Pleural TB also occurs frequently in the chest pain, perhaps with a low-grade fever and a non
immunocompromised individual. Mycobacterial productive cough, weight loss, and easy fatigability.
infection occurred in 27 of 1 ,26 1 patients (2. 1 %) who In general, patients with tuberculous pleuritis are
received renal transplants in Valencia, Spain and 3 of younger than patients with parenchymal TB . In one
these had pleural effusions (3 1 ) . TB occurred in 48 of recent series from Qatar, the mean age of 1 00 patients
330 patients ( 1 4. 5 %) who were on renal dialysis in with tuberculous pleuritis was 3 1 . 5 years (3 8). In
Saudi Arabia and 5 of them had pleural effusions (32) . industrialized countries, the mean age of patients
One might anticipate that the incidence of tuber with TB tends to be older. In a recent study from the
culous pleuritis would be relatively low in patients United States, the mean age of the 1 4,000 patients
with AIDS and TB because the patient with AIDS with tuberculous pleuritis reported to the Commu
has a compromised immunologic system, and pleural nicable Disease Center in the United States between
TB is thought to be due to hypersensitivity. However, 1 993 and 2003 was 49.9 years (27) . Patients with
overall it appears that the incidence of pleural effu pleural effusions secondary to reactivation tend to be
sions is higher in patients with AIDS. One possible older than those with postprimary pleural effusion ( 1 ) .
explanation for this apparent paradox is that the pleu Pleural effusions secondary t o tuberculous pleuri
ral effusion in patients with AIDS is related to pleural tis are usually unilateral and can be of any size. In one
invasion by the mycobacteria rather than to delayed series, the effusions occupied more than two thirds
hypersensitivity (33) . The fact that smears and cul of the hemithorax in 1 8%, between one third and two
tures are more often positive in the human immuno thirds of the hemithorax in 47%, and less than one
deficiency virus (HIV)-positive patient lends support third of the hemithorax in 34% (39) . In another series
to this hypothesis. of 46 patients with massive pleural effusions (40) , 4%
Although in the series referenced in the preced of the effusions were due to TB . In approximately 20
ing text from Burundi, a slightly smaller percentage to 25% of patients with pleural effusions secondary to
of HIV-positive patients (24%) than HIV-negative TB (39,4 1 ) , coexisting parenchymal disease is visible
patients (28%) had pleural effusions (29) , other series on the chest radiograph. If chest CT scans are done,
have shown that pleural effusions are more common approximately 90% will have parenchymal abnor
in HIV-positive patients. The percentage of patients malities (4 1 ,42) . In such patients, the pleural effusion
with thoracic TB who also had a pleural effusion was is almost always on the side of the parenchymal infil
higher in HIV-positive patients than in HIV-negative trate and invariably indicates active parenchymal dis
patients in series from South Africa (38% vs. 20%) ease. On rare occasions, pleural TB can present with
(29), Uganda (23% vs. 1 1 %) (34) , and Zimbabwe pleural-based nodules and thickening (43 ) .
(27% vs. 1 3%) (3 5 ) .
C l i n ical M a n ifestations i n
H IV-Positive Patie n ts
C LI N I CA L M A N I F E STATI O N S
The clinical manifestations of pleural TB tend to
Although TB is usually considered a chronic illness, be somewhat different in the HIV-positive patient.
tuberculous pleuritis most commonly manifests as an Patients with HIV tend to have a longer duration
acute illness. In one series of 71 patients, 25 (35%) of illness and a lower incidence of chest pain (44) .
250 PLE U RAL D I S EASES
Systemic signs and symptoms such as night sweats, of tuberculous pleuritis and initiate proper treatment.
fatigue, diarrhea, hepatomegaly, splenomegaly, and Moreover, patients in whom the diagnosis cannot be
lymphadenopathy are significantly more common established but is considered likely should also be
in HIV-infected patients (37) . Patients with HIV are treated.
more likely to have concomitant parenchymal lesions
(3) . Their pleural fluid is more likely to be smear
positive for acid-fast bacilli (AFB) and culture posi D IAG N O S I S
tive for AFB (44,45 ) . If the CD4 count is less than Th e diagnosis o f tuberculous pleuritis depends on
1 00, approximately 50% have a positive smear for the demonstration of tubercle bacilli in the sputum,
AFB on their pleural fluid (44) . HIV patients have pleural fluid, or pleural biopsy specimen, or the dem
significantly lower lymphocyte counts (45 ) . Interest onstration of granulomas in the pleura. The diagnosis
ingly, the viral load per mL pleural fluid was higher can also be established with reasonable certainty by
than that in simultaneously obtained serum in each demonstrating elevated levels of adenosine deaminase
of eight patients in one study (46) . (ADA) or interferon-gamma in the pleural fluid (50) .
Study of the peripheral blood is not useful; most
patients do not have leukocytosis (5) . The chest radio
N AT U RA L H I STORY O F U NTR EATE D graph usually demonstrates only the pleural fluid, but
TU B E RC U LO U S P L E U R I T I S as previously mentioned, approximately 20 to 25%
of the patients also have a parenchymal infiltrate due
Without treatment, tuberculous pleuritis usually
to TB (39) .
resolves spontaneously, only to return as active TB at a
later date. Patiala (47) followed up for at least 7 years
all 2,8 1 6 members of the Finnish Armed Forces who Tu bercu l i n S k i n Testi ng
developed pleural effusions between 1 939 and 1 94 5 . In the past, the tuberculin skin test was an important
They reported that 43% o f this large group of young diagnostic aid in patients suspected of having tuber
men developed TB during the follow-up period. Even culous pleuritis. However, a negative skin test does
in the I -year observation period 5 years following the not rule out the diagnosis of tuberculous pleuritis.
initial episode, 5% of the total population studied In one series from Spain, the PPD was positive in
developed active TB. only 66.5% of254 patients with tuberculous pleuritis
Confirmatory evidence for this large series comes (39) . In another series from Hong Kong, more than
from the series of Roper and Waring (48) in the one half of the patients tested had a negative PPD
United States, who followed up 1 4 1 military person ( 1 4) . The factors responsible for the negative skin test
nel first seen from 1 940 to 1 944 with a pleural effu in patients with tuberculous pleuritis are discussed
sion and a positive PPD test. In most patients, the earlier in this chapter. If a patient with a negative
effusions resolved and all the other symptoms disap tuberculin skin test and tuberculous pleuritis is skin
peared within 2 to 4 months. Nevertheless, 92 of the tested more than 8 weeks after the development of
1 4 1 individuals (65%) subsequently developed some symptoms, the skin test will almost always be posi
form of active TB. Manifest TB did not develop in tive. Therefore, in patients with an undiagnosed exu
the lung or elsewhere in any of the patients within dative pleural effusion, a negative tuberculin skin test
8 months of the onset of the original pleurisy. The performed 8 weeks after the development of symp
incidence of subsequent TB was 60% in those with toms can be used to exclude the diagnosis of tuber
initially negative pleural fluid cultures for TB and culous pleuritis. However, if the patient is markedly
6 5% in those with initially positive pleural fluid cul immunosuppressed with HIV infection or is severely
tures. In addition, the size of the original effusions malnourished, the PPD may remain negative.
and the presence or the absence of small radiologic
residual pleural disease were not correlated with the
P l e u ra l F l u i d A n a lysis
subsequent appearance of active TB (48) . The forego
ing series emphasize the importance of making the Pleural fluid analysis is useful in the diagnosis of
diagnosis of tuberculous pleuritis. tuberculous pleuritis. The fluid is invariably an
Because the administration of antituberculous exudate. Frequently, the pleural fluid protein level
chemotherapy reduces the incidence of subsequent is above 5 g/dL, and this finding suggests tubercu
TB (5,49), it is important to establish the diagnosis lous pleuritis. In most patients, the pleural fluid
C H A PT E R 1 3 / TU B E R C U LO U S P L E U RAL E F F U S I O N S 25 1
not have pleural fluid lymphocytosis. Indeed, if the pleuritis, the separation of ADA into its isoenzymes
diagnostic criteria for tuberculous pleuritis includes is not necessary in the vast majority of cases (73,74) .
a pleural fluid lymphocyte-to-neutrophil ratio of When preservatives are added to pleural fluid,
0.75 or more, the specificity of the test is increased the pleural fluid ADA levels remain stable during
(64-66) . This increase in specificity is due to exclud transportation. Miller et al. (75) have shown that
ing the cases with rheumatoid pleuritis or empy if 0.9 mL pleural fluid is maintained in a test tube
ema. The pleural fluid ADA is elevated in patients containing 0 . 1 0 mL of a mixture of 50% glycerol
with tuberculous pleuritis who have predominantly and 50 % ethylene glycol, the pleural fluid can be
PMNs in their pleural fluid (54) . mailed by regular mail with no loss of ADA activity.
There are a few other instances in which the pleural The pleural fluid ADA levels were virtually identical
fluid ADA level will be elevated. Patients with pleural in specimens that were shipped in the preservative
effusions due to Q fever (67) or brucellosis (68) will through regular mail and those that were shipped in
have elevated pleural fluid ADA levels. An occasional dry ice by air (75) . However, if pleural fluid is main
patient with a pleural effusion due to lymphoma or tained at ambient temperatures without preserva
malignancy will have a lymphocytic effusion with a tives, the ADA levels will linearly decrease (75) . The
high ADA (69) . ADA levels remain stable for long periods in pleural
Almost all patients with tuberculous pleuritis have fluid frozen at - 70°C (69) .
an ADA level above 40 U/L. Immunosuppressed In summary, measurement of the pleural fluid
patients with tuberculous pleuritis have elevated pleu ADA level is useful in the diagnosis of tuberculous
ral fluid ADA levels. The levels of ADA in patients pleurisy. In patients with lymphocytic pleural effu
with and without AIDS are comparable (62) . Renal sion, demonstration of an ADA level above 40 U/L
transplant patients who develop a pleural effusion is strongly suggestive of the diagnosis of tubercu
have elevated pleural fluid ADA levels (70) . lous pleurisy. If a lymphocytic pleural effusion is
The pleural fluid ADA level can be used to exclude not due to TB, the pleural fluid ADA level is almost
the diagnosis of tuberculous pleural effusions in always below 40 U/L. Pleural fluid for ADA assay
patients with undiagnosed pleural effusions. Ferrer can be transported at ambient room temperatures if
et al. (7 1 ) followed up 40 patients with undiagnosed preservatives are added to the pleural fluid (75 ) .
pleural effusions and a pleural fluid ADA level below
43 U/L for a mean of 5 years and reported that none
I nterfe ron-Ga m m a
of the patients developed TB. Lymphocytic pleural
effusions not due to tuberculous pleuritis usually have Another test that i s useful i n the diagnosis o f tuber
pleural fluid ADA levels below 40 U/L. Lee et al. (69) culous pleuritis is the level of interferon-gamma in
measured the pleural fluid ADA in 1 06 patients with the pleural fluid (59,76-80) . Interferon-gamma is
lymphocytic pleural effusions not due to TB, includ produced by the CD4• lymphocytes from patients
ing 45 post-coronary artery bypass pleural effusions with tuberculous pleuritis (78) . Patients with tuber
and 26 malignant pleural effusions. They reported culous pleuritis tend to have higher pleural fluid
that only 3 of the 1 06 fluids (3%) had ADA levels interferon-gamma levels than do patients with pleu
above 40 U/L (69) . The three patients included rwo ral effusions of other etiologies. In the largest series
with lymphoma and one with a late complicated published until now, Villena et al. (8 1 ) measured the
parapneumonic effusion (69) . Jimenez Castro et al. pleural fluid interferon-gamma levels in 5 9 5 patients,
(72) measured the pleural fluid ADA levels in 4 1 0 including 82 with TB, and reported that a cutoff
lymphocytic nontuberculous pleural fluids and found level of 3 .7 IU/mL yielded a sensitivity of 0 .98 and a
that the ADA was above 40 U/L in only 7 ( 1 .7%) . specificity of 0 . 9 8 . Most of the false positives in this
There are rwo molecular forms of ADA-ADAl study occurred in patients with hematologic malig
and ADA2 . ADA l is found in all cells, but has its nancies. Unfortunately, ADA levels were not mea
greatest activity in lymphocytes and monocytes (73) . sured in this article for comparison (8 1 ) . The pleural
ADA2 is found only in monocytes, and most of the fluid interferon-gamma levels were elevated in 1 4
ADA in tuberculous pleural fluid is ADA2, whereas immunocompromised patients and 3 transplantation
most of the ADA in other pleural fluids is AD Al (73 ) . patients with tuberculous pleuritis (8 1 ) . Comparable
Although the use of a ratio of the ADAl t o ADA total results have been reported in other series (76-79) , but
of less than 0.42 will slightly increase the sensitivity comparison of the series is difficult because the units
and specificity of the ADA in diagnosing tuberculous have differed from one study to another. Patients with
C H A PT E R 1 3 / TU B E R C U LO U S P L E U RAL E F F U S I O N S 253
empyema frequently have elevated pleural fluid levels Polym e rase C h a i n Reaction
of interferon-gamma (59) . Jiang et al. (80) performed
In the field of infectious diseases, the polymerase
a meta-analysis on the diagnostic usefulness of pleural
chain reaction (PCR) tests and other nucleic acid
fluid interferon-gamma levels in 22 studies with 782
amplification (NAA)-based tests have been quite
patient with tuberculous pleuritis and 1 , 3 1 9 patients
useful in establishing the diagnosis of viral diseases.
with pleural effusions due to other diseases. They
Therefore, it was hypothesized that the NAA-based
reported that the mean sensitivity was 0 . 8 9 , the mean
tests would be useful in diagnosing tuberculous
specificity was 0. 97, the mean positive likelihood
pleuritis. It appears, however, that the NAA-based
ratio was 23.45, and the mean negative likelihood
tests are certainly not superior to either the pleural
ratio was 0 . 1 1 (80) .
fluid ADA or interferon-gamma levels in establish
ing the diagnosis of tuberculous pleuritis. Pai et al.
Ad e n os i n e Dea m i nase vers u s (87) performed a meta-analysis on 1 4 studies utiliz
I nte rfe ron-Ga m m a ing NAA-based tests for the diagnosis of tuberculous
pleuritis using three different commercial kits includ
Which i s the better test fo r the diagnosis o f tubercu
ing 1 27 patients with tuberculous pleuritis and 1 ,400
lous pleuritis-ADA or interferon-gamma? Greco et
patients with pleural effusions due to other diseases.
al. (82) performed a meta-analysis comparing ADA
They reported that the mean sensitivity was only 0.62
and interferon-gamma for the diagnosis of tubercu
while the mean specificity was 0 .98 (87) . The mean
lous pleuritis. After a review of 3 1 papers on ADA
positive likelihood ratio was 25.4 while the mean
which included 4,738 patients and 1 3 papers on
negative likelihood ratio was 0.40 (87) . Currently,
interferon-gamma which included 1, 1 89 patients,
PCR of the pleural fluid should be considered to be
they reported that the summary receiver operating
an investigative test until there is more data available
characteristic curve yielded a maximum joint sensi
regarding its sensitivity and specificity.
tivity and specificity of 93% for ADA and 96% for
interferon-gamma (82) . They concluded that both
ADA and interferon-gamma are reasonably accurate P l e u ra l F l u i d Tu b e rc u l o u s
at diagnosing tuberculous pleuritis. Gupta et al. (83) Prote i n s o r Anti bod ies
have pointed out that in articles that directly compare In recent years, the possibility of establishing the
ADA and interferon-gamma, some conclude that diagnosis of tuberculous pleuritis by the demonstra
ADA is better and some conclude that interferon tion of tuberculous antigens or specific antibodies
gamma is better. In general, measurement of the against tuberculous proteins in the pleural fluid has
pleural fluid ADA level is preferred to measurement been investigated. None of these tests have proved
of the pleural fluid interferon-gamma level in evaluat to be useful in the diagnosis of tuberculous pleu
ing patients suspected of having pleural TB because ritis . The presence of antituberculous antibodies in
the ADA measurement is less expensive (84) . the pleural space appears to result from their passive
diffusion from the serum rather than local antibody
I nte rfe ron-Ga m m a Release Assays {IG RA) production (88) . Therefore, it is unlikely that mea
surement of such antibodies in the pleural fluid will
The IGRAs are inferior to the pleural fluid inter ever be diagnostically useful.
feron-gamma levels in diagnosis tuberculous pleuritis
(85,86). Zhou et al. (85) performed a meta-analysis of
Other Chem ical Tests
the available reports on IGRA on the pleural fluid for
diagnosing tuberculous pleuritis. The analyzed seven Other chemical analyses of the pleural fluid are of
reports with a total of 2 1 3 patients with tuberculous limited value in establishing the diagnosis of tuber
pleuritis and 1 5 3 patients with pleural effusions of culous pleuritis. Although in the past it was believed
other etiologies and reported that the mean sensi that the pleural fluid glucose level was reduced in
tivity was 0.75, the mean specificity was 0.82, the most cases of tuberculous pleuritis (89), more recent
mean positive likelihood ratio was 3 .49 and the mean studies show that most patients with tuberculous
negative predictive ratio was 0.24 ( 8 5 ) . Since these pleuritis have a pleural fluid glucose level above
results are markedly inferior to those with ADA or 60 mg/dL (5 ,44) . A low pleural fluid pH level was
interferon-gamma, the IGRA should not be used in once thought to be suggestive of tuberculous pleu
assessing whether patients have tuberculous pleuritis. ritis (90) , and I concluded that tuberculous pleural
2 54 PLE U RAL D I S EASES
effusions had a lower pleural fluid pH level than smear and culture of the sputum in 84 patients with
malignant pleural effusions in the first article Light et tuberculous pleuritis. Sputum was induced in those
al. wrote on pleural fluid pH levels (9 1 ) . Subsequent unable to spontaneously produce sputum. They
articles (92,93), however, and my own observations reported that the sputum studies were positive in
indicate that the pleural fluid pH has approximately 44 of the 84 patients (52%) (99 ) . In 10 of the 44
the same distribution in malignant as in tuberculous patients, sputum smears were positive, whereas cul
pleural effusions. The mean level of C-reactive pro tures were positive in all 44 (99) . Sputum was induced
tein (CRP) is higher in tuberculous effusions than in 64 patients with a normal radiograph (except for
in other exudative effusions (94) . Chierakul et al. the effusion) and was positive in 35 ( 5 5 % ) .
(94) reported that the optimal cutoff level for the
pleural fluid CRP level was 30 mg/dL and this level
P l e u ra l F l u i d Sta i n s a n d C u l t u res
had a sensitivity of 72% and a specificity of 93%.
Garcia-Pachon et al. (95) in a study of 1 44 patients For nonimmunosuppressed patients, routine smears
reported that a CRP level greater than 50 mg/L had of the pleural fluid for mycobacteria are not indicated
a 9 5 % specificity for TB, whereas a CRP level less because they are usually negative, unless the patient
than 30 mg/L had a 9 5 % sensitivity for excluding has a tuberculous empyema ( 1 3,39 ) . Cultures for
tuberculous pleuritis. However, both the pleural mycobacteria should be obtained, however. If the
fluid ADA and interferon gamma are superior to the patient is HIV-positive, the smears may be positive in
pleural fluid CRP levels in establishing the diagnosis more than 20% (44) . In most series of patients with
of tuberculous pleuritis. tuberculous pleuritis, the pleural fluid cultures are
The levels of lysozyme in the pleural fluid have positive for mycobacteria in fewer than 40% (5,39) .
been proposed to be useful diagnostically, and there For mycobacterial cultures, use of a BACTEC sys
is no doubt that the mean level of lysozyme in the tem with bedside inoculation provides higher yields
pleural fluid from patients with tuberculous pleuritis and faster results than do conventional methods. In
is higher than it is in other exudative pleural fluids one study, the median time for the BACTEC cultures
(96,97) . A value of 1 .2 : 1 for the ratio of the pleural to become positive was 1 8 days (range 3-40 days) ,
fluid to the serum lysozyme has been proposed as a whereas the median time for conventional cultures
good test for diagnosing tuberculous pleuritis (97) . was 3 3 . 5 days (range 2 1 -48 days) ( 1 00 ) . In a second
When the utility of this ratio is compared with that study, the pleural fluid cultures were positive by the
of the pleural fluid ADA or interferon-gamma level, BACTEC system in 24% of HIV-negative and 75%
the lysozyme ratio is distinctly inferior (59) . For this of HIV-positive individuals, whereas the cultures
reason, the measurement of the lysozyme ratio is not were positive by the Lowenstein-Jensen medium in
recommended (59) . Measurement of the levels of 1 2% of HIV-negative patients and 56% of HIV
tumor necrosis factor alpha (TNF-a) in the pleural positive patients (45 ) . In this same study, the mean
fluid or the ratio of the pleural fluid to the serum time to positive culture with the BACTEC system
TNF-a is inferior to measurements of the pleural was 3 . 5 weeks compared with 4.7 weeks with the
fluid ADA in making the diagnosis of tuberculous Lowenstein-Jensen medium (45 ) .
pleuritis (98) . There have been many other studies
on cytokine levels in the pleural fluid and although
Pleu ra l B i o psy
the mean cytokine levels in tuberculous effusions dif
fer significantly from those in pleural fluids of other For the last 40 years, the diagnosis of tuberculous
etiologies (see Chapter 5 ) , there is always so much pleuritis has been most commonly made with needle
variation in the levels that their measurement has biopsy of the pleura. The demonstration of granuloma
not proved as diagnostically useful as that of ADA or in the parietal pleura suggests tuberculous pleuritis;
interferon-gamma. caseous necrosis and AFB need not be demonstrated.
Although other disorders including fungal diseases,
sarcoidosis, tularemia ( 1 0 1 ) , and rheumatoid pleuri
S p u t u m S m e a rs a n d C u lt u res
tis may produce granulomatous pleuritis, more than
One test that is frequently overlooked in the diagnostic 9 5 % of patients with granulomatous pleuritis have
workup of patients with a pleural effusion is examina TB . Even when granulomas are not demonstrated
tion of the sputum for mycobacteria. Conde et al. (99) in the pleural biopsy, the biopsy specimen should be
prospectively evaluated the diagnostic yield of AFB examined for AFB because organisms are occasionally
C H A PT E R 1 3 / TU B E R C U LO U S P L E U RAL E F F U S I O N S 25 5
demonstrated when no granulomas are present in the typical ofTB and, particularly, if the pleural fluid has
biopsy. In one study of 248 patients with tuberculous a high percentage of lymphocytes, the possibility of
pleuritis who underwent needle biopsy of the pleura, TB can be further evaluated with thoracoscopy or
the biopsy showed granulomas in 1 9 8 patients (80%), needle biopsy of the pleura.
the acid-fast stain of the biopsy was positive in 64
(25 . 8%) , the culture of the biopsy tissue was positive
in 1 40 (56%), and at least one of the preceding three
tests was positive in 227 (9 1 %) (39) . PCR has also
TREAT M E N T
been tried on pleural biopsy specimens, but it is not Th e treatment o f tuberculous pleuritis has three goals:
clear that the PCR adds to the regular microscopic (a) to prevent the subsequent development of active
examination ( 1 02) . TB, (b) to relieve the patient symptoms, and (c) to
In recent years, the availability of tests on the prevent the development of a fibrothorax.
pleural fluid, such as the ADA, which are at least
as sensitive in diagnosing tuberculous pleuritis as
C h e m othera py
is needle biopsy of the pleura, have resulted in a
decreased use of needle biopsy of the pleura ( 1 03 ) . The recommendations for the treatment of all pul
A possible criticism of relying on the pleural fluid monary and extrapulmonary TB are as follows
tests rather than the pleural biopsy is that with pleu ( 1 04, 1 0 5 ) . The initial phase of a 6-month regimen
ral fluid markers, no culture results are obtained. should consist of a 2-month period of isoniazid
Accordingly, the sensitivities of the organisms can (INH), rifampin, and pyrazinamide. Ethambutol
not be determined. It should be noted that although should be included in the initial regimen until the
the cultures of the biopsy are positive in approxi results of drug susceptibility studies are available,
mately 5 5 % , cultures of the fluid itself are positive unless there is little possibility of drug resistance.
in 3 5 % . Therefore, the culture of the biopsy itself The second phase of the treatment should be INH
provides additional positive cultures in only 20% and rifampin given for 4 months. Directly observed
of the patients. One study reported that multidrug therapy (DOT) is recommended. Nine-month regi
resistance occurred in only approximately 1 % of mens using INH and rifampin are also effective when
patients with pleural TB and that any first-line drug the organisms are fully susceptible to the drug. One
resistance occurs in less than 1 0% (27) . study demonstrated that the penetration of INH into
pleural fluid was excellent, but the penetration of
pyrazinamide was very poor, whereas the penetration
Reco m m e n ded D i a g n ostic A p p roach i n
of rifampin was intermediate ( 1 06) . The clinical sig
Pati e n ts with U n d iag nosed Exudative
nificance of these findings remains to be determined.
Ple u ra l Effu s i o n
The recommendations mentioned in the pre
When a patient is seen with a pleural effusion, the ceding text may be somewhat intensive for isolated
diagnosis of tuberculous pleuritis should always be tuberculous pleuritis. Less intensive regimens appear
considered. When the initial thoracentesis is per to be effective. In one study, Canete et al. ( 1 07)
formed, the pleural fluid should be analyzed for the treated 1 30 patients with 5 mg/kg of INH and 1 0
ADA level and differential cell count, and the fluid mg/kg o f rifampin daily fo r 6 months and reported
should be cultured for mycobacteria. If the pleural no treatment failures. In a second study, Dutt et al.
fluid ADA is above 70 U/L and the pleural fluid ( 1 08) administered 300 mg of INH plus 600 mg of
has a lymphocyte-to-neutrophil ratio of more than rifampin daily for 1 month, followed by 900 mg of
0.75, the diagnosis of TB is virtually established. If INH plus 600 mg of rifampin rwice a week for the
the pleural fluid ADA level is berween 40 and 70 next 5 months to 1 98 patients. This regimen failed in
U/L and the patient has a lymphocyte-to-neutrophil only one patient ( 1 08) .
ratio of more than 0.75, a presumptive diagnosis of The patient with isolated tuberculous pleuritis
TB can be made. In this situation, consideration can appears to have a small bacterial burden because many
be given to further studies (thoracoscopy or needle of the symptoms are due to delayed hypersensitivity.
biopsy of the pleura) if the patient's clinical picture In the series of Patiala and Mattila (49), the admin
is not typical for tuberculous. If the patient's pleural istration of chemotherapy decreased the subsequent
fluid ADA level is below 40, the diagnosis of TB is incidence of TB from 28% to 9%, although most
unlikely. However, if the patient has a clinical picture of their patients received only one drug for less than
256 PLE U RAL D I S EASES
6 months. Falk and Stead ( 1 09) reported that anti a fibrinolytic may decrease the degree of residual
tuberculous therapy reduced the incidence of subse pleural thickening. Kwak et al. ( 1 1 9) randomized 43
quent TB from 1 9% to 4%, and, again, many of their patients to receive only antituberculous chemother
patients did not receive two drugs for even 6 months. apy or antituberculous chemotherapy plus 1 00,000
From the foregoing studies, it appears that 6 months urokinase daily administered through a pigtail cath
of INH and rifampin administration are sufficient if eter starting when the amount of pleural Buid drain
the patient does not have resistant organisms. age was less than 1 00 mL/ day and finishing when
With treatment, the patient's symptoms and the amount of pleural Buid was less than 50 mL/
radiologic abnormalities gradually abate. The average day. Patients received a mean of 3 . 8 urokinase instil
patient becomes afebrile within 2 weeks, but temper lations ( 1 1 9) . After the cessation of treatment, the
ature elevations may persist for as long as 2 months width of the pleural thickening was 0.46 cm in the
( 1 1 0) . If a therapeutic thoracentesis is performed at urokinase group and 1 . 86 cm in the control group
the same time that antituberculous therapy is initi ( 1 1 9) . Chung et al. ( 1 20) randomized 44 patients
ated, most patients become afebrile within 5 days with loculated TB pleural effusions to receive pig
( 1 1 1 , 1 1 2 ) . The mean duration for complete resorp tail drainage of the effusion with saline or strepto
tion of the pleural Buid is approximately 6 weeks, kinase irrigations for three consecutive days. They
but it can be as long as 1 2 weeks ( 1 1 0) . No reason reported that the group that received streptokinase
exists to keep the patient in bed (5), and patients need had less residual pleural thickening and had a faster
be isolated only if their sputum tests are positive for radiological improvement ( 1 20) .
mycobacteria. An occasional patient will develop paradoxi
The incidence of pleural thickening at 6 to cal worsening of their disease after antituberculous
12 months after beginning treatment is approximately therapy is initiated as part of the immune recon
50% ( 1 1 3) . In one study ( 1 1 4) , 8 of 8 1 patients stitution inflammatory syndrome. Al-Majed ( 1 2 1 )
( 1 0%) had a forced vital capacity (FVC) less than serially studied 6 1 patients with tuberculous pleural
80% of predicted at the end of the follow-up period effusion who were started on a standard regimen of
for their TB treatment. In this study, there was only rifampin, INH, pyrazinamide, and ethambutol for
a weakly negative correlation (r = -0 .298) between the first 2 months. He reported that the size of the
the degree of pleural thickening and the reduction effusion worsened in 1 0 of the 6 1 patients ( 1 7%)
in the FVC ( 1 1 4) . The incidence of residual pleural after the initiation of antituberculous therapy. Six of
thickening is not closely related to the initial pleural the patients developed increasing dyspnea and were
Buid findings; patients with a low glucose, high lac treated with pleural aspiration and oral prednisolone
tate dehydrogenase (LDH), and high cytokine levels with complete resolution of the pleural effusion.
are only slightly more likely to have residual pleural A second report suggested that such paradoxical
thickening ( 1 1 3 , 1 1 5) . One randomized controlled responses might be due to !NH-induced lupus pleu
study of 52 patients demonstrated that the intrapleu ritis ( 1 22) . An occasional patient with tuberculous
ral administration of 2 . 5 mL of a hyaluronate-based pleuritis will also develop a peripheral lung nodule
gel resulted in significantly faster Buid absorption while being treated for the pleuritis ( 1 23 ) . The nod
and significantly less pleural thickening at 3 months ules represent pulmonary TB and disappear when the
(0 . 5 7 vs. 1 . 1 4 cm) ( 1 1 6) . Residual pleural thicken antituberculous therapy is continued ( 1 23 ) .
ing is more common if the pleural effusion is initially A n occasional patient will also develop a pleu
loculated ( 1 1 7) . Residual pleural thickening will ral effusion during antituberculous chemotherapy.
continue to improve for at least 1 8 months after the Gupta et al. ( 1 24) reported 29 patients who devel
anti tuberculous therapy is completed ( 1 1 7) . oped pleural effusions while receiving chemotherapy
It appears that complete removal o f the pleural for pulmonary ( 1 6 cases) or extrapulmonary TB . The
Buid does not decrease the amount of residual pleural pleural effusion developed between the 5th and 8th
thickening. Lai et al. ( 1 1 8) randomized 6 1 patients week of starting chemotherapy in 1 3, between the
to receive pigtail drainage until the drainage was less 9th and 1 2th week in 9, and between the 1 3th and
than 50 mL/day or no drainage. The degree of resid 24th week in 5 ( 1 24) . The pleural Buid was exudative
ual pleural thickening was basically identical in both in all cases and the culture was positive for Mycobac
groups ( 1 1 8) . terium tuberculosis in four ( 1 24) . Most patients had a
I n patients with loculated tuberculous pleu good response to the same chemotherapeutic regimen
ral effusions, the intrapleural administration of without any interruption.
C H A PT E R 1 3 / TU B E R C U LO U S P L E U RAL E F F U S I O N S 257
Definitive surgical treatment consists of decor maximal tolerated dosages, because these patients
tication, which sometimes must be combined with have a strong tendency to develop resistant organ
thoracoplasty because the underlying lung has usually isms. This is probably because the antituberculous
been damaged by the TB to such an extent that it can drugs frequently do not reach their normal levels
not expand to fill the pleural space ( 1 29 , 1 3 1 ) . This in the pleural space owing to the thick, fibrous, and
is a major operation and is dangerous to the patient often calcified pleura ( 1 36) .
with severely damaged lungs. In Jensen's series ( 1 30)
of 15 patients with tuberculous bronchopleural fistu
ATYP I CA L M YCO BACT E R I A
las, 3 were cured with conservative treatment, 2 were
deemed unfit to undergo definitive surgical treatment Pleural effusions due t o atypical mycobacteria are
and died within a year, and 1 0 were operated on with rare. Pleural effusions without parenchymal disease
an operative mortality rate of 20%. analogous to the post primary pleural effusion with
Mycobacterium tuberculosis do not occur. However,
approximately 5% of patients with parenchymal dis
TU B E RC U LO U S E M PY E M A
ease due to either M. kansasii or M. intracellulare have
Tuberculous empyema i s a rare entity characterized an associated small pleural effusion, an incidence
by purulent pleural fluid that is loaded with tuber similar to that seen with parenchymal disease due to
culous organisms on AFB stains ( 1 32) . It usually M. tuberculosis ( 1 37) . Approximately 1 5 % of patients
develops in fibrous scar tissue resulting from pleu with parenchymal disease due to M. intracellulare or
risy, artificial pneumothorax, or thoracoplasty ( 1 33) . M. avium (collectively referred to as the Myobacte
In one series of 1 2 patients from Italy, 9 patients rium avium complex [MAC] ) have marked pleural
had received artificial pneumothorax therapy, 1 had thickening ( > 2 cm) , as compared with fewer than
received a thoracoplasty, and 2 had received inad 3% of patients with disease due to M. tuberculosis or
equate antituberculous therapy ( 1 3 3 ) . The mean M. kansasii ( 1 3 7) .
duration between the therapy with the artificial If the cultures of the pleural fluid yield a non
pneumothoraces and the development of the empy tuberculous mycobacterium, one must be cautious
ema was more than 40 years ( 1 3 3 ) . Frequently, the in attributing the pleural effusion to that organism .
underlying pleura is heavily calcified. The patient Gribetz et al. ( 1 3 8 ) reviewed the case records of 22
usually has a subacute or chronic illness character patients whose pleural fluid grew nontuberculous
ized by fatigue, low-grade fever, and weight loss. On mycobacterium. In 1 6 of the patients, there was
rare occasions, a tuberculous empyema may produce another explanation for the pleural effusion, and
an empyema necessitatis where the empyema rup in only 3 did the nontuberculous mycobacterium
tures through the chest wall ( 1 34) . Indeed, TB is the appear to be responsible for the pleural effusion.
most common cause of empyema necessitatis ( 1 34) . All three patients had nontuberculous mycobacte
Radiographically, there may be an obvious pleural rial infection of other tissues. These authors con
effusion, but frequently the chest radiograph only cluded that nontuberculous mycobacteria isolated
shows pleural thickening. The chest CT scan usu from pleural fluid should not be considered etio
ally demonstrates a thick, calcific pleural rind and rib logic, unless there is evidence of the same organism
thickening surrounding loculated pleural fluid. The infecting other tissues ( 1 3 8 ) . However, in another
diagnosis is established with diagnostic thoracente study ( 1 39) for a single center in Taiwan, pleural
sis, which yields thick pus on which the AFB smear fluid cultures from 3 5 patients grew atypical myco
is markedly positive. bacterium over a 7-year period including 1 6 MAC,
Treatment is difficult, and decortication, extra 6 M. fortuitum, 3 M. kansasii, 3 M. chenonae, and
pleural pneumonectomy, and thoracoplasty have all 7 others. Before the AIDS epidemic, disseminated
been recommended. All of these procedures have nontuberculous mycobacterial infections were very
substantial morbidity and some mortality, at least uncommon. However, disseminated disease due
in part, because of the compromised pulmonary sta MAC is an important cause of infection in patients
tus of the patient. Because intensive chemotherapy with AIDS ( 1 40 ) . Some autopsy studies have shown
coupled with serial thoracentesis can be curative that more than 50 % of patients with AIDS who die
at times ( 1 3 5 ) , this approach should be attempted have disseminated disease due to MAC ( 1 40 ) . Pleu
initially. It is important to use a multiple (three or ral effusions do occur in some patients with dissem
more) drug regimen and to employ agents at their inated disease due to MAC ( 1 40 ) , and pleural fluid
C H A PT E R 1 3 / TU B E R C U LO U S P L E U RAL E F F U S I O N S 259
cultures are sometimes posmve for MAC. Never 1 7. Ellner JJ . Pleural fluid and peripheral blood lymphocyte func
theless, it is unclear whether the atypical mycobac tion in tuberculosis. Ann Intern Med. 1 978;89:932-933.
18. Rossi GA, Balbi B, Manca F. Tuberculous pleural effusions:
teria are responsible for the effusion ( 1 40) . Overall, evidence for selective presence of PPD-specific T-lymphocyres
disease due to MAC accounts for at most only a at site of inflammation in the early phase of the infection. Am
small percentage of the pleural effusions in patients Rev Respir Dis. 1 987; 1 36: 575-579.
with AID S . 19. Ellner JJ, Barnes PF, Wallis RS, et al. The immunology of
tuberculous pleurisy. Semin Respir Infect. 1 9 88;3:335-342.
20. Mehra V, Gong JH, Iyer D, et al. Immune response to recom
binant mycobacterial proteins in patients with tuberculosis
infection and disease. ] Infect Dis. 1 996; 1 74:43 1 -434.
REFERENCES 2 1 . Apicella MA, Allen JC. A physiologic differentiation berween
delayed and immediate hypersensitivity. ] Clin Invest.
I. Moudgil H, Sridhar G, Leitch AG . Reactivation disease: 1 969;48:250-2 5 9 .
the commonest form of tuberculous pleural effusion in 2 2 . Leckie WJH, Tothill P. Albumin turnover in pleural effusions.
Edinburgh, 1 9 80- 1 99 1 . Respir Med. 1 994;88:30 1 -304. Clin Sci. 1 965;29:339-352.
2 . Ong A, Creasman J, Hopewell PC, et al. A molecular epide 23. Mohammed KA, Nasreen N, Hardwick ], et al. Mycobacteria
miological assessment of extrapulmonary tuberculosis in San induces pleural mesothelial permeability by down-regulating
Francisco. Clin Infect Dis. 2004;38:25-3 1 . beta-catenin expression. Lung. 2003; 1 8 1 :57-66.
3 . Qiu L, Teeter LO, Liu Z, et al. Diagnostic associations 24. Valdes L, Alvarez D, Valle JM, et al. The etiology of pleural
berween pleural and pulmonary tuberculosis. J Infect. effusions in an area with high incidence of tuberculosis. Chest.
2006;53:377-3 86. 1 996; 1 09: 1 5 8- 1 62.
4. Torgersen ], Dorman SE, Baruch N, et al. Molecular epide 2 5 . al-Qorain A, Larbi EB, al-Muhanna F, et al. Pattern of pleu
miology of pleural and ocher excrapulmonary tuberculosis: a ral effusion in eastern province of Saudi Arabia: a prospective
Maryland state review. Clin Infect Dis. 2006;42: 1 375- 1 382. study. East Afr Med]. 1 994;7 1 :246-249.
5 . Berger HW, Mejia E. Tuberculous pleurisy. Chest. 26. Mehta JB, Dutt A, Harvill L, et al. Epidemiology of extrapul
l 973;63 : 8 8-92. monary tuberculosis. Chest. 1 9 9 1 ;99: 1 1 34-1 1 3 8 .
6. Stead WW, Eichenholz A, Stauss H-K. Operative and patho 27. Baumann M H , Nolan R , Petrini M, e t al. Pleural tuberculo
logic findings in rwenty-four patients with syndrome of idio sis in the United States: incidence and drug resistance. Chest.
pathic pleurisy with effusion, presumably tuberculous. Am Rev 2007; 1 3 1 : 1 1 25-1 1 32.
Respir Dis. 1 9 5 5 ; 7 1 :473-5 02. 28. Frye MD, Pozsik CJ, Sahn SA. Tuberculous pleurisy is more
7. Allen JC, Apicella MA. Experimental pleural effusion as a common in AIDS than in non-AIDS patients with tuberculo
manifestation of delayed hypersensitivity to tuberculin PPD. sis. Chest. 1 9 97; 1 1 2: 393-397.
] Immunol. 1 968; 1 0 1 :48 1 -487. 29. Mlika-Cabanne N, Brauner M, Kamanfu G, et al. Radio
8 . Leibowitz S, Kennedy L, Lessof MH. The tuberculin reaction graphic abnormalities in tuberculosis and risk of coexisting
in the pleural cavity and its suppression by antilymphocyte human immunodeficiency virus infection. Am J Respir Crit
serum. Br] Exp Pathol. 1 973;54: 1 52- 1 62. Care Med. 1 99 5 ; 1 52:794-799.
9 . Antony VB, Sahn SA, Antony AC, et al. Bacillus Calmette 30. Saks AM, Posner R. Tuberculosis in HIV positive patients
Guerin-stimulaced neutrophils release chemotaxins for in South Africa: a comparative radiological study with HIV
monocytes in rabbit pleural space in vitro. J Clin Invest. negative patients. Clin Radio!. 1 992;46:3 87-390.
1 9 8 5 ;76: 1 5 1 4- 1 5 2 1 . 3 1 . Queipo JA, Broseta E, Santos M, et al. Mycobacterial infec
1 0 . Mohammed KA, Nasreen N, Ward MJ, et al. Mycobacterium tion in a series of 1 ,26 1 renal transplant recipients. Clin Micro
mediated chemokine expression in pleural mesothelial cells: biol Infect. 2003; 9 : 5 1 8-525.
role of C-C chemokines in tuberculous pleurisy. J Infect Dis. 32. Malik GH, Al-Harbi AS, Al-Mohaya S, et al. Eleven years of
1 998; 1 78 : 1 45 0- 1 4 5 6 . experience with dialysis associated tuberculosis. Clin Nephrol.
1 1 . Widstrom 0, Nilsson BS. Pleurisy induced b y intrapleu 2002; 5 8 :3 5 6-362.
ral BCG in immunized guinea pigs. Eur J Respir Dis. 33. Hodsdon WS, Luzze H, Hurst TJ, et al. HIV- I -related pleu
l 982;63 :425-434. ral tuberculosis: elevated production of IFN-gamma, but
12. Widstrom 0, Nilsson BS. Low in vitro response to PPD and failure of immunity to Mycobacterium tuberculosis. AIDS.
PHA in lymphocytes from ECG-induced pleurisy in guinea 200 1 ; 1 5 :467-475 .
pigs. EurJ Respir Dis. 1 9 82;63 :43 5-44 1 . 3 4 . Awil P O , Bowlin SJ, Daniel T M . Radiology o f pulmonary
1 3 . Bueno CE, Clemente G , Castro BC, et al. Cytologic and bac tuberculosis and human immunodeficiency virus infection in
teriologic analysis of fluid and pleural biopsy specimens with Gulu, Uganda. Eur Respir]. 1 997; 1 0 :6 1 5-6 1 8 .
Cope's needle. Arch Intern Med. 1 990; 1 50 : 1 1 90-1 1 94. 3 5. Pozniak AL, Macleod GA, N dlovu D, et al. Clinical and chest
1 4 . Chan CH, Arnold M, Chan CY, et al. Clinical and pathologi radiographic features of tuberculosis associated with human
cal features of tuberculous pleural effusion and its long-term immunodeficiency virus in Zimbabwe. Am ] Respir Crit Care
consequences. Respiration. 1 9 9 1 ; 5 8 : 1 7 1 - 1 7 5 . Med. 1 99 5 ; 1 52 : 1 5 5 8- 1 5 6 1 .
1 5 . Fujiwara H, Tsuyuguchi I. Frequency o f tuberculin-reactive 3 6 . Levine H , Szanto PB, Cugell OW. Tuberculous pleurisy: an
T-lymphocytes in pleural fluid and blood from patients with acute illness. Arch Intern Med. 1 968; 1 22 :329-332.
tuberculous pleurisy. Chest. 1 986;89: 530-532. 37. Richter C, Perenboom R, Mtoni I, et al. Clinical features of
16. Shimokata K, Kawachi H, Kishimoto H, et al. Local cel HIV-seropositive and HIV-seronegative patients with tuber
lular immunity in tuberculous pleurisy. Am Rev Respir Dis. culous pleural effusion in Dar es Salaam, Tanzania. Chest.
1 982; 1 2 8 : 822-824. 1 994; 1 06: 1 4 7 1 - 1 47 5 .
260 PLE U RAL D I S EASES
3 8 . Ibrahim WH, Ghadban W, Khinji A, et al. Does pleural tuber 6 0 . Burgess LJ, Maritz FJ, L e Roux I, e t al. Use o f adenosine
culosis disease pattern differ among developed and developing deaminase as a diagnostic tool for tuberculous pleurisy. Tho
countries. Respir Med. 2005;99: 1 038- 1 04 5 . rax. 1 995;50:672-674.
3 9 . Valdes L , Alvarez D, San Jose E, e t al. Tuberculous pleurisy: a 6 1 . Valdes L, Alvarez D, San Jose E, et al. Value of adenosine
study of 254 patients. Arch Intern Med. 1 998; 1 5 8 :20 1 7-202 1 . deaminase in the diagnosis of tuberculous pleural effusions in
40. Maher GG, Berger HW. Massive pleural effusion: malignant young patients in a region of high prevalence of tuberculosis.
and non-malignant causes in 46 patients. Am Rev Respir Dis. Thorax. 1 995;50:600-603.
1 972; I 05:45 8-460. 62. Riantawan P, Chaowalit P, Wongsangiem M, et al. Diagnos
4 1 . Seiscento M , Vargas FS, Bombarda S, et al. Pulmonary tic value of pleural fluid adenosine dearninase in tuberculous
involvement in pleural tuberculosis: How often does it mean pleuritis with reference to HIV coinfection and a Bayesian
disease activity? Respir Med. 20 1 1 ; 1 0 5 : 1 079-1 083. analysis. Chest. 1 999; 1 1 6:97- 1 03 .
42. Kim HJ, Lee HJ, Kwon SY, et al. The prevalence of pulmonary 63. Liang QL, S h i H Z , Wang K , e t a l . Diagnostic accuracy o f
parenchymal tuberculosis in patients with tuberculous pleuri adenosine dearninase in tuberculous pleurisy: A meta-analysis.
tis. Chest. 2006; 1 2 9 : 1 253- 1 2 5 8 . Respir Med. 2008; 1 02:744-754.
43. Ariyurek O M , Ci! B E . Atypical presentation o f pleural tuber 64. Burgess LJ, Maritz FJ, Le Roux I, et al. Combined use of pleu
culosis: CT fi n dings. Br] Radio!. 2000;73 :209-2 1 0 . ral adenosine deaminase with lymphocyre/neutrophil ratio.
44. Heyderman RS, Makunike R, Muza T, et al. Pleural tubercu Increased specificity for the diagnosis of tuberculous pleuritis.
losis in Harare, Zimbabwe: the relationship between human Chest. 1 996; 1 09 :4 1 4-4 1 9 .
immunodeficiency virus, CD4 lymphocyte count, granuloma 6 5 . Valdes L , San Jose MA, Pose A , e t al. Diagnosing tuberculous
formation and disseminated disease. Trop Med Int Health. pleural effusion using clinical data and pleural fluid analysis. A
1 998;3: 1 4-20. study of patients less than 40 years-old in an area with a high
4 5 . Luzze H, Elliott AM, Joloba ML, et al. Evaluation of sus incidence of tuberculosis. Respir Med. 20 1 0 ; 1 04: 1 2 1 1 - 1 2 1 7.
pected tuberculous pleurisy: clinical and diagnostic findings 66. Diacon AH, Van de Wal BW, Wyser C, et al. Diagnostic tools
in HIV- I -positive and HIV-negative adults in Uganda. Int J in tuberculous pleurisy: a direct comparative study. Eur Respir
Tuberc Lung Dis. 200 1 ; 5 :746-753. ]. 2003;22 : 5 8 9-5 9 1 .
46. Collins KR, Quinones-Mateu ME, Wu M, et al. Human 67. Esteban C , Oribe M , Fernandez A, et al. Increased adenosine
immunodeficiency virus type I (HIV- I ) quasispecies at the deaminase activity in Q fever pneumonia with pleural effu
sites of Mycobacterium tuberculosis infection contribute to sys sion. Chest. 1 994; 1 0 5 :648.
temic HIV- I heterogeneity. J Virol. 2002;76: 1 697- 1 706. 68. Dikensoy 0 , Namiduru M , Hocaoglu S , et al. Increased
47. Patiala J . Initial tuberculous pleuritis in the Finnish Armed pleural fluid adenosine deaminase in brucellosis is dif
Forces in 1 939- 1 945 with special reference to eventual post ficult to differentiate from tuberculosis. Respiration.
pleuritic tuberculosis. Acta Tuberc Scand Suppl. 1 9 54;36: 1-57. 2002;69 : 5 5 6- 5 5 9 .
4 8 . Roper WH, Waring ]] . Primary serofi b rinous pleural effusion 69. Lee YC, Rogers JT, Rodriguez RM , et al. Adenosine deami
in military personnel. Am Rev Respir Dis. 1 9 5 5 ;7 1 : 6 1 6-634. nase levels in nontuberculous lymphocytic pleural effusions.
49. Patiala J, Mattila M . Effect of chemotherapy of exudative Chest. 200 1 ; 1 20:356-36 1 .
tuberculous pleurisy on the incidence of post pleuritic tuber 70. Chung JH, Kim YS, Kim SI, e t al. The diagnostic value of
culosis. Acta Tuberc Scand. l 964;44:290-296. the adenosine deaminase activity in the pleural fluid of renal
5 0 . Light RW. Establishing the diagnosis of tuberculous pleuritis. transplant patients with tuberculous pleural effusion. Yonsei
Arch Intern Med. 1 99 8 ; 1 5 8 : 1 967- 1 96 8 . Med]. 2004;45 :66 1 -664.
5 1 . Yam LT. Diagnostic significance of lymphocytes i n pleural 7 1 . Ferrer JS, Munoz XG, Orriols RM , et al. Evolution of idio
effusions. Ann Intern Med. 1 967;66:972-9 82. pathic pleural effusion. A prospective, long-term follow-up
52. Light RW, Erozan YS, Ball WC. Cells in pleural fluid: study. Chest. 1 996; 1 09 : 1 508- 1 5 1 3 .
their value in differential diagnosis. Arch Intern Med. 7 2 . Jimenez Castro D , Diaz Nuevo G , Perez-Rodriguez E,
1 973; 1 32 : 8 54-860. et al. Diagnostic value of adenosine deaminase in non
5 3 . De Oliveira HG, Rossano ER, Prolla JC. Pleural fluid adenos cuberculous lymphocytic pleural effusions. Eur Respir ].
ine deaminase and lymphocyte proportion: clinical usefulness 2003;2 I :220-224.
in the diagnosis of tuberculosis. Cytopathology. 1 994; 5 : 27-32. 73. Perez-Rodriguez E, Castro DJ. The use of ADA and ADA iso
54. Bielsa S, Palma R, Esquerda A et al. Comparison of polymor enzymes in the diagnosis of tuberculous pleuritis. Curr Opin.
phonuclear and lymphocytic-rich tuberculous pleural effu 2000;6:259-266.
sions. J Tuberc Lung Dis. In press. 74. Valdes L, San Jose E, Alvarez D, et al. Adenosine deaminase
5 5 . Spriggs AI, Boddington MM. 1he Cytology of Effesions, 2nd (ADA) isoenzyme analysis in pleural effusions: diagnostic role,
ed. New York, NY: Grune & Stratton; 1 968. and relevance to the origin of increased ADA in tuberculous
56. Hurwitz S, Leiman G, Shapiro C. Mesothelial cells in pleural pleurisy. Eur Respir]. l 996;9:747-75 1 .
fluid: TB or not TB? S Afr Med]. 1 980;57:937-939. 7 5 . Miller KD, Barnette R, Light RW. Stability of adenosine deam
57. Jones D, Lieb T, Narita M, et al. Mesothelial cells in tuber inase during transportation. Chest. 2004; 1 26: 1 933- 1 937.
culous pleural effusions of HIV-infected patients. Chest. 76. Shimokata K, Saka H, Murate T, et al. Cytokine content in
2000; 1 1 7:289-29 1 . pleural effusion. Chest. 1 99 1 ;99: 1 1 03-1 1 07.
5 8 . Ocana I , Martinez-Vazquez JM, Segura RM , et al. Adenosine 77. Ribera E, Ocana I, Martinez-Vazquez JM, et al. High level
deaminase in pleural fluids: test for diagnosis of tuberculous of interferon gamma in tuberculous pleural effusion. Chest.
pleural effusion. Chest. 1 983;84:5 1 -53. I 988;93:308-3 1 1 .
5 9 . Valdes L, San Jose E, Alvarez D, et al. Diagnosis of tuberculous 78. Barnes PF, Mistry SD, Cooper CL, et al. Compartmentaliza
pleurisy using the biologic parameters adenosine deaminase, tion of a CD4• T lymphocyte subpopulation in tuberculous
lysozyme, and interferon gamma. Chest. 1 993; 1 03:45 8-465 . pleuritis. J Immunol. 1 989; 1 42 : 1 1 1 4- 1 1 1 9 .
C H A PT E R 1 3 I TU B E RC U LO U S P L E U RAL E F F U S I O N S 26 1
79. Villena V, Lopez-Encuentra A, Echave-Sustaeta J, et al. Inter 1 00 . Maartens G, Bateman ED. Tuberculous pleural effusions:
feron-gamma in 3 8 8 immunocompromised and immuno increased culture yield with bedside inoculation of pleural
competent patients for diagnosing pleural tuberculosis. Eur fluid and poor diagnostic value of adenosine deaminase. Tho
Respir]. 1 996;9:263 5-2639 . rax. 1 9 9 1 ;46:96-99 .
80. Jiang J, Shi H Z , Liang QL, e t al. Diagnostic value of inter 1 0 1 . Schmid GP, Catino D , Suffin SC, et al. Granulomatous
feron-gamma in tuberculous pleurisy: a metaanalysis. Chest. pleuritis caused by Francisella tularensis: possible con
2007; 1 3 1 : 1 1 33-1 1 4 1 . fusion with tuberculous pleuritis. Am Rev Respir Dis.
8 1 . Villena V, Lopez-Encuentra A, Pozo F, e t al. Interferon gamma 1 9 83; 1 28:3 1 4-3 1 6.
levels in pleural fluid for the diagnosis of tuberculosis. Am J 1 02 . Yum HK, Choi SJ . Detection of mycobacterial DNA using
Med. 2003; 1 1 5 :365-370. nested polymerase chain reaction of pleural biopsy speci
82. Greco S, Girardi E, Masciangelo R, et al. Adenosine deami mens: compared to pathologic findings. Korean J Intern Med.
nase and interferon gamma measurements for the diagnosis 2003 ; 1 8 : 8 9-93 .
of tuberculous pleurisy: a meta-analysis. Int] Tuberc Lung Dis. 1 03 . Light RW. Closed needle biopsy o f the pleura i s a valuable
2003;7:777-786. diagnostic procedure. Con closed needle biopsy. J Bronchol.
83. Gupta UA, Chhabra SK, Hiraki A, et al. Diagnosing tubercu 1 99 8 ; 5 :332-336.
lar pleural effusions. Chest. 2005 ; 1 27 : 1 078- 1 079. 1 04. Small PM, Fujiwara Pl. Management of tuberculosis in the
84. Sharma SK, Banga A. Pleural fluid interferon-gamma and United States. N Engl] Med. 200 1 ;345 : 1 8 9-200.
adenosine deaminase levels in tuberculosis pleural effusion: a 105. Blumberg HM, Burman WJ , Chaisson RE, et al. American
cost-effectiveness analysis. ] Clin Lab Anal. 2005; 1 9 :40-46. Thoracic Society/Centers for Disease Control and Prevention/
8 5 . Zhou Q, Chen YQ, Qin SM, et al. Diagnostic accuracy of Infectious Diseases Society of America: treatment of tuber
T-cell interferon-gamma release assays in tuberculous pleurisy: culosis. Am J Respir Crit Care Med. 2003; 1 67:603-662.
a meta-analysis. Respirology. 20 1 1 ; 1 06:473-480. 1 06. Jutte PC, Rutgers SR, Van Altena R, et al. Penetration of
86. Chegou NN, Walz! G, Bolliger CT, et al. Evaluation ofadapted isoniazid, rifampicin and pyrazinamide in tuberculous
whole-blood interferon-gamma release assays for the diagnosis pleural effusion and psoas abscess. Int J Tuberc Lung Dis.
of pleural tuberculosis. Respiration. 2008;76: 1 3 1 - 1 3 8 . 2004;8 : 1 368-1 372.
8 7 . Pai M, Flores LL, Hubbard A , e t al. Nucleic acid amplifica 1 07. Canete C, Galarza I, Granados A, et al. Tuberculous pleural
tion tests in the diagnosis of tuberculous pleuritis: a systematic effusion: experience with six months of treatment with iso
review and meta-analysis. BMC Infect Dis. 2004;4 : 1 - 1 4 . niazid and rifampicin. Thorax. 1 994;49: 1 1 60- 1 1 6 1 .
8 8 . Levy H, Wayne LG, Anderson B E , et al. Anti-mycobacterial 1 0 8 . Dutt AK, Moers D , Stead WW. Tuberculous pleural effu
antibody levels in pleural fluid reflect passive diffusion from sion: 6-month therapy with isoniazid and rifampin. Am Rev
serum. Chest. 1 990;97: 1 1 44- 1 1 47. Respir Dis. 1 992; 1 45 : 1 429- 1 432.
8 9 . Barber LM, Mazzadi L, Deakins DD, et al. Glucose level in 1 0 9 . Falk A, Stead WW. US Veterans Administration Armed
pleural fluid as a diagnostic aid. Dis Chest. 1 957;3 l :680-68 1 . Forces cooperative studies of tuberculosis: V. Antimi
90. Holton K. Diagnostic value o f some biochemical pleural fluid crobial theory in treatment of primary tuberculous
examinations. Scand] Respir Dis Suppl. 1 968;63 : 1 2 1 - 1 2 5 . pleurisy with effusion: the effect upon the incidence of
9 1 . Light RW, MacGregor Ml, Ball W C Jr, et a l . Diagnos subsequent tuberculous relapse. Am Rev Tuberc Pulmon Dis.
tic significance of pleural fluid pH and PCO 2 • Chest. 1 9 5 6;74:897-902.
1 973;64 : 5 9 1 -596. 1 1 0 . Tani P, Poppius H, Makipaja J. Cortisone therapy for exuda
92. Chavalittamrong B, Angsusingha K, Tuchinda M, et al. Diag tive tuberculous pleurisy in the light of the follow-up study.
nostic significance of pH, lactic acid dehydrogenase, lactate Acta Tuberc Scand. 1 964;44:303-309.
and glucose in pleural fluid. Respiration. 1 979;38: 1 1 2- 1 20 . 1 1 1 . Galarza I, Canete C, Granados A, et al. Randomised trial
93. G o o d JT Jr, Taryle DA, Maulitz RM, et a l . The diagnostic of corticosteroids in the treatment of tuberculous pleurisy.
value of pleural fluid pH. Chest. 1 9 80;78 : 5 5-59. Thorax. 1 99 5 ; 5 0 : 1 305-1 307.
94. Chierakul N, Kanitsap A, Chaiprasert A, et al. A simple 1 1 2 . Wyser C, Walz! G, Smedema JP, et al. Corticosteroids in the
C-reactive protein measurement for the differentiation treatment of tuberculous pleurisy. A double-blind, placebo
between tuberculous and malignant pleural effusion. Respirol controlled, randomized study. Chest. 1 996; 1 1 0 :333-338.
ogy. 2004;9 : 66-69. 1 1 3 . Barbas CSV, Cukier A, de Varvalho CRR, et al. The rela
9 5 . Garcia-Pachon E, Soler MJ, Padilla-Navas I, et al. C-reactive tionship between pleural fluid findings and the development
protein in lymphocyric pleural effusions: a diagnostic aid in of pleural thickening in patients with pleural tuberculosis.
tuberculous pleuritis. Respiration. 2005 ;72:486-489. Chest. 1 99 1 ; 1 00 : 1 264- 1 267.
96. Fontan Bueso J, Verea Hernando H, Garcia-Buela JP, et al. 1 1 4. Wong P-C. Management of tuberculous pleuritis: can we do
Diagnostic value of simultaneous determination of pleural better? Respirology. 2005; 1 0: 1 44- 1 4 8 .
adenosine deaminase and pleural lysozyme/serum lysozyme 1 1 5 . Candela A , Andujar J, Hernandez L, et al. Functional
ratio in pleural effusion. Chest. l 988;93:303-307. sequelae of tuberculous pleurisy in patients correctly treated.
97. Verea Hernando HR, Masa Jimenez JF, Dominguez Juncal L, Chest. 2003; 1 2 3 : 1 996-2000.
et al. Meaning and diagnostic value of determining the lyso 1 1 6. Zhou A, Guo L, Tang L. Effect of an intrathoracic injec
zyme level of pleural fluid. Chest. 1 987;9 1 : 342-345. tion of sodium hyaluronic acid on the prevention of pleural
9 8 . Ugurman F, Gozu A, Akkalyoncu B, et al. Tumour necrosis thickening in excess fluid of tuberculous thoracic cavity. Clin
factor-alpha in comparison ro adenosine deaminase in tuber Exp Pharmacol Physiol. 2003;30:203-205.
culous pleuritis. Respiration. 2003;70:270-274. 1 1 7. Han DH, Song JW, Chung HS, et al. Resolurion of resid
99. Conde MB, Loivos AC, Rezende VM, et al. Yield of sputum ual pleural disease according to time course in tuberculous
induction in the diagnosis of pleural tuberculosis. Am J Respir pleurisy during and afrer the termination of anti tuberculosis
Crit Care Med. 2003 ; 1 67:723-725 . medication. Chest. 2005; 1 28 :3240-3245 .
262 PLE U RAL D I S EASES
1 1 8 . Lai YF, Chao TY, Wang YH, et al. Pigtail drainage in the 129. Johnson TM, McCann W, Davey WN. Tuberculous bron
treatment of tuberculous pleural effusions: a randomized chopleural fi s tula. Am Rev Respir Dis. 1 973; 1 07:30-4 1 .
study. Thorax. 2003; 5 8 : 1 49-1 5 1 . 1 30. Jenssen AD. Chronic calcified pleural empyema. Scand J
1 1 9 . Kwak SM, Park CS, Cho JH, et al. The effects of urokinase Respir Dis. 1 969;50: 1 9-27.
instillation therapy via percutaneous transthoracic catheter 1 3 1 . Mouroux J, Maalouf ], Padovani B, et al. Surgical manage
in loculated tuberculous pleural effusion: a randomized pro ment of pleuropulmonary tuberculosis. J Thorac Cardiovasc
spective study. Yonsei MedJ 2004;45 : 822-828 . Surg. 1 996; 1 1 1 :662-670.
1 2 0 . Chung CL, Chen C H , Yeh CY, e t al. Early effective drainage 1 32. Sahn SA, Iseman MD. Tuberculous empyema. Semin Respir
in rhe treatment of loculared tuberculous pleurisy. Eur Respir Infect. 1 999; 1 4 : 82-87.
J 2008;3 1 : 1 2 6 1 - 1 267. 1 33 . Mancini P, Mazzei L, Zarzana A, et al. Post-tuberculosis
1 2 1 . Al-Majed SA. Study of paradoxical response to che chronic empyema of rhe "forty years after." Eur Rev Med
motherapy in tuberculous pleural effusion. Resp Med. Pharmacol Sci. 1 99 8 ;2:25-29 .
1 996;90:2 1 1 -2 1 4 . 1 34. Jover F, Andreu L , Cuadrado J M , e t al. Tuberculous empy
1 22. Hiraoka K , Nagata N, Kawajiri T, e t al. Paradoxical pleural ema necessitatis in a man infected with the human immuno
response to antituberculous chemotherapy and isoniazid deficiency virus. South Med J 2002;9 5 :75 1-752.
induced lupus. Review and report of rwo cases. Respiration. 135. Neihart RE, Hof DG. Successful nonsurgical treatment of
1 998;65 : 1 52- 1 5 5 . tuberculous empyema in an irreducible pleural space. Chest.
1 2 3 . Choi YW, Jeon SC, Seo H S , e t al. Tuberculous pleural effu 1 9 8 5 ; 8 8 :792-794.
sion: new pulmonary lesions during treatment. Radiology. 1 36. Iseman MD, Madsen LA. Chronic tuberculous empyema
2002;224:493-502. with bronchopleural fistula resulting in treatment failure and
1 24. Gupta RC, Dixie R, Purohit SD, et al. Development of pleu progressive drug resistance. Chest. 1 99 1 ; 1 00 : 1 24- 1 27.
ral effusion in patients during anti-tuberculous chemother 1 37. Christensen EE, Dietz GW, Ahn CH, et al. Initial roent
apy: analysis of rwenty-nine cases with review of literature. genographic manifestations of pulmonary Mycobacterium
Indian J Chest Dis Allied Sci. 2000;42: 1 6 1 - 1 66. tuberculosis, M. kansasii, and M. intracellularis infections.
125. Lee CH, Wang WJ, Lan RS, et al. Corticosteroids in the Chest. 1 9 8 1 ; 80: 1 32- 1 36.
treatment of tuberculous pleurisy: a double-blind, placebo 138. Gribetz AR, Damsker B, Marchevsky A, et al. Nontuber
controlled, randomized study. Chest. 1 9 88;94: 1 256-1 259. culous mycobacteria in pleural fluid: assessment of clinical
1 26. Elliott AM, Luzze H, Quigley MA, et al. A randomized, significance. Chest. 1 9 8 5 ; 87:495-498.
double-blind, placebo-controlled trial of the use of predniso 1 39 . Shu CC, Lee LN, Wang JT, et al. Non-tuberculous myco
lone as an adjunct to treatment in HIV- I -associated pleural bacterial pleurisy: an 8-year single-centre experience in
tuberculosis. ] Infect Dis. 2004; 1 90: 869-878. Taiwan. Int] Tuberc Lung Dis. 20 I O ; 1 4:63 5-64 1 .
1 27. Large SE, Levick RK. Aspiration in rhe treatment of primary 1 40. Aronchick JM, Miller WT. Disseminated nontuberculous
tuberculous pleural effusion. Br Med J 1 9 5 8 ; I : 1 5 1 2- 1 5 1 4 . mycobacterial infections in immunosuppressed patients.
1 2 8 . Engel M E , Matchaba PT, Volmink J. Corticosteroids Semin Roentgenol. 1 993;8 : 1 5 0- 1 57.
for tuberculous pleurisy. Cochrane Database Syst Rev.
2007;4:CDOO 1 876.
P l e u ra l E ff u s i o n S e co n d a ry to F u n g a l
I n fe ct i o n s, Act i n o m yco s i s, a n d N o ca rd i o s i s
In this chapter, pleural disease resulting from fun Th e second situation i n which pleural aspergillosis
gal infections is discussed. Although fungal diseases occurs is postoperatively after lobectomy or pneumo
account for less than 1 % of all pleural effusions, it nectomy for lung cancer, tuberculosis, or aspergillosis
is important to identify correctly patients with fun (3,4,6) . A bronchopleural fistula is almost invari
gal disease of the pleura because effective treatment ably present. The clinical picture is similar to that of
is available. Actinomycosis and nocardiosis are also a pleural bacterial infection after lung resection (see
included in this chapter because they produce a Chapter 1 2 ) . On rare occasions, the pleural fluid
chronic disease similar to that caused by the fungi, becomes infected with aspergillus in the immunosup
although they actually are bacteria. pressed patient with systemic aspergillosis (7) . One
report cited two patients with pleural effusion com
plicating allergic bronchopulmonary aspergillosis (8),
AS P E RG I LLO S I S but the relationship between the pleural effusion and
Occasionally, the pleural space becomes infected with the allergic aspergillosis was not convincing.
the Aspergillus species of fungus. The usual infect There has been one case report of a patient with
ing organism is Aspergillus fomigatus ( l ) , but other allergic bronchopulmonary aspergillosis who pre
species such as Aspergillus niger may also be respon sented with a large recurrent pleural effusion (9) .
sible (2) . Pleural aspergillosis is uncommon, but The pleural fluid was an exudate and the effusion
1 3 cases were observed in one institution during a responded to corticosteroid therapy (9) .
recent 5-year period (3) .
D i a g nosis
Cl i n ical M a n ifestati o n s
The diagnosis of pleural aspergillosis should be sus
Pleural aspergillosis usually occurs in one of two pected in any patient with a history of artificial pneu
settings. Most commonly, it occurs in patients who mothorax therapy for tuberculosis who has a chronic
were treated in the past with artificial pneumothorax pleural infection, particularly when a bronchopleu
therapy for tuberculosis (2-4) . Such patients have ral fistula is present. Similarly, the diagnosis should
signs and symptoms of a chronic infection, includ be suspected in any patient with a pleural infection
ing weight loss, malaise, a low-grade fever, and a after lung resection. The diagnosis is confirmed by
chronic, productive cough ( 1 ) . The chest radiograph the demonstration of aspergillus on fungal cultures
reveals increasing degrees of pleural thickening and of the pleural fluid. The presence of brown clumps
usually an air-fluid level in the pleural space indi containing fungal hyphae in the pleural fluid suggests
cating the presence of a bronchopleural fistula ( 1 ) . the diagnosis (4) . Patients with pleural aspergillosis
Fungus balls, although uncommon, may b e evident almost always have positive precipitin blood tests for
radiographically either in the lungs or the pleural antibodies against aspergillus ( 1 ,4) . Aspergillus anti
space ( 1 , 5 ) . gens can also be demonstrated in the pleural fluid
263
264 PLE U RAL D I S EASES
infection and may or may not have concomitant patients with disseminated disease died within a short
parenchymal involvement. The second rype occurs period, whereas the other three patients had minor
when a coccidioidal caviry adjacent to the pleura rup complications and were treated with amphotericin B .
tures to produce a hydropneumothorax with a bron Although the CF titers are high i n patients with pleu
chopleural fistula. ral coccidioidomycosis (24) and high (> 1 : 1 6) CF
titers are used by some as an indication of dissemi
nation (26) , patients with pleural coccidioidomycosis
Pri m a ry I n fection
and high CF titers should be treated only if their skin
The pleura is frequently involved in primary infec tests are negative or if other evidence of dissemination
tions with C. immitis. As many as 70% of patients exists. The treatment of choice is either itraconazole
have pleuritic chest pain, and approximately 20% or amphotericin B ( 1 2) . If the patient has a coccidio
have blunting of the costophrenic angles radiologi dal empyema, treatment should be with Buconazole
cally (24) . Approximately 7% of all symptomatic or itraconazole plus tube thoracostomy (25 ) .
patients with primary coccidioidomycosis have pleu
ral effusions (24) . However, on one study (25 ) , 22
Ru ptu re o f Coccid i o i d a l Cavity
of 1 46 patients ( 1 5%) hospitalized with coccidioido
mycosis had a pleural effusion. Patients with pleural The second situation in which pleural disease occurs
effusions secondary to coccidioidomycosis are almost with coccidioidomycosis is when a coccidioidal caviry
always febrile, and more than 80% have pleuritic adj acent to the pleura ruptures into the pleural space
chest pain (24) . Approximately 50% of patients to produce a bronchopleural fistula and a hydropneu
have either erythema nodosum or erythema multi mothorax. Hydropneumothoraces develop in 1 % to
forme (24) . The chest radiograph reveals parenchy 5% of patients with chronic cavitary coccidioidomy
mal infiltrates in addition to the pleural effusion in cosis and occasionally occur without a prior cavita
approximately 50% of patients. The pleural effusion tion (23) . Many patients who experience rupture of a
varies in size, but it often occupies more than 50% of coccidioidal caviry have no history of coccidioidomy
the hemithorax (24) . In one series of 28 patients, all cosis (27) . Accordingly, in endemic areas, all patients
pleural effusions were unilateral (24) . with spontaneous hydropneumothoraces should be
Pleural Buid analysis reveals an exudate that usu evaluated for the possibiliry of coccidioidomycosis.
ally contains predominantly small lymphocytes When a coccidioidal caviry ruptures into the pleu
(24) . Although approximately 50% of patients have ral space, the patient usually becomes acutely ill, with
peripheral eosinophilia, pleural Buid eosinophilia is systemic signs of toxiciry. The pleural Buid cultures
uncommon and occurred in only 1 of 1 5 patients in are usually positive for C. immitis and the CF tests
one series (24) . However, in another series (25) the are almost always positive (27) . In one case report,
mean percentage of eosinophils in the pleural Buid the pleural Buid glucose was 0 mg/dL and the lac
was 1 0 .3%. The pleural Buid glucose level is almost tic acid dehydrogenase was more than 3 , 5 00 U/mL
always above 60 mg/dL (24) , and the pH is normal (26) . Patients with a hydropneumothorax should
unless the patient has an empyema, which occurred in have a chest tube inserted immediately to drain the
5 of 1 5 patients (33%) in one series (25 ) . The pleural air and the Buid from the pleural space. They should
Buid cultures are positive for C. immitis in approxi also be given itraconazole or amphotericin B systemi
mately 20% of patients, but cultures of pleural biopsy cally. Most of them require additional surgery such
specimens are positive in almost all patients (24) . In as a partial lobectomy for control of the caviry. In
one series, eight of eight pleural biopsy cultures were one series of 23 patients, all but 2 required surgical
positive, and cocci spherules were identified in six of treatment in addition to the tube thoracostomy (27) .
eight specimens (24) . The pleural biopsy may reveal In view of this observation, it is recommended that
caseating or noncaseating granulomas (24) . The cocci patients who have a persistent bronchopleural fistula
skin tests are usually positive, and the mean comple for more than 7 days be subjected to surgery.
ment fixation (CF) titer 6 weeks after the onset of
symptoms is 1 :32 (24) .
Most patients with primary coccidioidomycosis
C RY PTOCOCCO S I S
and pleural effusion require no systemic antifungal
therapy (24) . In a series of28 such patients, 23 (82%) Cryptococcus neoformans, a fungus distributed world
recovered completely without specific therapy. Two wide, lives in soil, particularly that contaminated by
266 PLE U RAL D I S EASES
pigeon excreta. On rare occasions, infection with lymphoma, diabetes mellitus, or sarcoidosis, and
C. neoformans produces a pleural effusion. Until patients receiving corticosteroids or immunosuppres
1 980, only 30 cryptococcal pleural effusions had sant agents should also be treated (37) . If any of the
been reported (28 ). With the advent of the acquired foregoing conditions are met, it is recommended that
immunodeficiency syndrome (AIDS) epidemic, pleu the patient be treated with 400 mg/day of fluconazole
ral effusions secondary to cryptococcosis have become for 6 months (22) . On rare occasions in immunosup
much more frequent. In one series of 12 patients with pressed individuals, the pleural infection is so over
pulmonary cryptococcal involvement proved by cul whelming that tube thoracostomy is indicated (38) .
ture, 3 (25%) had a pleural effusion (29) . In another
series of 75 patients with AIDS and pleural effusion
H I STO P LA S M O S I S
from Paris, 4 (5%) had pleural cryptococcosis (30) .
Pleural cryptococcosis appears to result from an Histoplasma capsulatum is a fungus that lives i n the
extension of a primary subpleural cryptococcal nod mycelial form in soil and is distributed throughout
ule into the pleural space (3 1 ) . the temperate zones of the world but is most heavily
I n patients with pleural cryptococcosis, the disease endemic in central United States (39) . Infection with
is localized to the hemithorax in approximately 50% H. capsulatum only rarely produces pleural effusions.
and is disseminated in the remaining 50% (28) . More Although it has been estimated that 500,000 persons
than half the patients have serious underlying dis are infected annually in the United States (39), fewer
ease, most commonly leukemia, lymphoma, or AIDS than 20 pleural effusions secondary to histoplasmosis
(28,32) . In 27 of the 30 cases reported up to 1 9 80, have been reported. In a review of the radiographic
the pleural effusion was unilateral (28) and most of manifestations of pulmonary histoplasmosis, only 1
the effusions associated with AIDS are also unilateral. patient of 259 with abnormal chest radiographs had a
The size of the pleural effusion ranges from minimal pleural effusion (40) .
to massive (28 ). Most patients also have an accompa Patients with pleural effusions secondary to his
nying parenchymal lesion in the form of a nodule, a toplasmosis usually have a subacute illness character
mass, or an interstitial infiltrate (28) . The pleural fluid ized by a low-grade fever and pleuritic chest pain.
is an exudate, usually with a predominance of small The chest radiograph usually reveals an infiltrate or
lymphocytes. In one case report (33), the pleural a subpleural nodule in addition to the pleural effu
fluid adenosine deaminase level was elevated making sion (4 1 -44) . Pleural fluid analysis reveals an exudate
the differentiation from tuberculous pleuritis difficult. containing predominantly lymphocytes. In two of
One case report cited an effusion in which the pleu the reported cases (4 1 ,43) , pleural fluid eosinophilia
ral fluid contained 1 5% eosinophils (34) . Cultures of was present. The pleural biopsy may reveal noncaseat
the pleural fluid were positive in 1 1 of 26 patients in ing granulomas. The diagnosis is made by culturing
the 1 9 80 series (28) . In the remaining patients, the H. capsulatum from the pleural fluid, sputum, or biopsy
diagnosis was made by histologic study or by culture material by routine fungal cultures or by demonstrat
of lung tissue obtained at operation or autopsy (28) . ing the organism in biopsy material with appropri
Patients with cryptococcal pleural effusion have high ate stains. A presumptive diagnosis can sometimes be
titers of cryptococcal antigen in their pleural fluid established by demonstrating a high histoplasmosis
and serum (28) . CF titer (> 1 : 32) or an M band on counterimmu
It is not clear whether treatment with systemic noelectrophoresis (45 ) . It appears that treatment is
antifungal agents is necessary for all patients with not necessary for pleural effusions secondary to his
pleural cryptococcosis. Several patients have recov toplasmosis (42,44,46) . The pleural effusion usually
ered without any specific therapy (28 , 3 5 ,36) . There resolves spontaneously over several weeks (4 1 ,42,47) .
fore, it is recommended that blood and cerebrospinal On rare occasions, however, a patient develops a
fluid (CSF) be studied for cryptococcal antigen. fibrosing pleuritis for which a decortication should
If cryptococcal antigen is detected in the CSF, the be considered if the patient is symptomatic (42,48) .
patient should be treated with amphotericin B and An isolated pleural effusion due to infection
5 -fluorocytosine (37) . If the cryptococcal antigen with H. capsulatum has been reported in a patient
in the CSF is negative and the patient is symptom with AIDS (49) . This patient presented with fever
atic or the blood antigen is positive, then the patient and bilateral small pleural effusions. Thoracentesis
should be treated with fluconazole (37) . lmmunosup revealed an exudate with many organisms typical of
pressed patients, such as those with AIDS, leukemia, H. capsulatum on the Wright-Giemsa stain of the
C HAPT E R 1 4 / PLE U RAL E F F U S I O N S E C O N DARY TO F U N GAL I N F ECTI O N S, ACTI N O MYCOS I S, A N D N O CARD I O S I S 267
pleural fluid. This patient appeared to respond to areas where parenchymal actinomycosis has extended
therapy with amphotericin B (49) . Pleural effusions through the chest wall to produce a chest wall abscess
can also occur in patients with AIDS and dissemi or a draining sinus. In a more recent series using com
nated histoplasmosis (50) , but pleural involvement is puted tomography (CT) scans, pleural effusions were
not a prominent part of the disease picture. present in five of eight patients (62. 5%) with acti
On rare occasions, patients with parenchymal nomycosis, although there was only enough pleural
histoplasmosis may develop a bronchopleural fistula fluid for thoracentesis in three (57) . Pleural thicken
with the subsequent development of a loculated fun ing was demonstrated with CT scans in all eight of
gal empyema (5 1 ) . Such patients should be treated the patients in this latter series (57) .
with drainage and decortication. The pleural fluid with actinomycosis may be either
frank pus with predominantly polymorphonuclear
leukocytes (58) or serous fluid with predominantly
CA N D I D I A S I S lymphocytes (59) . I have seen a patient with tho
racic actinomycosis in which the associated pleural
O n occasion, pleural fluid cultures are positive for
effusion was serous and contained more than 50%
candida. Pleural infection with candida usually
eosinophils.
occurs because of a leak from the gastrointestinal
tract and often occurs in the postoperative period
(52, 53) . There is coinfection with bacteria in most D i a g nosis
patients. In patients with pleural infection with can
The diagnosis of thoracic actinomycosis should be
dida, the possibility of esophageal perforation should
suspected in any patient with a chronic infiltrative
always be suspected (see Chapter 1 8) . The most
pulmonary disease, particularly when the paren
important aspect in the treatment of these patients
chymal disease crosses lung fissures. The presence
is to repair the esophageal perforation if one is pres
of chest wall abscesses or draining sinus tracts sug
ent. In addition, appropriate antibiotics should be
gests the diagnosis, as do bone changes consisting
administered and purulent fluid should be drained
of periosteal proliferation or bone destruction (56) .
through percutaneous catheter drainage (52) . For
Thoracic actinomycosis sometimes becomes dissemi
antifungal therapy both conventional amphotericin B
nated and produces peripheral abscesses in the skin,
(0 .7 mg/kg daily) and fluconazole (400 mg daily) are
subcutaneous tissues, or muscles (60) . The diagnosis
options suggested in the 2004 Infectious Diseases
is suggested by the presence of sulfur granules in the
Society of America (IDSA) guidelines (54) .
draining exudate or the pleural fluid. These granules
are 1 to 2 mm in diameter and consist of clumps of
thin bacterial filaments that possess peripheral radia
ACTI N O MYCO S I S tions with or without clubbing at their ends. Sulfur
Actinomyces israelii, an anaerobic o r microaerophilic granules may be associated with cutaneous nocar
gram-positive bacterium, is a normal inhabitant of diosis, but their presence in viscera only occurs in
the mouth and oropharynx. Although this organ actinomycosis.
ism and Nocardia asteroides are actually bacteria, they Gram's stains of the exudate should be carefully
are usually grouped with fungi because they cause examined for the presence of the slender, gram
chronic illness. positive, long-branching filaments characteristic of
actinomycosis (60) . The definitive diagnosis is estab
lished with the demonstration ofA. israelii by anaero
C l i n ical M a n ifestati o n s
bic cultures. The diagnosis of thoracic actinomycosis
Actinomycosis is characterized by the formation of cannot be established from cultures of expectorated
abscesses and multiple sinus tracts ( 5 5 ) . The infection sputum or bronchoscopic washings because A. israelii
arises from endogenous sources such as infected gums, can frequently be cultured from such specimens in
infected tonsils, or carious teeth ( 5 5 ) . The pleura is the absence of invasive disease. Bacterial culture of
involved in more than 50% of patients with thoracic the pleural fluid frequently reveals other organisms
actinomycosis. In one series of 1 5 cases of this dis in addition to A. israelii (58,60) . The organism most
order, 6 patients had pleural effusions, and an addi commonly isolated is Actinobacillus actinomycetem
tional 6 patients had marked pleural thickening (56) . comitans, a gram-negative aerobic coccobacillus (60) .
Th e pleura is particularly likely to b e thickened in It has been suggested that the aerobic actinobacillus
268 PLE U RAL D I S EASES
lymphocytes or eosinophils, the insertion of chest with positive sputum cultures for N asteroides, 9 of
tubes is not necessary. Alternately, if the pleural fluid the patients (45 %) did not have radiographic abnor
is frank pus, tube thoracostomy should be performed malities (68) .
(6 1 ) . Decortication is sometimes necessary for resolu
tion of the process ( 5 8 ) . Treatment
5 . Colp CR, Cook WA. Successful treatment of pleural aspergil 28. Young EJ, Hirsh DD, Fainstein V, et al. Pleural effusions
losis and bronchopleural fistula. Chest. 1 975;68:96-9 8 . due to Cryptococcus neoformans: a review of the literature and
6. Shiraishi Y, Katsuragi N, Nakajima Y, e t al. Pneumonectomy report of two cases with cryptococcal antigen determinations.
for complex aspergilloma: is it still dangerous? Eur J Cardio Am Rev Respir Dis. 1 9 80; 1 2 1 :743-747.
thorac Surg. 2006;29:9-l 3. 29. Chechani V, Kamholz SL. Pulmonary manifestations of dis
7. Walsh TJ, Bulkley BH. Aspergillus pericarditis: clinical and seminated cryptococcosis in patients with AIDS. Chest.
pathologic features in the immunocompromised patient. Can 1 990;98: 1 060- 1 06 5 .
cer. 1 982;49:48-54. 30. Cadranel J L , Chouaid C, Denis M, e t al. Causes o f pleural
8 . Murphy D, Lane DJ. Pleural effusion in allergic broncho effusion in 75 HIV-infected patients. Chest. 1 993; 1 04:65 5 .
pulmonary aspergillosis: two case reports. Br J Dis Chest. 3 1 . Salyer WR, Salyer D C . Pleural involvement i n cryptococcosis.
1 9 8 1 ;75:9 1 -9 5 . Chest. 1 974;66: 1 39-140.
9 . O'Connor T M , O'Donnell A , Hurley M, e t al. Allergic bron 32. Wasser L , Talavera W. Pulmonary cryptococcosis in AIDS.
chopulmonary aspergillosis: a rare cause of pleural effusion. Chest. 1 987;92:692-6 9 5 .
Respirology. 200 1 ;6:36 1-363. 33. Yoshino Y, Kitazawa T, Tatsuno K , et a l . Cryptococcal pleuritis
10. Weiner MH. Antigenemia detected by radioimmunoassay in containing a high level of adenosine deaminase in a patient
systemic aspergillosis. Ann Intern Med. 1 980;92:793-796. with AIDS : a case report. Respiration. 2 0 1 0;79: 1 5 3- 1 5 6.
1 1 . Metzger JB, Garagusi VF, Kerwin OM. Pulmonary oxa 34. Epstein R, Cole R, Hunt KK Jr. Pleural effusion secondary to
losis caused by Aspergillus niger. Am Rev Respir Dis. pulmonary cryptococcosis. Chest. 1 972;6 l :296-298.
1 984; 1 2 9 : 5 0 1 -502. 35. Duperval R, Hermans PE, Brewer N S , et al. Cryptococ
12. Klein NC, Cunha BA. New antifungal drugs for pulmonary cosis, with emphasis on the significance of isolation of
mycoses. Chest. 1 996; 1 1 0 : 525-532. Cryptococcus neoformans from the respiratory tract. Chest.
13. Herbrecht R, Denning OW, Patterson TF, et al. Voriconazole 1 977;72: 1 3- 1 9 .
versus amphotericin B for primary therapy of invasive asper 3 6 . Saag MS, Graybill RJ, Larsen RA , e t al. Practice guidelines for
gillosis. N Engl] Med. 2002;347:408-4 1 5 . the management of cryptococcal disease. Infectious diseases
1 4 . Shirakusa T, Ueda H , Saito T, e t al. Surgical treatment o f pul society of America. Clin Infect Dis. 2000;30:7 1 0-7 1 8 .
monary aspergilloma and Aspergillus empyema. Ann Thorac 37. Warr W, Bates J H , Stone A . Th e spectrum o f pulmonary cryp
Surg. 1 9 89;48 :779-782. tococcosis. Ann Intern Med. 1 968;69: 1 1 09-1 1 1 6.
1 5 . Matsuda T, Koreeda Y, Mataki H, et al. A case of Aspergil 3 8 . Tenholder MF, Ewald FW Jr, Khankhanian NK, et al. Com
lus empyema successfully treated with combination therapy plex cryptococcal empyema. Chest. 1 992; 1 0 1 : 5 8 6-5 8 8 .
of voriconazole and micafungin: excellent penetration of 39. Goodwin RA Jr, Des Prez RM . Histoplasmosis. A m Rev Respir
voriconazole and micafungin into pleural fluid. Intern Med. Dis. 1 978; 1 1 7:929-956.
20 1 0;49 : 1 1 63- 1 1 69 . 40. Connell JV Jr, Muhm JR. Radiographic manifestations
1 6. Blastomycosis Cooperative Study of the Veterans Administra of pulmonary histoplasmosis: a I 0-year review. Radiology.
tion. Blastomycosis. A review of 1 98 collected cases in Veterans 1 976; 1 2 1 :28 1 -2 8 5 .
Administration Hospitals. Am Rev Respir Dis. 1 964;89:659--072. 4 1 . Brewer P L , Himmelwright JP. Pleural effusion due t o infec
1 7 . Sheflin JR, Campbell JA, Thompson GP. Pulmonary blasto tion with Histoplasma capsulatum. Chest. 1 970; 5 8 :76-79.
mycosis: findings on chest radiographs in 63 patients. AJR Am 42. Schub HM, Spivey CG Jr, Baird GD. Pleural involvement in
J Roentgenol. 1 990; 1 54: 1 1 77- 1 1 80 . histoplasmosis. Am Rev Respir Dis. I 966;94:225-232 .
1 8 . Faila PJ, Cerise FP, Karam GH, e t al. Blastomycosis: pulmonary 43. Campbell GD, Webb WR. Eosinophilic pleural effusion. Am
and pleural manifestations. South MedJ l 9 9 5 ; 8 8 :405-4 1 0 . Rev Respir Dis. 1 964;90: 1 94-20 1 .
1 9 . Kinasewitz GT, Penn RL, George R B . Th e spectrum and 44. Weissbluch M . Pleural effusion in histoplasmosis. J Pediatr.
significance of pleural disease in blastomycosis. Chest. 1 976; 8 8 : 894-8 9 5 .
1 9 84; 86:5 80-5 84. 4 5 . Davies SF, Knox KS, Sarosi GA. Fungal infections. I n : Mason
20. Nelson 0, Light RW Granulomatous pleuritis secondary to RJ, Murray JF, Broaddus VC, et al. eds. Textbook of Respira
blastomycosis. Chest. 1 977;71 :433-434. tory Medicine, 4th ed. Philadelphia, PA: Elsevier, Saunders,
2 1 . Sarosi GA, Armstrong D, Davies SF, et al. Laboratory diag 2005 : 1 044- 1 082.
nosis of mycotic and specific fungal infections. Am Rev Respir 46. Quasney MW, Leggiadro RJ . Pleural effusion associated with
Dis. 1 9 8 5 ; 1 32: 1 373- 1 380. histoplasmosis. Pediatr Infect Dis] 1 9 93; 1 2 :4 1 5-4 1 8 .
22. Limper AH, Knox KS, Sarosi GA, et al. An official American 47. Ericsson CD, Pickering LK, Salmon GW. Pleural effusion in
Thoracic Society statement: treatment of fungal infections in histoplasmosis. J Pediatr. 1 977;90:326-327.
adult pulmonary and critical care patients. Am J Respir Crit 48. Kilburn CD, McKinsey OS. Recurrent massive pleural effu
Care Med. 2 0 1 1 ; 1 83:96- 1 2 8 . sion due to pleural, pericardia!, and epicardial fibrosis in his
23. Drutz DJ, Catanzaro A . Coccidioidomycosis. Am Rev Respir toplasmosis. Chest. 1 99 1 ; 1 00 : 1 7 1 5- 1 7 1 7.
Dis. 1 978; 1 1 7:727-77 1 . 49. Marshall BC, Cox JK Jr, Carroll KC, et al. Histoplasmosis as
24. Lonky SA, Catanzaro A, Moser KM , et al. Acute coccidioidal a cause of pleural effusion in the acquired immunodeficiency
pleural effusion. Am Rev Respir Dis. 1 976; 1 1 4:68 1-6 8 8 . syndrome. Am ] Med Sci. 1 990;300:98- 1 0 1 .
2 5 . Merchant M, Romero A O , Libke R D , e t a l . Pleural effusion 5 0 . Ankobiah WA, Vaidya K , Powell S, e t al. Disseminated
in hospitalized patients with coccidioidomycosis. Respir Med. histoplasmosis in AIDS. Clinicopathologic features in
2008;68 :4360-4368. seven patients from a non-endemic area. N Y State J Med.
26. Youssef SS, Ramu V, Sarubbi FA. Unusual case of pyopneu 1 990;90 :234-238.
mothorax in Tennessee. South MedJ 2005;98: 1 1 39-1 1 4 1 . 5 1 . Richardson JV, George RB. Bronchopleural fistula and lym
27. Cunningham RT, Einstein H . Coccidioidal pulmonary cavi phocytic empyema due to Histoplasma capsulatum. Chest.
ties with rupture. ] Thorac Cardiovasc Surg. 1 982;84: 1 72- 1 77. 1 997; 1 1 2: 1 1 30- 1 1 32.
270 PLE U RAL D I S EASES
52. Varghese JC, Hahn PF, Harisinghani MG, et al. Fungus 6 1 . McQuarrie DG, Hall WH . Acrinomycosis of the lung and
infected fluid collections in thorax or abdomen: effec chest wall. Surgery. l 968;64:905-9 1 1 .
tiveness of percutaneous catheter drainage. Radiology. 62. Neu HC, Silva M , Hazen E, et al. Necrotizing nocardial pneu
2005;236:730-73 8 . monitis. Ann Intern Med. 1 967;66:274-284.
5 3 . Ishiguro T, Takayanagi N, Ikeya T, e t al. Isolation of Can 63. Feigin DS. Nocardiosis of the lung: chest radiographic find
dida species is an important clue for suspecting gastrointes ings in 21 cases. Radiology. 1 986; 1 5 9:9- 1 4.
tinal tract perforation as a cause of empyema. Intern Med. 64. Uttamchandani RB, Daikos GL, Reyes RR, et al. Nocardio
20 1 0;49: 1 957- 1 964. sis in 30 patients with advanced human immunodeficiency
54. Pappas PG, Rex JH, Sobel JD, et al. Guidelines for treatment virus infection: clinical features and outcome. Clin Infect Dis.
of candidiasis. Clin Infect Dis. 2004;3 8 : 1 6 1 - 1 6 9 . 1 994; 1 8:348-3 5 3 .
5 5 . Peabody J W Jr, Seabury JH. Actinomycosis and nocar 6 5 . Rubin E , Shin MS. Pleural and exrrapleural disease i n nocar
diosis: a review of basic differences in therapy. Am J Med. dia infections. Can Assoc Radio!j. 1 984;3 5 : 1 89-1 9 1 .
l 960;28:99-1 1 5 . 66. Kramer MR, Uttamchandani RB. The radiographic appear
5 6 . Flynn MW, Felson B . Th e roentgen manifestations o f thoracic ance of pulmonary nocardiosis associated with AIDS. Chest.
actinomycosis. AJR Am J Roentgenol. 1 970; 1 1 0:707-7 1 6. 1 990;98:382-3 8 5 .
57. Kwong JS, Muller NL, Godwin JD, et al. Thoracic acti 6 7 . Palmer D L , Harvey RL, Wheeler J K . Diagnostic and ther
nomycosis: CT findings in eight patients. Radiology. apeutic considerations in Nocardia asteroides infection.
1 992; 1 83 : 1 89- 1 92. Medicine. 1 974;53:39 1 -40 1 .
58. Karetzky MS, Garvey JW. Empyema due to Actinomyces naes 68. Frazier AR, Rosenow E C III, Roberts GD. Nocardiosis: a
lundii. Chest. 1 974;65 :229-230. review of 25 cases occurring during 24 months. Mayo Clin
5 9 . Barker CS. Thoracic actinomycosis. Can Med Assoc j. Proc. 1 975;50:657-663.
1 9 54;7 1 : 332-334. 69. Green M, Avery RK. Nocardiosis. Am J Transplant.
60. Varkey B, Landis FB, Tang TT, et al. Thoracic actinomycosis: 2004;4(suppl 1 0) :47-50 .
dissemination to skin, subcutaneous tissue and muscle. Arch
Intern Med. 1 974; 1 34:689-693.
P l e u ra l E ff u s i o n D u e to Pa ra s i t i c I n fect i o n
27 1
272 PLE U RAL D I S EASES
The diagnosis of amebiasis should be considered The diagnosis is established by the characteristic
in all patients with right-sided pleural effusions for appearance of the pleural fluid and can be confirmed
which no other explanation is obvious. Ultrasonic by serologic tests for amebiasis. Ultrasound or CT
studies and computed tomography (CT) scan can scanning of the abdomen can delineate the extent of
demonstrate the hepatic abscess but cannot differ the intrahepatic disease and the presence or absence
entiate the pyogenic from amebic abscesses (7) . The of a subphrenic abscess. Patients with transdiaphrag
diagnosis is aided by the use of serologic tests. The matic rupture should be treated with the same drugs
sensitivity of antibody tests for E. histolytica is 95% for as patients with sympathetic pleural effusions due to
patients with extraintestinal amebiasis (8) . Serology is amebic hepatic abscess. Patients with transdiaphrag
limited as a diagnostic tool in highly endemic areas, as matic rupture should also undergo percutaneous
individuals will remain seropositive for years after an drainage of both the liver abscess and the collection
infection has cleared and seropositivity rates of greater of material in the pleural space. The drainage can be
than 25% may exist in some areas (8). Polymerase accomplished with small tubes ( 1 2 to 14 F) ( 1 0) .
chain reaction (PCR) techniques are being more fre Th e combination o f the drugs and the percutaneous
quently used for the diagnosis of amebiasis but there drainage tubes results in clinical cure in almost all
is yet no standardized commercially available test (8) . patients ( 1 0) .
Approximately one third o f patients with transh
epatic rupture also have a bacterial infection of their
Treatment
pleural space (4,9) . Such patients should be treated
The treatment of choice is metronidazole, 500 to 750 with the appropriate antibiotics. In addition, an
mg t.i.d. orally, for 10 days (8) . If the patient is dys open-drainage procedure or decortication is fre
pneic from the pleural effusion, a single therapeutic quently necessary, indications for which are outlined
thoracentesis is usually sufficient to control the symp in Chapter 1 2. In patients who undergo decortica
toms. More than 90% of patients can be cured with tion, the visceral pleura is found to be covered with
the aforementioned regimen. Catheter drainage adds a thick membrane (9), but this membrane can eas
no significant benefit to amebicidal therapy alone ( 1 ) . ily be stripped off the visceral pleura (9) . Even when
no bacterial superinfection is present, decortica
tion should be performed if the lung has not fully
Tra n sd i a p h ra g m atic Ru ptu re
expanded in 1 0 days (9) . The prognosis with trans
of Liver Abscess
diaphragmatic rupture is excellent if the patient is
The transdiaphragmatic rupture of an amebic liver not too debilitated initially or if the diagnosis is not
abscess is usually signaled by an abrupt exacerba delayed (3,4,9) .
tion of pain in the right upper quadrant and may be
accompanied by a tearing sensation (3) . These symp
E C H I N OCOCCO S I S ( H Y DATI D D I S EA S E )
toms are followed by the development of rapidly
progressive respiratory distress and sepsis, occasion Echinococcosis i s caused b y the tapeworm Echino
ally with shock (3) . The pleural effusion is frequently coccus granulosus. The definitive host for this small
massive, with opacification of the entire hemithorax tapeworm is the dog or wolf When dog feces con
and shift of the mediastinum to the contralateral taining the parasite's eggs are ingested by humans,
side (3) . The rupture is into the right pleural space larvae emerge in the duodenum, enter the blood, and
in more than 90% of patients. The symptoms are usually lodge in either the liver or the lung. In these
sometimes subacute or chronic in nature (9) . The tissues, the parasite grows gradually, and years may
diagnosis of amebic abscess with transdiaphragmatic pass before symptoms appear. It takes approximately
rupture is suggested by the discovery of anchovy paste 6 months for the cyst to reach a diameter of 1 cm and
or chocolate sauce pleural fluid on diagnostic thora thereafter it increases in size by 2 to 3 cm/year ( 1 1 ) .
centesis. Amebas can be demonstrated in the pleural Th e dog becomes infected by eating meat containing
fluid in fewer than 1 0% of patients. Concomitant the larvae. Echinococcosis is seen in most sheep- and
rupture into the airways occurs in approximately cattle-raising areas of the world including Australia,
30% of patients (3) , and this complication is usually New Zealand, Argentina, Uruguay, Chile, parts of
manifested by the expectoration of chocolate sauce Africa, Eastern Europe, and the Middle East. The
sputum that may be confused with hemoptysis by the disease is particularly common in Turkey, Lebanon,
patient and the physician. and Greece.
C H A PT E R 1 5 I P L E U RAL E F F U S I O N D U E TO PARASITIC I N F E CTI O N 273
form develops in the snails and is eventually liberated The characteristics of the pleural fluid are virtually
as cercariae that penetrate crayfish and crabs, to com pathognomonic for paragonimiasis. The pleural fluid
plete the cycle (29 ,30) . with paragonimiasis is an exudate with a low glucose
level (< 1 0 mg/dL) , a low pH (<7. 1 0) , and a high
Path ogenesis a n d I n cide nce lactic dehydrogenase (LDH) level (> 1 ,000 IU/L)
(29,34) . Pseudochylothorax with cholesterol crys
The pleural disease associated with paragonimiasis is
tals in the pleural fluid occurs frequently with pleu
thought to arise when the parasites traverse the pleural
ral paragonimiasis (37) . Most patients with pleural
space and penetrate the visceral pleura. Pleural disease
paragonimiasis have significant eosinophilia in their
is common with paragonimiasis (3 1 ,32) . In a series
pleural fluid (26,34,38 ,39) . In patients with pleural
of 71 cases of pleuropulmonary paragonimiasis from
paragonimiasis the pleural fluid interleukin 5 levels
Korea, 43 patients (6 1 %) had pleural disease includ
are markedly elevated and correlate with the percent
ing 26 with pleural effusions, 1 2 with hydrothorax,
age of eosinophils in the pleural fluid (39) . The only
and 5 with pleural thickening (32) . Twelve of these
other disease that produces an eosinophilic exudate
patients had bilateral pleural effusions or hydrothora
with a low glucose level and a low pH is the Churg
ces (32) . In another series from Japan, 9 of 1 3 patients
Strauss syndrome (40) .
( 69%) had a pleural effusion (3 1 ) . The prevalence of
Yokogawa et al. (26) reported that pleural fluid
pleural disease was less in a more recent study of 3 1
immunoglobulin E (Ig-E) levels are elevated and are
patients from Korea who underwent chest CT scans
higher than the simultaneous serum IgE levels in
(33) . In this study, only nine patients (29%) had
patients with pleural paragonimiasis. Subsequently
pleural disease that included six pleural effusions, two
Ikeda et al. (4 1 ) used ELISA to measure P. westermani
hydropneumothoraces, and one pneumothorax (3) .
specific IgE and IgG in seven patients. They reported
Although paragonimiasis is confined mainly
that the levels of parasite-specific IgE and IgG were
to residents of the Far East, there was an increased
significantly higher in the pleural effusion than in the
incidence of this disease in the United States in the
serum in all patients (4 1 ) . This latter study suggests
1 9 80s with the influx of refugees from Southeast
that measurement of parasite-specific IgE and IgG is
Asia (29,34) . Johnson and Johnson (34) reported
useful diagnostically and indicates that these antibod
a series of 25 cases of paragonimiasis that occurred
ies are produced in the pleural space.
in Indo-Chinese refugees between March 1 9 8 0 and
December 1 982. Seventeen cases occurred in Min
neapolis, whereas eight cases occurred in Seattle (34) . Treatment
Twelve of the patients (48%) had pleural effusions. The treatment of choice is praziquantel, 25 mg/kg
The effusions were bilateral in three individuals body weight three times a day for 3 days. If symptoms
and were massive in six. Non-Asian individuals in relapse, retreatment with praziquantel should be con
the United States have developed paragonimiasis sidered (42) . Bithionol, 35 to 50 mg/kg on alternate
from ingesting infected crayfish (35) or crabs (36) . days for 1 0 to 1 5 doses, is also effective, but its toxic
Indeed between July 2006 and September 20 1 0, gastrointestinal effects can be troublesome (4 1 ) . If
nine patients in Missouri developed paragonimiasis pleural disease has been present for such a prolonged
after eating raw or undercooked crayfish and all had period that the pleural surfaces are abnormally thick
a pleural effusion (28 ) . ened, penetration of the drugs into the pleural space
is insufficient to eradicate the infection, and thora
D i a g nosis cotomy with decortication may be necessary (29,43) .
The diagnosis of pleural paragonimiasis should be
suspected in Asian patients or in patients with a pleu
OTH E R PARAS IT I C I N F E CTI O N S
ral effusion who have recently traveled to the East.
The diagnosis of paragonimiasis is made either by Pleural disease due to other parasites is uncommon.
detecting eggs in sputum, stool, fluid from broncho A rare patient with Pneumocystis carinii pneumonia
scopic lavage, or biopsy specimens, or by a positive has a pleural effusion, but the effusion hardly ever
anti-Paragonimus antibody test (29) . Enzyme-linked dominates the clinical picture. Because P. carinii
immunosorbent assay (ELISA) is highly sensitive infection usually occurs in patients with acquired
and specific in detecting antibodies, whereas eggs are immunodeficiency syndrome, this entity is dis
demonstrable in less than 50% of cases (32) . cussed in Chapter 1 6. Patients who die from malaria
C H A PT E R 1 5 I P L E U RAL E F F U S I O N D U E TO PARASITIC I N F E CTI O N 275
frequently have pleural effusions (6) . The pleural effu 20. Jacobson ES. A case of secondary echinococcosis diagnosed
sions in this circumstance are probably secondary to by cyrologic examination of pleural fluid and needle biopsy of
pleura. Acta Cytol. 1 973; 1 7:76-79.
the pulmonary edema and have little, if any, clini 2 1 . GouliamosAD, KalovidourisA, PapailiouJ, et al. CT appearance
cal significance (5) . There have been case reports of of pulmonary hydatid disease. Chest. 1 9 9 1 ; 1 00: 1 578-1 5 8 1 .
pleural effusions due to Strongyloides sterocoralis (44) , 22. Kurkcuoglu IC, Eroglu A, Karaoglanoglu N , et al. Tension
Trichomonas (45 ) , Toxocara canis (46) , loiasis (47) , pneumothorax associated with hydatid cyst rupture. ] Thorac
gnathostomiasis (48) , anisakiasis (49) , toxoplasmosis Imaging. 2002; 1 7:78-80.
23. Men S, Hekimoglu B, Yucesoy C, et al. Percutaneous treat
(50), and sporotrichosis (5 1 ) . ment of hepatic hydatid cysts: an alternative to surgery. A]R
Am ] Roentgenol. 1 999; 1 72:83-89.
24. Paterson HS, Blyth DF. Thoracoscopic evacuation of dead
hydatid cyst. ] Thorac Cardiovasc Surg. 1 996; 1 1 1 : 1 280- 1 28 1 .
REFERENCES 2 5 . Drugs fo r parasitic infections. Med Lett. 1 99 8 ;40: 1 -7.
26. Yokogawa M, Kojima S, Araki K, et al. Immunoglobulin E:
I. Lyche KD , Jensen WA. Pleuropulmonary amebiasis. Semin raised levels in sera and pleural exudates of patients with para
Respir Infect. 1 997; 1 2 : 1 06-1 1 2 . gonimiasis. Am ] Trop Med Hyg. 1 976;2 5 : 5 8 1-586.
2 . Lyche KD, Jensen WA, Kirsch C M , e t al. Pleuropulmo 27. Devi KR, Narain K, Bhattacharya S, et al. Pleuropulmonary
nary manifestations of hepatic amebiasis. Wt-st J Med. paragonimiasis due ro Paragonimus heterotremus: molecular
1 990; 1 53:275-278. diagnosis, prevalence of infection and clinicoradiological fea
3 . Ibarra-Perez C. Thoracic complications of amebic abscess of tures in an endemic area of northeastern India. Trans R Soc
the liver. Chest. 1 9 8 1 ;79:672-676. Trop Med Hyg. 2007; 1 0 1 :786-792.
4. Cameron EWJ . The treatment of pleuropulmonary amebiasis 28. Centers for Disease Control and Prevention (CDC) . Human
with metronidazole. Chest. 1 978;73 :647-650. paragonimiasis after eating raw or undercooked crayfish
5 . Shamsuzzaman SM, Hashiguchi Y. Thoracic amebiasis. Clin Missouri, July 2006-September 2 0 1 0 . MMWR Morb Mortal
Chest Med. 2002;23:479-492. Wkly Rep. 20 1 0 ; 5 9 : 1 573- 1 576.
6. Sharma OP, Maheshwari A. Lung diseases in the tropics. 29. Minh V-D, Engle P, Greenwood JR, et al. Pleural para
Part 2 : common tropical lung diseases: diagnosis and manage gonimiasis in a southeast Asian refugee. Am Rev Respir Dis.
ment. Tuber Lung Dis. 1 993;74:359-370. 1 9 8 1 ; 1 24: 1 86- 1 8 8 .
7. Salles JM, Moraes LA, Salles MC. Hepatic amebiasis. Braz ] 3 0 . Eikas J, Kim PK. Clinical investigation o f paragonimiasis.
Infect Dis. 2003;7:96- 1 1 0. Acta Tuberc Scand. 1 960;39 : 1 40- 1 47.
8. Talaat KR, Nutman TB. Parasitic diseases. In: Mason RJ, 3 1 . Nawa Y. Recent trends of Paragonimiasis westermani in
Broaddus VC, Murray JF, et al. eds. Textbook of Respiratory Miyazaki Prefecture, Japan. Southeast Asian J Trop Med Public
Medicine, 4th ed. Philadelphia, PA: Elsevier, WB Saunders; Health. 1 9 9 1 ;22 (suppl 22) :342-344.
2005 : 1 083- 1 1 1 3 . 32. Im JG, Whang HY, Kim WS, et al. Pleuropulmonary para
9 . Rasaretnam R , Paul ATS, Yoganathan M . Pleural empy gonimiasis: radiologic findings in 71 patients. A]RAm J Roent
ema due ro ruptured amoebic liver abscess. Br J Surg. genol. 1 992; 1 5 9 :39-43 .
1 974;6 1 :7 1 3-7 1 5 . 33. Kim TS, Han J, Shim SS, et al. Pleuropulmonary paragoni
1 0 . Baijal SS, Agarwal DK, Roy S, e t al. Complex ruptured ame miasis: CT findings in 3 1 patients. A]R Am J Roentgenol.
bic liver abscesses: the role of percutaneous catheter drainage. 2005 ; 1 8 5 : 6 1 6-62 1 .
EurJ Radio!. 1 99 5 ;20:65-67. 34. Johnson RJ, Johnson JR. Paragonimiasis i n Indochinese refu
1 1 . von Sinner WN. Ultrasound, CT and MRI of ruptured and gees: roentgenographic findings with clinical correlations. Am
disseminated hydatid cysts. EurJ Radio!. 1 990; 1 1 :3 1 -37. Rev Respir Dis. 1 983; 1 2 8 : 5 34-5 3 8 .
12. Rakower J, Milwidsky H . Hydatid pleural disease. Am Rev 3 5 . Pachucki CT, Levandowski RA , Brown VA, e t al. American
Respir Dis. l 964;90:623-63 1 . paragonimiasis treated with praziquantel. N Engl ] Med.
1 3 . Barzilai A, Pollack S, Kaftori JK, et al. Splenic echinococcal cyst 1 9 84;3 1 1 : 5 82-5 84.
burrowing inro left pleural space. Chest. 1 977;72:543-545 . 36. Sharma OP. The man who loved drunken crabs: a case of pul
1 4 . Jerray M, Benzarti M, Garrouche A , e t al. Hydatid disease of monary paragonimiasis. Chest. 1 989;95:670-672.
the lung. Am Rev Respir Dis. 1 992; 1 46: 1 8 5- 1 8 9 . 37. Inoue Y, Kawaguchi T, Yoshida A, et al. Paragonimiasis miyaza
1 5 . von Sinner W Pleural complications of hydatid disease (Echino kii associated with bilateral pseudochylothorax. Intern Med.
coccus granulosus) . Rofo Fortschr Geb Rontgenstr Nuklearmedizin. 2000;39 :579-5 82.
1 990; 1 52 : 7 1 8-722. 38. Matsumoto N, Mukae H, Nakamura-Uchiyama F, et al. Elevated
16. Ozvaran MK, Ersoy Y, Uskul B, et al. Pleural complications of levels of thymus and activation-regulated chemokine (TARC) in
pulmonary hydatid disease. Respirology. 2004;9 : 1 1 5- 1 1 9 . pleural effusion samples from patients infested with Paragonimus
1 7 . Aktogu Ozkan S, Erer OF, A Yalcin Y, e t al. Hydatid cyst westermani. Clin Exp lmmunol. 2002; 1 30:3 1 4-3 1 8 .
presenting as an eosinophilic pleural effusion. Respirology. 39. Taniguchi H, Mukae H, Matsumoto N, et al. Elevated IL-5
2007; 1 2:462-464. levels in pleural fluid of patients with Paragonimiasis wester
1 8 . Keramidas D, Mavridis G, Souris M, et al. Medical treatment mani. Clin Exp lmmunol. 2000; 1 23 : 94-9 8 .
of pulmonary hydatidosis: complications and surgical man 4 0 . Erzurum S E , Underwood GA, Hamilos D L , e t a l . Pleural effu
agement. Pediatr Surg Int. 2004; 1 9 :774-776. sion in Churg-Strauss syndrome. Chest. 1 989;9 5 : 1 357- 1 3 5 9 .
1 9 . Xanthakis DS, Katsaras E, Efthimiadis M, et al. Hyda 4 1 . Ikeda T, Oikawa Y, Owhashi M, e t al. Parasite-specific IgE and
tid cyst of the liver with intrathoracic rupture. Thorax. IgG levels in the serum and pleural effusion of Paragonimiasis
1 9 8 1 ;36:497-5 0 1 . westermani patients. Am ] Trop Med Hyg. 1 992;47: 1 04- 1 07.
276 PLE U RAL D I S EASES
42. Oh IJ, Kim YI, Chi SY, et al. Can pleuropulmonary paragoni 47. Klion AD, Eisenstein EM, Smirniotopoulos TT, et al.
miasis be cured by only the I st set of chemotherapy? Treatment Pulmonary involvement in loiasis. Am Rev Respir Dis.
outcome and clinical features of recently developed pleuropul 1 992; 1 4 5 : 9 6 1 -963.
monary paragonimiasis. Intern Med. 20 1 1 ; 50: 1 365-1 370. 48. Parola P, Bordmann G , Brouqui P, et al. Eosinophilic
43. Dietrick RB, Sade RM, Pak JS. Results of decortication in pleural effusion in gnathostomiasis. Emerg Infect Dis.
chronic empyema with special reference to paragonimiasis. 2004; 1 0 : 1 690- 1 6 9 1 .
J Thorac Cardiovasc Surg. 1 98 1 ; 8 2 : 5 8-62. 49. Saito W, Kawakami K, Kuroki R, et al. Pulmonary anisa
44. Emad A. Exudative eosinophilic pleural effusion due to kiasis presenting as eosinophilic pleural effusion. Respirology.
Strongyloides stercoralis in a diabetic man. South Med ]. 2005; 1 0:26 1 -262.
1 999;92: 5 8-60. 5 0 . Collet G, Marty P, Le Fichoux Y, et al. Pleural effusion as the
45. Lewis KL, Doherty DE, Ribes J, et al. Empyema caused by first manifestation of pulmonary toxoplasmosis in a bone mar
trichomonas. Chest. 2003 ; 1 23:29 1 -292. row transplant recipient. Acta Cytol. 2004;48 : 1 1 4- 1 1 6 .
46. Jeanfaivre T, Cimon B, Tolstuchow N, et al. Pleural effusion 5 1 . Morrissey R , Caso R . Pleural sporotrichosis. Chest.
and toxocariasis. Thorax. 1 996;5 1 : 1 06- 1 07. 1 983;84:507.
P l e u r a l E ff u s i o n D u e t o A c q u i r e d
I m m u n o d ef i c i e n cy Sy n d ro m e, Ot h e r V i r u ses,
Mycoplasma Pn eumonia e, a n d R i c ketts i a e
Over the last three decades, the acquired immuno Kaposi's sarcoma (KS) (2) , 1 8% had aerobic bacterial
deficiency syndrome (AIDS) epidemic has had a infections, 1 5 % had tuberculosis, 1 0% had opportu
profound impact on the practice of medicine. Accord nistic infections, and 5% had effusions due to other
ingly, the first part of this chapter deals with pleural malignancies (2) . In a third series from Rwanda (3) ,
effusions in patients with AIDS. Because the organ tuberculosis was responsible for the pleural effusion
ism responsible for AIDS is a virus, pleural diseases in 82 of 9 1 (90%) of patients who were HIV positive.
due to viruses are included in this chapter. Addition
ally, the pleural effusions resulting from infection with Ka posi's S a rco m a
Mycoplasma pneumoniae, ehrlichia, and the rickettsial
KS is o ne of the more common causes of a pleural
diseases, Q fever, and Rocky Mountain spotted fever
effusion in patients with AIDS . KS occurs almost
are discussed in this chapter because they produce
exclusively in male homosexual patients with AIDS
clinical pictures simulating viral pneumonias.
(4) . Human herpesvirus 8 (HHV-8) is associated
with the development of this malignancy. Although
HHV-8 is necessary, additional factors not yet fully
P L E U RA L E F F U S I O N S I N
delineated need to be present for a person to develop
PAT I E NTS WITH A I D S
KS (5,6) . It is thought that HHV-8 is particularly
Pleural effusions are not uncommon i n patients with likely to be spread by male homosexual contacts. In
AIDS . In one series of 1 ,225 consecutive hospital the past, KS occurred in 20% to 25% of individu
admissions of patients with AIDS in Jacksonville, als with AIDS (4) . In recent years, there has been a
Florida, the incidence of pleural effusion was 1 4.6% ( 1 ) . precipitous decline in the prevalence of KS (5). For
I n a n older series from Metropolitan Hospital Cen example, in the state of Washington, the prevalence
ter in New York, the incidence was 1 .7% in a series of KS in the homosexual HIV-positive population
of 4, 5 1 1 hospitalized human immunodeficiency was 6% in 1 990, but had fallen to 2% by 1 997 and
virus (HIV) -positive patients ( 1 ) . The distribu apparently has fallen more in the next 3 years (5) . It is
tion of the diseases responsible for pleural effusions likely that the decrease in the prevalence of KS is, at
in patients with AIDS varies widely from series to least in part, related to the introduction of the highly
series. In the series of 1 60 patients from Jacksonville, active antiretroviral therapy (HAART) (5) .
the five leading causes of pleural effusions were pneu Cutaneous violaceous plaques are the most com
monia and empyema (33%), renal failure (9%) , mon presentation of KS . At autopsy, 50% to 75% of
hypoalbuminemia (8%), tuberculosis (6%), and patients with cutaneous KS have pulmonary involve
pancreatitis (4%) ( 1 ) . In a second, older series of 6 1 ment (5), but clinically apparent pulmonary involve
patients from Paris, 52% o f the effusions were due to ment is less common during life (5) . Most patients
277
278 PLE U RAL D I S EASES
with pleuropulmonary KS present with progressive architectural appearance and not a particular neoplas
shortness of breath, nonproductive cough, and fever. tic cell type ( 1 2) . Mesothelial cells from patients with
Patients with pulmonary KS generally have abnormal KS are infected with the HHV-8 virus ( 1 3) . Needle
chest roentgenograms characterized by bilateral infil biopsy of the pleura does not establish the diagnosis
trates (4) . The incidence of pleural effusion with pul of KS because the parietal pleura is not involved (4) .
monary KS is approximately 50% (4) . Most patients It is probable that the diagnosis could be established
with a pleural effusion due to KS also have bilateral with thoracoscopy given the characteristic appearance
parenchymal infiltrates (4) . The pleural effusions may of the KS lesions on the visceral pleura (4) .
be unilateral or bilateral. Bilateral pleural effusions, The prognosis of a patient with pleuropulmo
focal airspace consolidation, intrapulmonary nodules nary KS is poor. In one study, the average interval
and/or hilar adenopathy are very suggestive of KS (7) . from diagnosis of pulmonary KS to death was 4 ±
The computed tomography (CT) scans are somewhat 3 months (4) . The presence of the pleural effusion is
characteristic with KS . In a review of 53 patients with a significant problem for many patients with pleu
pulmonary KS, 42 patients (79%) had nodules, 3 5 ropulmonary KS . In one series, recurrent, massive,
(66%) had bronchovascular bundle thickening, 28 progressive effusions dominated the final days of a
(53%) had tumoral masses, and 29 ( 5 5 %) had pleu substantial percentage of patients and contributed
ral effusions (8) . The effusions were bilateral in 40 significantly to the death of approximately 50% (4) .
(76%) and were usually at least medium sized (8) . The treatment of the pleural effusion associated
Autopsy studies demonstrate multiple cherry red to with KS is difficult. Tube thoracostomy with the
purple lesions on the visceral surface but not on the instillation of tetracycline is usually not successful
parietal pleural surface (4) . (4) . If the diagnosis is made with thoracoscopy, pleu
The pathogenesis of the pleural effusions in KS is ral abrasion or parietal pleurectomy is probably the
not conclusively defined. Approximately 20% of the treatment of choice. Otherwise, the best alternatives
pleural effusions with KS are chylothoraces, and, in are probably the insertion of a pleuroperitoneal shunt
these instances, the pleural effusion is probably due to or an indwelling catheter (e.g., PleurX) .
involvement of the thoracic duct by the sarcoma (4) . In
the remaining patients, lymphatic blockade is probably
Pri m a ry Effusion Lym p h o m a
not the responsible mechanism because the lymphatic
drainage of the pleura is through the parietal pleura This rare lymphoma occurs almost exclusively in
and KS does not involve the parietal pleura. It has been patients with HIV infection and is discussed in
hypothesized that the effusion is due to the elabora Chapter 1 1 .
tion of vascular endothelial growth factor (VEGF)
by the tumor (9) . VEGF increases the permeability
Pa ra p n e u m o n i c Effu s i o n a n d
of the microvessels and is present in large quantities in
E m pye m a a n d A I D S
AIDS-KS cell-derived conditioned media ( 1 0) .
Th e diagnosis o f pulmonary KS i s usually estab Community-acquired bacterial pneumonia (CAP)
lished at bronchoscopy, which demonstrates ery occurs frequently in patients with AIDS . It appears
thematous or violaceous macules or papules in the that patients with AIDS are probably more likely to
respiratory tree ( 1 1 ) . It is important to remember that develop pleural complications with their pneumo
many patients with pulmonary KS have a coexisting nias than are other patients because they are more
opportunistic infection (5) . If the CT scan reveals likely to have bacteremia with their pneumonia ( 1 4) .
ground-glass opacities, alternative diagnoses must be Moreover, once there are bacteria in the pleural space,
sought (8) . The definitive diagnosis of pleural KS is patients with low CD4+ counts probably have more
not easy and is virtually one of exclusion. The pleu trouble clearing the bacteria. When Staphylococcus
ral fluid is an exudate that is usually serosanguineous aureus bacteria are injected intrapleurally in CD4+
or hemorrhagic. The differential cell count shows a knockout mice, there is a decreased pleural chemokine
mononuclear cell-predominant pattern (4) . In one response, decreased neutrophil influx into the pleural
series of 1 0 patients, the pleural fluid glucose and space, and impaired bacterial clearance from the pleu
pH levels were normal in 9 patients but were reduced ral fluid ( 1 5) . In one series of 8 1 cases of community
to 63 mg/dL and 7.02, respectively, in 1 patient (4) . acquired pneumonia, pleural effusion occurred in 2 1
Cytologic examination of the pleural fluid is not (26%) patients ( 1 4) . Th e pleural fluid was culture pos
helpful because the diagnosis requires a characteristic itive in 1 1 of these 21 (52%) patients. However, in a
C H A PT E R 1 6 I PLE U RAL E F F U S I O N D U E TO A I D S OTH E R V I R U S E S 279
more recent study of 1 ,4 1 5 patients hospitalized with patients with and without AIDS is similar, but there
HIV-associated CAP from 1 99 5 to 1 997 at 86 hos are some differences. In patients with tuberculosis,
pitals in seven metropolitan areas, the prevalence of some reports have demonstrated that a higher per
pleural effusion was only 7.8% and the presence of a centage of patients with AIDS have a pleural effu
pleural effusion was not associated with a higher mor sion (20), whereas others report a similar incidence
tality ( 1 6) . The distribution of organisms responsible (2 1 ) . The percentage of cases of tuberculosis that have
for CAP in AIDS is similar to that of patients without pleural effusions in patients with AIDS is higher in
AIDS ( 1 4) . The management ofthe patient wirh AIDS patients with CD4 + counts above 200 than in those
and a parapneumonic effusion or an empyema is the with CD4+ counts below 200/mm3 (22) .
same as that for any patient with a parapneumonic The purified protein derivative (PPD) skin test is
effusion (see Chapter 1 2 ) . HIV-positive patients with less frequently positive in patients with AIDS who
empyema and CD4+ counts less rhan 200/mm3 more have tuberculous pleuritis. In one series of patients
commonly have complex empyemas that require open with tuberculous pleuritis, the PPD was positive in
decortication and drainage ( 1 7) . 76% of patients without AIDS but in only 4 1 % of
patients with AIDS (23) . The lower the CD4 + count,
the less likely the PPD is to be positive. HIV-positive
P n e u m ocysti s j i roveci P n e u m o n i a a n d A I D S
patients have a lower percentage of CD4+ lympho
Although pleural effusions due t o Pneumocystis cytes and a higher percentage of CDs + lymphocytes
jiroveci account for only a small percentage of pleural in their pleural fluid than do HIV-negative patients
effusions in patients with AIDS , they do occur. By with tuberculous pleuritis (24) . In patients without
1 993, a total of seven cases of pleural effusion due to AIDS, the pleural fluid acid-fast bacilli (AFB) stain is
P. jiroveci infection had been reported ( 1 8 , 1 9) . In only rarely positive (- 1 %), but in one series, the AFB
most cases, the diagnosis was established by visual pleural fluid smear was positive in 1 5 % of patients
ization of Pneumocystis in pleural fluid stained with with AIDS (23) . In another series, the pleural fluid
Gomori methenamine silver. All seven of the reported smear was positive in 37% of patients with AIDS and
patients were receiving aerosolized pentamidine, and a CD4+ count less than 200/mm3 (2 5 ) . The pleural
five of the seven had documented underlying P.jiroveci fluid cultures for AFB are more likely to be positive
pneumonia. Two patients presented with primary in patients with AIDS . In one study, the pleural fluid
pleural infection with Pneumocystis. It appears that culture was positive in 75% and 24% of the HIV
Pneumocystis pleural disease is an anatomic extension positive and HIV-negative patients with BACTEC,
of smoldering subpleural Pneumocystis pneumonia, respectively; with Lowenstein-Jensen medium, com
and the prognosis is not worse than with pneumonia parable numbers were 43% and 1 2%, respectively
alone. Four of the seven patients with pleural Pneu (26) . The levels of pleural fluid interferon-gamma are
mocystis also had a bronchopleural fistula ( 1 8) . higher in HIV-positive than in HIV-negative patients
Th e pleural fluid is an exudate with pleural Pneu (24) . The granuloma on pleural biopsy are less well
mocystis. The pleural fluid lactate dehydrogenase formed in some patients with AIDS, and there are
(LDH) has been higher than 400 IU/L, and the ratio numerous AFB (27) . Patients with poorly defined
of the pleural fluid to serum LDH level has exceeded granuloma appear to respond less well to antituber
1 : 0 . Interestingly, the pleural fluid protein level has culous therapy (27) . The incidence of granuloma on
been below 3 . 0 g/dL, and the ratio of the pleural pleural biopsy is comparable in patients with and
fluid to the serum protein has been below 0 : 5 0 in without AIDS (23) . The treatment of the patient with
all patients. The pleural fluid glucose and pH levels AIDS and tuberculous pleuritis is the same as the
are not reduced, and the differential cell count can treatment of the HIV-negative individual. However,
reveal either neutrophils or mononuclear cells ( 1 8) . prednisone should not be given to the AIDS patient
Th e treatment o f pleural Pneumocystis is the same as with tuberculous pleuritis because its administration
that of pulmonary Pneumocystis. is associated with a risk of developing KS (28) .
atypical mycobacteria (32) are at times responsible phases of disease are not usually obtained for diag
for a pleural effusion in patients with AID S . In their nosis. Moreover, most hospitals are not equipped to
terminal stages, some patients with AIDS develop culture viruses. In recent years, the diagnosis of more
hypoproteinemia and this may lead to a transudative and more viral infections have been made by PCR
pleural effusion (33 ) . Patients with AIDS may also (3 5 ) . In one study (35) of patients with a positive
develop hypervolemia owing to heart failure or renal PCR and symptoms of a lower respiratory tract infec
failure, which can lead to a pleural effusion (34) . In tion, 1 9 of 9 1 patients (20%) had a pleural effusion.
one series, pancreatitis was the fifth leading cause of The most interesting epidemic of pleural effu
pleural effusion in patients with AIDS ( 1 ) . The diag sions attributed to viral infection occurred in Turkey
nosis and management of pleural effusions due to in 1 9 5 5 , when 5 5 9 individuals at a military base
these different entities are described in the appropri developed a pleural effusion in conj unction with an
ate chapters in this book. acute illness characterized by fever, cough, malaise,
anorexia, and shortness of breath (36) . None of the
A p p roach to the Patient with A I D S patients had parenchymal infiltrates, but approxi
a n d P l e u ra l Effusion mately 30% had an enlarged hilar shadow. The
peripheral white blood cell (WBC) count was nor
Patients with AIDS and pleural effusion should mal or reduced, with an increased percentage of lym
undergo a diagnostic thoracentesis. Studies on the phocytes. The differential WBC on the pleural fluid
fluid should include smears and cultures for bacte revealed mostly mononuclear cells. The disease was
ria, mycobacteria, and fungi; cytology with special self-limited, and almost all patients recovered com
consideration for primary effusion lymphoma; and pletely within 90 days. Because all bacterial cultures
either an interferon-gamma or an adenosine deami were negative, as were serologic tests for Q fever, and
nase measurement for pleural tuberculosis. If the because the patients recovered without any specific
patient has a positive PPD ( > 5 mm) or if there are no therapy, it was concluded that the disease was due to a
mesothelial cells in the pleural fluid, chemotherapy viral infection (36) . This report is important because
with isoniazid and rifampin for 9 months is recom it documents that viral infections can cause pleural
mended. If the patient is receiving aerosolized pen effusions and in large numbers. One wonders what
tamidine, silver stains of the pleural fluid should be fraction of undiagnosed pleural effusions is due to
obtained to rule out P jiroveci. viral infections.
If the diagnosis is not apparent after the thoracen Small pleural effusions frequently accompany
tesis and the patient has an exudative pleural effu primary atypical pneumonia. Fine et al. (37) pro
sion, what should be the next diagnostic procedure? spectively studied 59 patients with atypical pneu
Possible courses of action include a needle biopsy of monia that satisfied serologic criteria for association
the pleura, thoracoscopy, bronchoscopy, or observa with either a mycoplasma, viral, or cold-agglutinin
tion. In general, a thoracoscopy is recommended, if positive pneumonia. Twelve of these patients (20%)
any procedure is going to be done. With thoracos had small pleural effusions, and in four patients the
copy one can establish the diagnosis of KS, other effusions were evident only on the lateral decubi
intrathoracic malignancies, tuberculous pleuritis, or tus radiographs. This finding compares with a 45%
other opportunistic pleural infections. In addition, incidence of pleural effusions in patients with acute
a pleural abrasion or a partial parietal pleurectomy bacterial pneumonia (38) . In the series of Fine et al.
can be performed to prevent reaccumulation of the (37) , 6 of 29 (2 1 %) patients with M. pneumonia, 1
pleural fluid. of 7 ( 1 4%) with adenoviral pneumonia, 1 of 4 (25%)
with influenza pneumonia, and 4 of 19 (2 1 %) with
only increased titers of cold agglutinins had pleural
VIRUSES
effusions.
Viral infections probably account fo r a larger per In patients with viral infections, the pleural effu
centage of pleural effusions than is generally realized. sions are usually small (36,37) , but may occasionally
The diagnosis usually depends on isolation of the be large (39) . The pleural fluid is an exudate (37), and
virus or the demonstration of a significant increase usually mononuclear cells are predominant on the
in the antibodies to the virus. Because most pleural pleural fluid differential WBC (36,40) . I have seen
effusions secondary to viruses are self-limiting, paired a patient with a viral pneumonia, however, in whom
sera from patients in the acute and convalescent the initial thoracentesis revealed predominantly
C H A PT E R 1 6 I PLE U RAL E F F U S I O N D U E TO A I D S OTH E R V I R U S E S 28 1
polymorphonuclear leukocytes, but a subsequent shows bibasilar infiltrates, which rapidly spread to
thoracentesis 48 hours later revealed predominantly include all four quadrants of the lung. The heart size
mononuclear cells. The diagnosis of pleural effusions is normal. Patients with the hantavirus pulmonary
secondary to viral infections is established by docu syndrome tend to decompensate rapidly with refrac
menting increasing titers with the specific serologic tory hypoxemia and hypotension. The mean duration
tests or by culturing viruses from the pleural fluid of hospitalization before death is only approximately
(40,4 1 ) . At times, with pleural effusions secondary 3 days.
to herpes infections or cytomegalovirus, the cytologic Pleural effusions were common in the 23 patients
findings in the pleural fluid, consisting of intranuclear seen at the University of New Mexico Hospital (44) .
inclusions and multinuclear giant cells with gelati At the time of admission, 2 1 of the 23 patients had
nous nuclear changes, suggest the diagnosis (42,43) . pleural effusions. Effusions of similar size were pres
ent bilaterally in most patients. The maximum size of
the effusion was achieved within 24 to 48 hours of
H a ntav i r u s I nfections admission. Three of the patients had effusions of suf
The hantavirus pulmonary syndrome is due to infec ficient size that chest tubes were inserted. A thoracen
tion with a previously unknown hantavirus species tesis was performed in four of the patients. Although
now called Sin Nombre virus. As of May 3 1 , 1 996, none of the fluids was very inflammatory, all four met
1 39 cases had been confirmed from 24 states, repre exudative criteria with a pleural fluid LDH greater
senting all regions of the United States, with a mor than two thirds of the upper normal serum limit.
tality rate of 49.6% (44) . An additional 1 2 cases had However, the pleural fluid protein level was below
been reported from Canada (44) . Most cases have 2 . 5 g/dL in three of the four patients, and the pleural
occurred in the Four Corners Region, where New fluid WBC was below 200/mm3 in all patients (44) .
Mexico, Arizona, Colorado, and Utah meet. The deer At autopsy, patients dying of the hantavirus pulmo
mouse, Peromyscus maniculatus, has been identified as nary syndrome have large serous effusions with severe
the likely principal reservoir of the Sin Nombre virus edema of the lungs. It is probable that the pleural
(45 ,46) . There are several other hantaviruses that can effusion results from interstitial fluid traversing the
produce a similar syndrome including the New York visceral pleura to the pleural space.
virus, which produces disease in New York State; the It is important to make the diagnosis of the hanta
Bayou virus, which produces disease in Louisiana; virus pulmonary syndrome early because antiviral
and the Black Creek Canal virus, which produces dis therapy requires time to be beneficial. The diagno
ease in Florida and southeastern United States (47) . sis can be established by serologic tests for immu
Hantaviruses have also been recognized in several noglobulin M (Ig-M) and IgG antibody, which are
locations in South America (47) . usually demonstrated by enzyme-linked immuno
The hantavirus pulmonary syndrome is charac sorbent assay (ELISA) (47) . The broad-spectrum
terized by a brief prodromal illness, followed by rap antiviral agent ribavirin is active against hantavirus
idly progressive, noncardiogenic pulmonary edema in vitro, but an open-label trial of intravenous riba
(45 ,46) . The median age of infected individuals is virin in patients with the hantavirus syndrome was
approximately 30 years, and 50% of the patients have inconclusive (47) .
been Native American Indians (45 ) . Most patients An animal model of the hantavirus syndrome
present with fever or chills, and gastrointestinal com has been described. When adult Syrian hamsters
plaints such as nausea or vomiting, abdominal pain, are exposed to the Andes virus (ANDV) , a South
or diarrhea. Most patients report myalgias and cough. American hantavirus, they develop a syndrome very
Dyspnea tends to be a late-developing symptom, similar to the hantavirus syndrome (48) . Animals that
occurring j ust before respiratory decompensation. died had a large volume (3-5 mL) pleural effusion.
Patients who present with the hantavirus pulmo
nary syndrome have many abnormal laboratory tests.
Ad e n ovirus P n e u m o n i a
They characteristically have the triad of thrombocy
topenia, a left shift in the myeloid series, and large After M. pneumoniae, adenoviruses are the second
immunoblastoid lymphocytes. The Pao / Fio 2 is usu leading cause of primary atypical pneumonia. Pleu
ally severely reduced, and 50% of the patients require ral effusions occur in 1 5 % to 62% of patients with
mechanical ventilation. The chest radiograph of adenoviral pneumonia (49 ,50 ). The pleural effusions
patients who progress to respiratory failure initially are usually bilateral, and most are moderate to large
282 PLE U RAL D I S EASES
in size. When adenovirus infection results in a pleural patient had a moderate-sized, left-sided pleural effu
effusion, a concomitant parenchymal infiltrate is usu sion without any parenchymal infiltrates (59) . The
ally present (49,50). pleural effusions are exudates and usually take several
months to resolve (60,6 1 ) .
I nfecti o u s H e patitis
D e n g u e Hemorrh a g i c Feve r
Pleural effusions occasionally occur in conj unction
with infectious hepatitis and at times precede the Dengue fever is caused by four antigenically distinct
development of icterus (40, 5 1 -5 5 ) . In a review of dengue viruses, and the disease is transmitted to
2,500 patients with viral hepatitis, 4 patients (0. 1 6%) human beings by mosquitoes. Classic dengue fever is
had pleural effusions (5 1 ) . In another prospective not uncommon among travelers to tropical areas but
study of 1 5 6 patients with hepatitis, however, 70% dengue hemorrhagic fever is rare. The characteristics
of the patients had at least a small pleural effusion of dengue hemorrhagic fever are increased capillary
(5 5 ) . Patients with pleural effusions secondary to viral permeability with leakage of plasma and abnormal
hepatitis do not have parenchymal infiltrates. The hemostasis (62) . In one study (63) involving 363
pleural Buid is an exudate with predominantly mono patients with dengue hemorrhagic, 25% had small
nuclear cells (40,56). The pleural effusion frequently effusion, 4 . 5 % had moderate effusion, and 1 .9% had
resolves before the hepatitis (52) . One must be careful massive effusion. Thoracentesis was performed in
in handling pleural Buid when infectious hepatitis is seven patients and all were transudates. The effusions
suspected because the infectious hepatitis B e antigen are usually bilateral but sometimes are right sided
has been demonstrated in pleural Buid secondary to only, and rarely, if ever, left sided only (64) . It appears
hepatitis (54,56) . that the pathogenesis of the effusions is similar to that
with the hantavirus syndrome, namely, a systemic
increase in the permeability of capillaries induced by
E pste i n - B a rr V i r u s a n d I nfect i o u s cytokines (63) . In one study, the interleukin 8 (IL-8)
M o n o n ucleosis levels in the pleural Buid were very high (65) .
It appears that infection with the Epstein-Barr virus
(EBV) can cause pleural effusions. The EBV infects
I nfl u e nza V i ruses
more than 90% of the population worldwide and is
able to establish a lifelong latent infection with inter Pleural effusions can occur in patients with influ
mittent reactivation to lyric replication (57) . Primary enza infections (37,66,67) . Pleural effusions appear
infection usually occurs subclinically in infancy and to be particularly common with avian influenza
childhood. Thijsen et al. (58) reported that the poly A (H 5N l ) . In one report (66) , 17 of 1 9 patients
merase chain reaction (PCR) for the EBV was posi (89%) of patients with H 5 N l had a pleural effusion.
tive in 24 of 60 (40%) of pleural Buids including 20 Thoracentesis was done in nine patients and showed a
of 34 (59%) with no etiology. The PCR in the serum variable WBC and differential (66) . Pleural effusions
was negative in 12 of the 1 8 Buids in which the pleu also occur in patients with influenza A (H l N l ) also
ral Buid PCR was positive. The pleural Buid PCR was known as swine Bu (67) . In one series of 42 patients
also positive in pleural Buids from 1 5 % of patients (67) , 1 2 had abnormal chest radiographs and pleu
with another clear diagnosis (57). The possibility ral effusions were present in 7 (58%) . The pleural
that EBV DNA came from latently infected B cells effusion does not seem to be a big problem in these
present in the Buid rather than from lytic replication patients.
could not be excluded. It is possible that many undi
agnosed pleural effusions are due to EBV
Other Viral I nfections
Infectious mononucleosis is due to infection with
the EBV, and pleural effusions occasionally occur Pleural effusions have also been reported to result
in the course of infectious mononucleosis (59-6 1 ) . from infection with respiratory syncytial virus
Lander and Palayew (59) reviewed the chest radio (68,69), measles after the administration of inacti
graphs of 59 patients with infectious mononucleosis vated virus vaccine (70), herpes simplex virus (7 1 ) ,
and reported that 3 (5%) had pleural effusions. Two Lassa fever virus (39), and HHV-6 associated pleurisy
of the patients had bilateral interstitial infiltrates and after hematopoietic stem cell transplantation (72) .
small bilateral pleural effusions, whereas the third Pleural effusions probably result from infection by
C H A PT E R 1 6 I PLE U RAL E F F U S I O N D U E TO A I D S OTH E R V I R U S E S 2 83
many other viruses as well. However, pleural effu In another review, 5 of 25 patients (20%) with chest
sions appear to be distinctly uncommon with the radiographic abnormalities due to Q fever had a pleu
severe acute respiratory syndrome (SARS) . In one ral effusion, and in one of these patients the effusion
study of 1 0 8 patients, no pleural effusions were was large (80) . The pleural fluid is an exudate, and
identified (73 ) . the differential reveals predominantly mononuclear
cells (8 1 ) or eosinophils (82) . In one report, the
Mycoplasma Pneumoniae pleural fluid adenosine deaminase level was increased
This organism is actually a small bacterium rather to 64 IU/L with Q fever (8 1 ) . The diagnosis is usu
than a virus. It is included in this chapter because ally established by demonstrating a fourfold increase
the disease it produces more closely resembles a in the antibody titers in the patient's serum. This
viral than a bacterial disease. Pleural effusions occur increase becomes apparent in most patients within
in 5% to 20% of patients with pneumonias due to 2 weeks of the onset of the illness. The treatment of
M. pneumoniae (37,74) . The effusions are usually small choice is tetracycline or doxycycline, which appear to
(37) but can be large (74-77) . In one series of 1 0 be superior to erythromycin (79).
patients with pneumonia and pleural effusion due to
M. pneumoniae, 4 of the patients' pleural fluid were Rocky M o u nta i n S potted Feve r
positive by PCR for M. pneumoniae DNA (77) . The
clinical courses of the patients with a positive pleu Rocky Mountain spotted fever is due to Rickettsia
ral fluid PCR were more prolonged than those of rickettsii, and humans acquire the infection after a tick
patients with negative pleural fluid PCR (77) . The bite. Most infections occur in the southeastern and
diagnosis, suggested by increased titers of cold agglu coastal Atlantic states. Classically, Rocky Mountain
tinins, is established by increasing specific antibody spotted fever is manifested by the triad of fever, rash,
titers. It may take several weeks for a fourfold rise and a history of tick exposure. The usual onset of ill
in specific antibody titer to become evident. In an ness is 5 to 7 days after the tick bite. Fever, malaise,
occasional patient, the diagnosis can be established by frontal headache, myalgia, and vomiting are com
isolating M. pneumoniae from the pleural fluid (78) . mon. A pleural effusion is present in 1 0% to 36%,
The treatment of choice is tetracycline or erythromy and a pulmonary infiltrate is present in a comparable
cin administration. No specific treatment need be percentage (83) . The infiltrates are probably due to
directed toward the pleural effusion, but a diagnostic vasculitis with increased permeability of the blood
thoracentesis should be performed to ensure that a vessels. The pleural fluid probably develops from a
complicated parapneumonic effusion is not present. capillary leak syndrome (84) . The pleural fluid can
be either a transudate or an exudate, but when it is an
exudate the LDH and protein levels are relatively low
R I C K ETTS IAE (84) . The treatment of choice is doxycycline 200 mg/day
in two divided doses.
Q Feve r
nausea, vomiting, and anorexia. Pulmonary infiltrates 16. Arozullah AM , Parada J, Bennett CL, er al. A rapid staging
develop in approximately 50% of patients (83) and system for predicting mortality from HIV-associated commu
nity-acquired pneumonia. Chest. 2003; 1 2 3 : 1 1 5 1 - 1 1 60 .
at least some patients develop a pleural effusion (8 5 ) . 1 7 . Khwaja S, Rosenbaum DH, Paul MC, et al. Surgical treat
Frequently the chest radiograph i s normal initially, ment of thoracic empyema in HIV-infected patients: severity
but the condition then progresses rapidly with the and treatment modality is associated with CD4 count status.
development of bilateral pulmonary infiltrates and Chest. 2005; 1 28:246-249.
1 8 . Horowitz ML, Schiff M, Samuels J, et al. Pneumocystis carinii
pleural effusions. It is probable that the infiltrates
pleural effusion. Pathogenesis and pleural fluid analysis. Am
represent noncardiogenic pulmonary edema with Rev Respir Dis. 1 993; 1 4 8 :232-234.
a pathogenesis similar to those with hantavirus 1 9 . Jayes RL, Kamerow HN, Hasselquist SM, et al. Dissemi
syndrome or Rocky Mountain spotted fever. The nated pneumocysrosis presenting as a pleural effusion. Chest.
treatment of choice for ehrlichiosis is doxycycline. 1 993; I 03:306-308.
20. Frye MD, Pozsik CJ, Sahn SA. Tuberculous pleurisy is more
common in AIDS rhan in non-AIDS patients with tuberculo
sis. Chest. 1 997; 1 1 2:393-397.
2 1 . Cordero PJ, Gil Suay V, Greses JV, et al. The clinical charac
REFERENCES
teristics of pleural tuberculosis in parients with and without
I . Afessa B. Pleural effusion and pneumothorax in hospitalized human immunodeficiency virus infection [Translated from
patients with HIV infection: the pulmonary complications, Spanish] . Arch Bronconeumol. 1 99 5 ;3 1 : 5 1 2-5 1 8 .
ICU support, and prognostic facrors of hospitalized patients 22. Jones BE, Young SMM, Anroniskis D, et al. Relationship
with HIV (PIP) study. Chest. 2000; 1 1 7: 1 03 1 - 1 037. of the manifestations of tuberculosis ro CD4 cell counts in
2 . Cadranel JL, Chouaid C, Denis M, et al. Causes of pleu patients with human immunodeficiency virus infection. Am
ral effusion in 75 HIV-infected patients (Letter) . Chest. Rev Respir Dis. 1 993; 1 4 8 : 1 292- 1 297.
1 993; 1 04:65 5 . 23. Relkin F, Aranda CP, Garay SM, et al. Pleural tuberculosis and
3 . Batungwanayo ], Tadman H, Allen S, e t al. Pleural effusion, HIV infection. Chest. 1 994; 1 0 5 : 1 33 8- 1 34 1 .
tuberculosis and HIV- I infection in Kigali, Rwanda. AIDS. 24. Hodsdon WS, Luzze H , Hurst TJ, e t al. HIV- I -related pleu
1 993;7:73-79. ral tuberculosis: elevated production of IFN-gamma, but
4. O'Brien RF, Cohn DL. Serosanguineous pleural effusions in failure of immunity ro Mycobacterium tuberculosis. AIDS.
AIDS-associated Kaposi's sarcoma. Chest. I 989;96:460--466. 200 1 ; 1 5 :467--475 .
5 . Aboulafia DM. The epidemiologic, pathologic, and clinical 2 5 . Heyderman RS, Makunike R , Muza T, e t al. Pleural tubercu
features of AIDS-associated pulmonary Kaposi's sarcoma. losis in Harare, Zimbabwe: the relationship between human
Chest. 2000; 1 1 7: 1 1 28- 1 1 4 5 . immunodeficiency virus, CD4 lymphocyte count, granuloma
6. Bubman D, Cesarman E. Pathogenesis of Kaposi's sarcoma. formation and disseminated disease. Trop Med Int Health.
Hematol Oncol Clin North Am. 2003; 1 7 : 7 1 7-74 5 . 1 998;3: 1 4-20.
7. Miller RF, Howling SJ, Reid AJ, e t al. Pleural effusions i n 26. Kitinya JN, Richter C, Perenboom R, et al. Influence of HIV
patients with AIDS. Sex Transm Infect. 2000;76: 1 22- 1 2 5 . status on pathological changes in tuberculous pleuritis. Tuber
8 . Khalil AM , Carette MF, Cadranel JL, e t al. lntrathoracic Lung Dis. 1 994;75 : 1 95-1 9 8 .
Kaposi's sarcoma. CT findings. Chest. 1 99 5 ; 1 0 8 : 1 622- 1 626. 2 7 . Luzze H, Elliotr AM , Joloba M L , et al. Evaluation of sus
9 . Light RW, Hamm H . Pleural disease and the acquired immune pected tuberculous pleurisy: clinical and diagnostic findings
deficiency syndrome. Eur Respir]. 1 997; 1 0:2638-2643. in HIV- I -positive and HIV-negative adults in Uganda. Int
1 0 . Nakamura S , Murakami-Mori K, Rao N, et al. Vascular J Tuberc Lung Dis. 200 1 ; 5 :746-753.
endothelial growth facror is a potent angiogenic facror in 2 8 . Elliott AM , Luzze H, Quigley MA, et al. A randomized,
AIDS-associated Kaposi's sarcoma-derived spindle cells . double-blind, placebo-controlled trial of the use of predniso
J Immunol. 1 997; 1 5 8 :4992-500 1 . lone as an adjunct to treatment in HIV- I -associated pleural
1 1 . Huang L, Schnapp LM, Gruden JF, et al. Presentation of tuberculosis. J Infect Dis. 2004; 1 90: 869-878.
AIDS-relared pulmonary Kaposi's sarcoma diagnosed by bron 29. Newman TG, Soni A, Acaron S, et al. Pleural cryprococcosis in the
choscopy. Am ] Respir Grit Care Med. 1 996; 1 5 3 : 1 385-1 390. acquired immune deficiency syndrome. Chest. 1 987;91 :459--460.
12. Ognibene FP, Shelhamer JH. Kaposi's sarcoma. Clin Chest 30. Ankobiah WA, Vaidya K, Powell S, et al. Disseminated histo
Med. 1 988;9:45 9-463. plasmosis in AIDS. Clinicoparhologic features in seven patients
13. Bryant-Greenwood P, Sorbara L, Filie AC, et al. Infection of from a non-endemic area. N Y State J Med. I 990;90:234-23 8.
mesothelial cells with human herpes virus 8 in human immu 3 1 . Uttamchandani RB, Daikos GL, Reyes RR, et al. Nocardio
nodeficiency virus-infected patients with Kaposi's sarcoma, sis in 30 patients with advanced human immunodeficiency
Castleman's disease, and recurrent pleural effusions. Mod virus infection: clinical features and outcome. Clin Infect Dis.
Pathol. 2003; 1 6 : 1 45- 1 5 3 . 1 994; 1 8 :348-353.
1 4 . Suay V, Cordero PJ, Martinez E, e t al. Parapneumonic effu 32. Aronchick JM, Miller WT. Disseminated nontuberculous
sions secondary ro community-acquired bacterial pneumonia mycobacterial infections in immunosuppressed patients.
in human immunodeficiency virus-infected parients. Eur Semin Roentgenol. 1 993;28: 1 5 0- 1 57.
Respir]. 1 99 5 ; 8 : 1 934- 1 939. 33. Joseph J, Strange C, Sahn SA. Pleural effusions in hospitalized
15. Mohammed KA, Nasreen N, Ward MJ, et al. Induction of patients with AIDS. Ann Intern Med. 1 99 3 ; 1 1 8 : 8 5 6-869.
acute pleural inflammation by Staphylococcus aureus. I. CD4 + 34. Lababidi HMS, Gupta K, Newman T, et al. A retrospecrive
T cells play a critical role in experimental empyema. J Infect analysis of pleural effusion in human immunodeficiency virus
Dis. 2000; 1 8 1 : 1 693-1 699. infected patients. Chest. 1 994; I 06:86S.
C H A PT E R 1 6 I PLE U RAL E F F U S I O N D U E TO A I D S OTH E R V I R U S E S 285
3 5 . Miller WT Jr, Barbosa E Jr, Mickus TJ, et al. Chest com 5 8 . Thijsen SF, Luderer R, van Gorp JM, et al. A possible role for
puted tomographic imaging characteristics of viral acute lower Epstein-Barr virus in the pathogenesis of pleural effusion. Eur
respiratory tract illnesses: a case-control study. J Comput Assist Respir]. 2005;26:662-666.
Tomogr. 20 1 1 ;3 5 : 524-5 30. 59. Lander P, Palayew MJ . Infectious mononucleosis: a review of
36. Alpcekin F. An epidemic of pleurisy with effusion in Biclis, chest roentgenographic manifestations. J Can Assoc Radio!.
Turkey: scudy of559 cases. U SArmedForcesMedj 1 958;9: 1-1 1 . 1 974;25:303-306.
37. Fine NL, Smith LR, Sheedy P F. Frequency o f pleural effu 60. Fermaglich DR. Pulmonary involvement in infectious mono
sions in mycoplasma and viral pneumonias. N Engl J Med. nucleosis. ] Pediatr. 1 975;86:93-9 5 .
1 970;283:790-793. 6 1 . Sarkar TK. Infectious mononucleosis with pleural effusion.
3 8 . Light RW, Girard WM, Jenkinson SG, et al. Parapneumonic Chest. l 969;56:359-360.
effusions. Am ] Med. 1 980;69:5 07-5 1 1 . 62. Laferi H. Pleural effusion and ascices on return from Pakistan.
39. Monath TP, Maher M , Casals J , et al. Lassa fever in the East Lancet. 1 997;3 5 0 : 1 072.
ern Province of Sierra Leone, 1 970- 1 972. II. Clinical obser 63. Wang CC, Wu CC, Liu JW, et al. Chest radiographic presen
vations and virological studies on selected hospital cases. Am tation in patients with dengue hemorrhagic fever. Am J Trop
] Trop Med Hyg. 1 974;23 : 1 140- 1 1 4 9 . Med Hyg. 2007;77:29 1 -296.
40. Gross PA, Gerding DN. Pleural effusion associated with viral 64. Setiawan MW, Samsi TK, Wulur H, et al. Dengue haemor
hepatitis. Gastroenterology. l 971 ;60:898-902. rhagic fever: ultrasound as an aid to predict the severity of the
4 1 . Cho CT, Hiatt WO, Behbehami AM . Pneumonia and massive disease. Pediatr Radio!. 1 998;28: 1 -4.
pleural effusion associated with adenovirus type 7. Am J Dis 65. Avirutnan P, Malasit P, Seliger B, et al. Dengue virus infec
Child. 1 973; 1 26:92-94. tion of human endothelial cells leads to chemokine produc
42. Goodman ZD, Gupta PK, Frost JK, et al. Cycodiagnosis of viral tion, complement activation, and apoptosis. J Immunol.
infections in body cavity fluids. Acta Cytol. l 979;23:204-208. 1 998; 1 6 1 :6338-6346.
43. Charles RE, Katz RL, Ordonez NG, et al. Varicella-zos 66. Soepandi PZ, Buchan E, Mangunnegoro H, et al. Clinical
ter infection with pleural involvement. Am J Clin Pathol. course of avian influenza A (H5N 1) in patients at the Per
1 9 86;85: 522-526. sahabatan Hospital, Jakarta, Indonesia, 2005-2008. Chest.
44. Bustamante EA, Levy H , Simpson SQ. Pleural fluid char 20 1 0; 1 3 8 :665-673.
acteristics in hancavirus pulmonary syndrome. Chest. 67. McEwen RE, Scriven JE, Green CA, et al. Chest radiography
1 997; 1 1 2: 1 1 33-1 1 36. findings in adults with pandemic H l N l 2009 influenza. Br]
4 5 . Levy H, Simpson SQ. Hantavirus pulmonary syndrome. Am Radio!. 20 1 0;83 :499-504.
J Respir Crit Care Med. 1 994; 1 4 9 : 1 7 1 0- 1 7 1 3 . 68. Milder JE, McDearmon SC, Walzer PD. Presumed respira
4 6 . Duchin J S , Koster FT, Peters CJ, e t al. Th e Hantavirus Study tory syncytial virus pneumonia in an adolescent compromised
Group. Hantavirus pulmonary syndrome: a clinical descrip hose. South Med] 1 979;72: 1 1 9 5- 1 1 98 .
tion of 17 patients with a newly recognized disease. N Engl 69. Fulginiti VA, Eller JJ, Downie AW, e t al. Altered reactivity to
] Med. 1 994;330: 949-9 5 5 . measles virus. JAMA. 1 967;202: 1 075- 1 080.
4 7 . Treanor JJ, Hayden FG. Viral infections. I n : Murray JF, Nadel 70. Trudo FJ , Gopez EV, Gupta PK, et al. Pleural effusion due to
JA, eds. Textbook ofRespiratory Medicine, 3rd ed. Philadelphia, herpes simplex type II infection in an immunocompromised
PA: WB Saunders; 2000: 929-984. host. Am ] Respir Crit Care Med. 1 997; 1 5 5 :371-373.
48. Hooper jw, Larsen T, Custer DM, et al. A lethal disease model 7 1 . Kern S, Uhl M, Berner R, et al. Respiratory syncytial virus
for hantavirus pulmonary syndrome. Virology. 200 1 ;289:6-14. infection of the lower respiratory tract: radiological findings in
49. Simila S, Ylikorkala 0, Wasz-Hockert 0 . Type 7 adenovirus 1 08 children. Eur Radio/. 200 1 ; 1 1 :25 8 1 -2584.
pneumonia. ] Pediatr. 1 97 1 ;79:605-6 1 1 . 72. Suminoe A, Matsuzaki A, Koga Y, et al. Human herpesvirus 6
5 0 . Han BK, Son JA, Yoon HK, et al. Epidemic adenoviral (HHV-6)-associated pleurisy after unrelated cord blood trans
lower respiratory tract infection in pediatric patients: radio plantation in children with chemotherapy-resistant malignant
graphic and clinical characteristics. A]R Am J Roentgenol. lymphoma. J Pediatr Hematol Oncol. 2007;29:709-7 1 2 .
1 998; 1 70: 1 077- 1 080. 7 3 . Wong KT, Antonio GE, Hui D S , e t al. Severe acute respira
5 1 . Kacsilabros L, Triandafillou G, Kontoyiannis P, et al. Pleural tory syndrome: radiographic appearances and pattern of pro
effusion and hepatitis. Gastroenterology. 1 972;63:7 1 8 . gression in 1 3 8 patients. Radiology. 2003;228:40 1 -406.
5 2 . Cocchi P, Silenzi M . Pleural effusion i n HBsAG-positive hepa 74. Mansel JK, Rosenow EC III, Smith TF, et al. Mycoplasma
titis. ] Pediatr. 1 976;89:329-330. pneumoniae pneumonia. Chest. 1 989;95:639-646.
5 3 . Owen RL, Shapiro H . Pleural effusion, rash, and anergy in 7 5 . Decancq HG Jr, Lee FA. Mycoplasma pneumoniae pneumonia.
icteric hepatitis. N Engl] Med. 1 974;29 1 : 963-964. JAMA. 1 965; 1 94: 1 0 1 0-1 0 1 1 .
54. Tabor E, Russell RP, Gerety RJ, et al. Hepatitis B surface 76. Grix A, Giammona ST. Pneumonitis with pleural effusion in
antigen and e antigen in pleural effusion: a case report. children due to Mycoplasma pneumoniae. Am Rev Respir Dis.
Gastroenterology. 1 977;73 : 1 1 57- 1 1 5 9 . 1 974; 1 0 9 :665-67 1 .
5 5 . Sposito M, Petroni VA, Valeri L. Importanza diagnoscica 77. Narica M , Macsuzono Y, Icakura 0 , et al. Analysis of myco
dei piccoli versamenti pleurici nella virus epatite. Epatologia. plasmal pleural effusion by the polymerase chain reaction.
1 966; 1 2:228-23 1 . Arch Dis Child. 1 998;78 :67-69.
56. Lee HS, Yang PM, Liu BF, et al. Pleural effusion coinciding 7 8 . Nagayama Y, Sakurai N, Tamai K, et al. Isolation of Myco
with acute exacerbations in a patient with chronic hepatitis B. plasma pneumoniae from pleural fluid and/or cerebrospinal
Gastroenterology. 1 9 89;96: 1 604- 1 606. fluid: report of four cases. Scand] Infect Dis. 1 987; 1 9 : 5 2 1 -524.
57. Marero E, Ausina V. The role of Epstein-Barr virus in pleu 79. Sobradillo V, Ansola P, Baranda F, et al. Q fever pneumonia: a
ral effusions of unknown aetiology: an interesting clinical review of 1 64 community-acquired cases in the Basque coun
perspective. Eur Respir J 2005;26:566-5 68. try. Eur Respir J l 989;2:263-266.
286 PLE U RAL D I S EASES
8 0 . Gordon JK, MacKeen AD, Marrie TJ, et al. The radiographic 83. Byrd RP Jr, Vasquez ], Roy TM. Respiratory manifestations of
features of epidemic and sporadic Q fever pneumonia. J Can tick-borne diseases in the Southeastern United States. South
Assoc Radio!. 1 984;3 5 : 293-296. Med] 1 997;90: !-4.
8 1 . Esteban C, Oribe M, Fernandez A, et al. Increased adenosine 84. Donohue JF. Lower respiratory tract involvement in Rocky
deaminase activity in Q fever pneumonia with pleural effu Mountain spotted fever. Arch Intern Med. 1 9 80; 1 40:223-227.
sion. Chest. 1 994; 1 0 5 :648. 8 5 . Fordham LA, Chung CJ, Specter BB, et al. Ehrlichiosis: find
82. Murphy PP, Richardson SG. Q fever pneumonia presenting as ings on chest radiographs in three pediatric patients. A]R Am
an eosinophilic pleural effusion. Thorax. 1 989 ;44:228-229. ] Roentgenol. 1 998 ; 1 7 1 : 1 42 1 - 1 424.
P l e u ra l E ff u s i o n D u e to P u l m o n a ry
E m bo l i zat i o n
The disorder most commonly overlooked in the pleural biopsy. Of the 1 9 patients who did not have
workup of a patient with pleural effusion is pul malignant disease, 2 subsequently died, and both had
monary embolization ( 1 ,2) . The possibility of pul pulmonary emboli at autopsy. One wonders how
monary embolization should be excluded in every many of the remaining 1 7 patients might have had
patient with a pleural effusion of uncertain origin. pulmonary emboli if this diagnosis had been consid
ered. Along the same lines, Storey et al. ( 1 0) reported
on a series of 1 33 patients with pleural effusions in
INCIDENCE
which only 3 were due to pulmonary emboli, but
It is estimated that at least 800,000 persons have a causes were not determined in 25 patients. Because
pulmonary embolic event each year in this country these authors do not mention any evaluation of their
(3) . The incidence of pulmonary embolus increased patients for pulmonary emboli, one wonders how
50% between 1 998 and 2005 (3) . Because pleural many of the 25 patients would have been switched
effusions occur in 30% to 50% of patients with pul from the undetermined category to the pulmonary
monary emboli (4-6), 240,000 to 400,000 pleural embolus category if the possibility of pulmonary
effusions secondary to pulmonary emboli should embolus had been explored.
occur annually. Therefore, one should expect to see
more cases of pleural effusions secondary to pulmo
PAT H O P H YS I O LO G I C M E C H A N I S M S
nary embolization than due to bronchogenic carci
noma. Nevertheless, in most large series, pulmonary The primary mechanism by which pulmonary emboli
embolization accounts for less than 5% of the pleural produce pleural effusion is by increasing the perme
effusions. There are two explanations for this discrep ability of the capillaries in the lung. The interstitial
ancy. First, individuals interested in pleural effusion fluid that results from this increased permeability tra
do not see many of the patients that have pleural verses the visceral pleura and leads to the accumulation
effusions due to pulmonary emboli as the effusions of pleural fluid. In the experimental situation, it has
are small and a thoracentesis is not performed (7) . been shown that more than 20% of the fluid formed
Second, the diagnosis of pulmonary embolus is fre in the lung with increased permeability pulmonary
quently not considered in patients with undiagnosed edema is cleared through the pleural space ( 1 1 ) . It
pleural effusions. Indeed, in an epidemiologic study is probable that ischemia of the capillaries in the vis
from the Czech Republic, pulmonary embolism was ceral pleura plays at most a minor role because these
the fourth leading cause of pleural effusion (8) . capillaries are supplied by the bronchial circulation
It is likely that pulmonary embolism is respon ( 1 2) . Leckie and Tothill ( 1 3) have demonstrated that
sible for a substantial fraction of undiagnosed pleu patients with a pleural effusion secondary to pulmo
ral effusions. Gunnels (9) followed 27 patients with nary emboli have a large amount of protein entering
exudative pleural effusions in whom no diagnosis and leaving the pleural space. The main factor respon
was established after an initial workup, including sible for the increased permeability of the pulmonary
287
288 PLE U RAL D I S EASES
capillaries is probably the release of inflammatory pleural effusions secondary to pulmonary embolism,
mediators from the platelet-rich thrombi. It is pos 28 (45%) had no associated infiltrate (4) , but in
sible that vascular endothelial growth factor (VEGF) another series of 20 patients, only 1 (5%) did not
may play a role in the formation of pleural fluid in at have an associated infiltrate ( 1 9) . In a third series
least some patients. Indeed a very high pleural fluid of 10 patients with pulmonary emboli and bilateral
VEGF level was reported in one patient with pulmo pleural effusions, only 3 (30%) had parenchymal
nary embolism ( 1 4) . The release of such mediators infiltrates (20) . Infiltrates are usually in the lower
can increase the permeability of the capillaries in lobes, are pleural based, and are convex toward the
either the visceral pleura or the lung. lschemia of the hilum. Patients with an embolic occlusion of segmen
pulmonary capillaries distal to the embolus may also tal pulmonary arteries are more likely to have infil
contribute to the increased permeability. trates than those with an embolic occlusion of the
central arteries ( 1 9) .
The pleural effusions secondary to pulmonary
C LI N I CA L M A N I F E STATI O N S emboli are small, with the mean size equal to approxi
mately 1 5 % of the hemithorax (4) . In the PI OPED
Sym pto m s a n d S i g n s study, 48 of the 56 effusions (86%) were manifest
There are three symptom complexes associated with only as blunting of the costophrenic angle and in no
pulmonary emboli: (a) pleuritic pain or hemoptysis, patient did the pleural effusion occupy more than one
(b) isolated dyspnea, and (c) circulatory collapse. In third of a hemithorax (7, 1 6) . In a second series of
the Prospective Investigation of Pulmonary Embo 73 patients (7) , effusions occupied less than a third
lism Diagnosis (PIOPED) study, 56% of the 1 1 9 of the hemithorax in 66 (90%) , occupied 50% of
patients with pleuritic chest pain or hemoptysis had the hemithorax in 3 (4%), and occupied more than
pleural effusion, 26% of the 3 1 with isolated dys two thirds the hemithorax in 4 (6%) . If parenchymal
pnea had pleural effusion, and none of the 5 with infiltrates are present, the pleural effusions are larger.
circulatory collapse had pleural effusion ( 1 5) . More In one series, the pleural effusion occupied greater
than 75% of patients with pleural effusions secondary than 1 5 % of the hemithorax in 74% of the patients
to pulmonary emboli have pleuritic chest pain ( 1 6) , with parenchymal infiltrates but in only 2 1 % of those
which i s almost invariably on the side o f the effusion without parenchymal infiltrates (4) . The pleural effu
(4) . Indeed, the presence of pleuritic chest pain in a sions are usually unilateral, but about 1 5 % to 3 5 % are
patient with pleural effusion is suggestive of pulmo bilateral (7,2 1 ) . The pleural effusions with pulmonary
nary embolus. In one series, pulmonary emboli were emboli may be loculated, particularly if the diagno
present in 1 2 of 22 patients (55%) younger than the sis has been delayed for more than 10 days (7,22) .
age of 40 who presented as outpatients with pleural There is not a close relationship between the sided
effusion and pleuritic chest pain ( 1 7) . Dyspnea, also ness of the pleural effusion and that of the pulmonary
present in more than 70% of patients ( 1 6, 1 8) , is usu em bolus (7) . In one study of 93 patients (7) , the pul
ally out of proportion to the size of the pleural effusion. monary embolus was unilateral in 6 1 and the pleural
Cough and apprehension are present in approximately effusion was ipsilateral in 3 8, on the contralateral side
50% of patients ( 1 6, 1 8) . Approximately 50% of these in 7, and bilateral in 1 6 .
patients are febrile ( 1 8) but less than 1 0% have tem
peratures above 3 8 . 5°C ( 1 6, 1 8) . Approximately 1 5 % P l e u ra l F l u i d F i n d i n g s
have hemoptysis ( 1 6) . Most patients have a respira
tory rate above 20 per minute, and a heart rate above In patients with pulmonary emboli, analysis of the
1 00 per minute occurs in approximately 40% (7, 1 5) . pleural fluid is not helpful in establishing the diagnosis
I n the PIOPED study, 1 1 3 o f 1 1 7 patients (97%) because the pleural fluid associated with pulmonary
with no preexisting cardiac or pulmonary disease had emboli can vary widely. Nevertheless, a thoracentesis
dyspnea or tachypnea or pleuritic chest pain ( 1 6) . should be performed in patients suspected of having
pulmonary emboli to exclude other causes of pleural
effusion such as tuberculosis, malignant disease, or
Ch est Rad iograph
pneumonia with a parapneumonic effusion.
When a pleural effusion is secondary to pulmonary Although in the past it has been stated that the
emboli, an associated parenchymal infiltrate may or pleural effusion with pulmonary embolus may be a
may not be present. In one series of 62 patients with transudate or an exudate, it appears that almost all
C HAPT E R 1 7 I P L E U RAL E F F U S I O N D U E TO P U L M O NARY E M B O L I ZATI O N 289
by factor XIII and is the primary component of during the subsequent 3 months (26) . One disturb
thrombus material. Increased levels of D-dimer are ing aspect of the study by Wells et al. is that the
found in conditions that result in the activation of D-dimer test was negative in 1 8 of the 66 patients
the fibrinolytic system. Therefore, D-dimer tests with pulmonary embolism (26) . However, Kelly and
lack specificity for thromboembolism, although it Hunt (34) concluded that patients presenting with
appears that normal levels are useful in excluding it a low pretest probability of pulmonary embolus and
(3 1 ) . Elevated levels of D-dimer are also found in a negative D-dimer test with a modern test need
patients with recent surgery, malignancy, and liver not undergo imaging or treatment for pulmonary
disease (32) . In fact, more than 5 0 % of hospitalized embolism (34) . It has been shown that the higher
patients have elevated levels of D-dimer. Neverthe the D-dimer levels, the more likely the diagnosis of
less, if the levels are normal and an appropriate test pulmonary embolism (3 5 ) .
is used, the diagnosis of pulmonary embolus can be
excluded. However, if the D-dimer test is positive,
L u n g Sca n s
an additional test is necessary to definitely diagnose
pulmonary embolism (3 1 ) . In general, the perfusion lung scan has significant
I t is important fo r physicians to understand the limitations in the diagnosis of pulmonary embolism.
characteristics and limitations of the test that is If the perfusion lung scan is negative, a pulmonary
performed at their hospital (3 1 ,33) . For example, embolus is virtually ruled out. If the perfusion lung
I recently ordered a D-dimer test on a young lady scan is a high-probability one, 87% of the patients
with acute shortness of breath to rule out pulmo will have pulmonary emboli, and when coupled with
nary embolism and subsequently discovered that a high clinical probability of embolism, the positive
the only D-dimer test available at my hospital was predictive value increases to 96% (36) . However, the
a latex agglutination test, which was only useful in patients most likely to have a pleural effusion are
the detection of disseminated intravascular coagu those with pleuritic pain or hemoptysis, and less than
lation and had no role in screening for deep vein one third of these patients will have a high-probabil
thrombosis. ity lung scan ( 1 5) .
The enzyme-linked immunosorbent assay I f a pleural effusion i s present, the perfusion lung
(ELISA) tests have a high sensitivity in screening for scan is even more difficult to interpret (Fig. 1 7 . 1 ) .
pulmonary embolism, but are expensive and must be A large effusion severely restricts the ability o f the
performed in batches. The latex agglutination tests in lung to expand and causes a shift of perfusion to
general are much less sensitive (29) . Recently newer the contralateral lung (37) . Small, mobile effusions
tests such as the Nycocard assay, the VIDAS D-dimer of any origin may gravitate to different regions of
assay and the Minutex D-dimer assay have been the pleural space, depending on the position of the
developed that appear to have a higher sensitivity and patient at the time of the examination. For example,
can be performed rapidly (3 1 ) . fluid may enter the major fissures when the patient
Th e utility o f using a D-dimer test fo r screening lies down and produce a perfusion defect on the lung
is demonstrated by a study by Wells et al. (26) . These scan, when no comparable defect is seen on the erect
researchers prospectively studied 930 patients sus chest radiograph. Similarly, mismatching of the ven
pected of having pulmonary embolus. In their initial tilation and perfusion lung scans can be produced by
evaluation, they obtained a D-dimer test (Sim the pleural fluid itself when the scans are obtained in
pliRED whole-blood agglutination) and assessed the different positions (Fig. 1 7 . 1 ) (3 7) . For these reasons,
clinical probability of pulmonary embolus as high consideration should be given to performing a thera
( n = 64) , moderate ( n = 3 3 9 ) , and low ( n = 5 27) . peutic thoracentesis (see Chapter 28) before obtain
If the D-dimer test was negative and the clinical ing the lung scan.
probability was low, no additional tests were per
formed. If the D-dimer test was negative but the
clinical probability was moderate or high, additional Com p uted To m o g ra p hy A n g i og ra p h y
diagnostic tests were performed. When the D-dimer In more and more centers, a CT angiography scan is
test was performed on the 527 patients with low being used rather than the perfusion scan as the initial
probability, it was negative in 437 (83%) . These imaging study for pulmonary emboli. The diagnosis
patients did not undergo further evaluation, and of pulmonary embolism is based on the presence of
only one (O . l %) developed a pulmonary embolus partial or complete filling defects in the pulmonary
C HAPT E R 1 7 I P L E U RAL E F F U S I O N D U E TO P U L M O NARY E M B O L I ZATI O N 29 1
artery on the contrast-enhanced CT (Fig. 1 7.2) . CT angiography has its greatest sensitivity for detect
In general, CT angiography detects between 75% ing emboli in the main, lobar, or segmental pulmo
and 1 00% of emboli and the specificity of the test nary arteries (4 1 ) . This should not be too much of
exceeds 90% (38,39) . These numbers are improving a problem because in the PIOPED study, only 6%
with the latest generation of scanners (40) . Indeed in of the patients with pulmonary emboli had isolated
a recent survey, 86.7% of respondents believed that subsegmental emboli (42) .
CT angiography was the most useful procedure for I believe that CT angiography is the best way
patients with acute pulmonary embolism (40) . If CT to evaluate the possibility of pulmonary emboli in
angiography is used for the diagnosis of pulmonary patients with a pleural effusion. Patients with a pleural
emboli, it is important that the pulmonary arteries effusion are likely to have an embolus in the central,
be examined on the video monitor and not j ust on lobar, segmental, or subsegmental pulmonary arter
the hard copies of the scans. False-positive and false ies, and these are the areas in which CT angiography
negative results are common when the length of the can detect an embolus. The additional advantage of
pulmonary artery is not scrutinized on the monitor. obtaining a CT angiogram in a patient with pleural
292 PLE U RAL D I S EASES
TREAT M E N T
Th e treatment o f the patient with pleural effusion
secondary to pulmonary embolization is the same as
that for any patient with pulmonary emboli. If the
patient has a moderate or a high probability of having
a pulmonary embolus, treatment should be started
immediately before any imaging studies are obtained.
Low-molecular-weight heparins (LMWHs) are now
the initial drugs of choice for the treatment of pul
monary embolism (46) . The advantages of LMWHs,
compared with unfractionated heparin, include the
following: (a) improved bioavailability and a longer
half-life; (b) a more predictable dose-response curve
so that laboratory monitoring is rarely needed; (c)
less frequent heparin-induced thrombocytopenia;
and (d) less heparin-associated osteopenia. After the
initial administration of LMWHs, the patient should
FIGURE 1 7 .2 • Com p uted tomog ra p h i c a n g iog
raphy s h ow i n g a sma l l r i g ht p l e u r a l effusion, a be treated with oral anticoagulants. With treatment,
p e r i p h e ra l i nfi ltrate in t h e r i g h t l ower l o be, a n d pleural effusions gradually resolve, particularly if no
a d a rk a rea i n t h e a rtery t o the r i g ht l ower lobe infiltrates are present. In one series, the effusions had
(arro w) that rep resents the p u l m o n a ry e m b o l us. cleared completely after 7 days of therapy in 1 8 of 28
patients (64%) without parenchymal infiltrates, but
in none of the 30 patients with parenchymal infil
trates (4) .
effusion who is being evaluated for pulmonary embo An alternative to heparin in the treatment of pul
lism is that the CT angiogram can also demonstrate monary emboli is fondaparinux. This is a synthetic
pulmonary infiltrates, pulmonary masses, pleural antithrombotic agent. In one large study involving
masses, or mediastinal abnormalities that may pro more than 2,000 patients, its efficacy was equivalent
vide clues to the etiology of the pleural effusion if a to heparin (47) . It is the agent of choice in patients
pulmonary embolus is not present. with heparin-induced thrombocytopenia.
An alternative to Coumadin for the treatment of
pulmonary embolism is rivaroxaban. It is approved
D u p lex U ltrasonography of Leg Ve i n s
in Europe and Canada for the prevention of venous
An alternative approach i n diagnosing pulmonary thromboembolism in patients undergoing major
embolism is to study the legs to see if there is any orthopedic surgery. Its mechanism of action is as an
evidence of deep venous thrombosis. The basis for active inhibitor of coagulation factor Xa. It is an oral
this approach is that approximately 90% of pulmo once a day agent that does not require blood moni
nary emboli originate in the legs ( 1 ) . Probably the toring. Trials are under way assessing whether it is an
best method to evaluate the proximal veins of the effective agent for treating pulmonary embolism and
legs is duplex ultrasonography with venous compres deep venous thrombosis (48) .
sion. The sensitivity of this exceeds 90% while the The presence of bloody pleural fluid is not a
specificity exceeds 9 5 % (43 ) . If the ultrasonography contraindication to the administration of heparin
demonstrates deep venous thrombosis, the diagnosis or LMWHs. Bynum and Wilson (24) treated three
of pulmonary embolism is likely. The patient needs patients who had pleural fluid RBC greater than
to be anticoagulated and usually no further diagnos 1 00,000/mm3 with intravenous heparin, and in none
tic tests directed toward the pleural effusion are indi did the effusion increase in size.
cated. Unfortunately, deep venous thrombosis is not If the pleural effusion increases in size with anti
demonstrable in approximately 40% of patients with coagulant therapy or if a contralateral pleural effusion
pulmonary embolism (44,45) and additional tests are develops, the patient probably has recurrent emboli
necessary in these patients to delineate the etiology of or another complication. In one series, two patients
the pleural effusion ( 1 ) . developed an enlarged ipsilateral effusion and one
C HAPT E R 1 7 I P L E U RAL E F F U S I O N D U E TO P U L M O NARY E M B O L I ZATI O N 29 3
had recurrent pulmonary emboli, whereas the other typically reveals distended veins with enhancing
had developed infected pleural fluid (4) . Two other walls, low-attenuation intraluminal filling defects,
patients developed contralateral pleural effusions, and localized soft tissue edema (53) . The treatment
and both had recurrent emboli. should include an extended course of a ,8-lactamase
On rare occasions, the administration of antico resistant antibiotic and surgical drainage of any puru
agulants to patients with pulmonary emboli can lead lent fluid collection (53) .
to the development of a hemothorax. Rostand et al. Pleural effusions occur with Lemierre syndrome as
(49) reviewed 1 1 such cases and reported that the with any septic emboli. In one review of 5 9 patients
clotting studies were within an acceptable range when reported in the literature between 1 990 and 2000, 1 9
the hemothorax occurred in 7 of the patients. Hemo (32%) had pleural effusion, 8 ( 1 4%) had empyema,
thorax developed within the first week of anticoagula and 7 ( 1 2%) had pneumothorax (54) . If the pleu
tion in 9 of the 1 1 reported cases and developed on ral effusion is infected, tube thoracostomy or tho
the side of the embolus in each of these patients. In racoscopy with the breakdown of adhesions may be
two patients, the hemothorax developed while they necessary.
were receiving long-term anticoagulation. When a
pleural effusion increases in size in a patient with pul P L E U RA L E F F U S I O N S WITH S I C K LE
monary emboli, a diagnostic thoracentesis should be C E LL A N E M IA
performed to rule out a complicated parapneumonic
effusion or a hemothorax. If bloody pleural fluid is There is a high incidence of pleural effusions in
obtained, the hematocrit of the pleural fluid should patients with sickle cell anemia who develop the
be determined. If the hematocrit on the pleural fluid acute chest syndrome. In one series of 1 07 episodes
is greater than 50% of that of the peripheral blood, of the acute chest syndrome in 77 adults, unilat
anticoagulation should be discontinued and chest eral pleural effusions were present in 3 5 % , whereas
tubes should be inserted (see Chapter 25) . bilateral pleural effusions were present in 1 4% (55) .
The pathogenesis of these effusions is not definitely
known but is believed to be either in situ thrombosis
P L E U RA L E F F U S I O N S WITH R I G H T
of the pulmonary arteries or fat embolization ( 5 5 ) .
S I D E D E N D O CA R D I T I S A N D S E PT I C
The characteristics of the pleural fluid associated with
E M B O LI
the acute chest syndrome with sickle cell anemia are
Patients who have right-sided endocarditis may have not known.
septic pulmonary emboli and pleural effusions (50) .
The incidence of pleural effusion with right-sided
endocarditis is approximately 25% to 67% (50, 5 1 ) . REFERENCES
Th e pleural fluid i s an exudate, and the cultures are
1 . Light RW. Pleural effusion in pulmonary embolism. Semin
usually negative. The differential cell count can reveal Respir Crit Care Med. 20 1 0;3 1 :7 1 6-722.
neutrophils, lymphocytes, or monomesothelial cells 2 . Findik S . Pleural effusion in pulmonary embolism. Curr Opin
(49 ). The pleural fluid lactate dehydrogenase (LDH) Pulm Med. 20 1 2 ; 1 8;347-3 54.
level is usually higher than the serum LDH level and 3 . Park B, Messina L, Dargon P, et al. Recent trends in clinical
outcomes and resource utilization for pulmonary embolism
may be elevated to 1 0 times or more the upper limit
in the United States: findings from the nationwide inpatient
for serum. sample. Chest. 2009; 1 36: 983-990.
4. Bynum LJ, Wilson JE III. Radiographic features of pleu
ral effusions in pulmonary embolism. Am Rev Respir Dis.
P L E U RA L E F F U S I O N S WITH L E M I E R R E 1 978; 1 1 7 : 829-834.
SYN D RO M E 5 . Worsley DF, Alavi A, Aronchick JM, et al. Chest radiographic
findings in patients with acute pulmonary embolism: observa
Lemierre syndrome is a rare life-threatening septic tions from the PIOPED study. Radiology. 1 993; 1 89 : 1 33- 1 36.
thrombophlebitis of a j ugular vein with anaerobic 6. Stein PD, Athanasoulis C, Greenspan RH, et al. Relation of
septicemia (52) . The most common pathogen asso plain chest radiographic findings to pulmonary arterial pres
ciated with the syndrome is Fusobacterium necropho sure and arterial blood oxygen levels in patients with acute
pulmonary embolism. Am J Cardiol. l 992;69:394-396.
rum. The diagnosis is established with evidence of
7. Poree! JM, Madronero AB, Pardina M , et al. Analysis of
metastatic infection and internal j ugular vein throm pleural effusions in acute pulmonary embolism: radiologi
bophlebitis. The diagnostic procedure of choice is cal and pleural fluid data from 230 patients. Respirology.
a contrast-enhanced CT of the neck and chest that 2007; 1 2:234-239 .
294 PLE U RAL D I S EASES
8 . Mare! M, Arustova M, Stasny B, et al. Incidence of pleural 2 9 . Race GA, Scheifley CH, Edward JE. Hydrothorax i n conges
effusion in a well-defined region: epidemiologic study in cen tive heart failure. Am ] Med. 1 957;22 :83-89.
tral Bohemia. Chest. 1 993; 1 04: 1 486- 1 4 8 9 . 30. Fraser RS , Muller NL, Colman N, et al. Diagnosis of Dis
9 . Gunnels JJ. Perplexing pleural effusion. Chest. eases of the Chest, 4th ed. Philadelphia, PA: WB Saunders;
1 978;74:390-393. 2000:2659-26 9 5 .
1 0. Storey DD, Dines DE, Coles DT. Pleural effusion: a diagnos 3 1 . Ahearn GS, Bounameaux H . The role of D-dimer in the diag
tic dilemma. JAMA. l 976;236 : 2 1 83-2 1 86. nosis of venous thromboembolism. Semin Respir Crit Care
1 1 . Wiener-Kronish JP, Broaddus VC, Albertine KH, et al. Med. 2000;2 1 : 52 1-535.
Relationship of pleural effusions to increased permeabil 32. Quinn DA, Fogel RB, Smith CD, et al. D-dimers in the diag
ity pulmonary edema in anesthetized sheep. J Clin Invest. nosis of pulmonary embolism. Am J Respir Crit Care Med.
1 988;82: 1 422- 1 429. 1 999; 1 5 9 : 1 445-1 449.
1 2 . Albertine KH, Wiener-Kronish JP, Roos PJ, et al. Structure, 33. Elliott CG. The diagnostic approach to deep venous thrombo
blood supply, and lymphatic vessels of the sheep's visceral sis. Semin Respir Crit Care Med. 2000;2 1 : 5 1 1-5 1 9 .
pleura. Am ]Anat. 1 982; 1 6 5 :277-294. 3 4 . Kelly J, Hunt BJ. A clinical probability assessment and
13. Leckie WJH, Tothill P. Albumin turnover in pleural effusions. D-dimer measurement should be the initial step in the
Clin Sci. 1 965;29:339-352. investigation of suspected venous thromboembolism. Chest.
1 4 . Cheng C-S, Rodriguez RM, Perkecc EA, et al. Vascular endothe 2003; 1 24: 1 1 1 6- 1 1 1 9 .
lial growth factor in pleural fluid. Chest. 1 99 9; 1 1 5 :760-765 . 3 5 . Tick LW, Nijkeuter M , Kramer MH, e t al. High D-dimer lev
1 5 . Stein P D , Henry J W Clinical characteristics o f patients with els increase the likelihood of pulmonary embolism. J Intern
acute pulmonary embolism stratified according to their pre Med. 2008;264: 1 9 5-200.
senting syndromes. Chest. 1 997; 1 1 2:974-979. 36. Fedullo PF. Pulmonary thromboembolism. In: Mason RJ,
16. Stein PD, Terrin ML, Hales CA, et al. Clinical, laboratory, Broaddus VC, Murray JF, et al. eds. Textbook of Respira
roentgenographic, and electrocardiographic findings in tory Medicine, 4th ed. Philadelphia, PA: WB Saunders;
patients with acute pulmonary embolism and no pre-existing 200 5 : 1 425- 1 4 5 8 .
cardiac or pulmonary disease. Chest. 1 99 1 ; 1 00 : 5 9 8-603 . 3 7 . Baum S, Vincent N R , Lyons KP, e t al. Atlas ofNuclear Medi
1 7. Branch WR, McNeil BJ. Analysis of the differential diagnosis cine Imaging. New York, NY: Appleton-Century-Crofts;
and assessment of pleuritic chest pain in young adulcs. Am J 1 98 1 .
Med. 1 9 83;75 :67 1 -679. 3 8 . Goodman PC. Spiral CT fo r pulmonary embolism. Semin
1 8 . Bell WR, Simon TL, DeMets DL. The clinical features of Respir Crit Care Med. 2000;2 1 : 503-5 1 0 .
submassive and massive pulmonary emboli. Am J Med. 39. d e Monye W, van Strijen MJ, Huisman MV, e t al. Suspected
1 977;62: 3 5 5-360. pulmonary embolism. Prevalence and anatomic distribution
19. Dalen JE, Haffajee CI, Alpert JS III, et al. Pulmonary embo in 487 consecutive patients; advances in new technologies
lism, pulmonary hemorrhage and pulmonary infarction. N evaluating the localization of pulmonary emboli (ANTE
Engl] Med. 1 977;296: 1 4 3 1 - 1 43 5 . LOPE) group. Radiology. 2000;2 1 5 : 1 84-1 8 5 .
20. Rabin C B , Blackman NS. Bilateral pleural effusion: its sig 4 0 . Weiss CR, Scatarige J C , Diecce GB, e t al. C T pulmonary
nificance in association with a heart of normal size. J Mt Sinai angiography is the first-line imaging test for acute pulmo
Hosp. 1 9 57;24:45-63. nary embolism: a survey of US clinicians. Acad Radio!.
2 1 . Goldberg SN, Richardson DD, Palmer EL, et al. Pleural effu 2006; 1 3 :434-446.
sion and ventilation/perfusion scan interpretation for acute 4 1 . Clinical Practice Guideline. The diagnostic approach to
pulmonary embolus. J Nucl Med. 1 996;37: 1 3 1 0- 1 3 1 3 . acute venous thromboembolism. Am J Respir Crit Care Med.
22. Erkan L, Findyk S, Uzun 0, et al. A new radiologic appear 1 999; 1 60: 1 043-1 066.
ance of pulmonary thromboembolism: mulciloculated pleural 42. The PIOPED Investigators. Value of the ventilation/
effusions. Chest. 2004; 1 26:298-302. perfusion scan in acute pulmonary embolism. JAMA.
23. Romero Candeira S, Hernandez Blasco L, Soler MJ, et al. Bio 1 990;263:2753-2759.
chemical and cytologic characteristics of pleural effusions sec 43. Segal JB, Eng J, Tamariz LJ, et al. Review of the evidence on
ondary to pulmonary embolism. Chest. 2002; 1 2 1 :465-469 . diagnosis of deep venous thrombosis and pulmonary embo
2 4 . Bynum LJ, Wilson JE I I I . Characteristics of pleural effu lism. Ann Fam Med. 2007;5 :63-73 .
sions associated with pulmonary embolism. Arch Intern Med. 4 4 . Monreal M, Munoz-Torrero JF, Naraine V S , e t al. Pulmonary
1 976; 1 36: 1 5 9- 1 62 . embolism in patients with chronic obstructive pulmonary dis
2 5 . Spriggs A l , Boddington MM. The Cytology of Effesions, 2nd ease or congestive heart failure. Am] Med. 2006; 1 1 9 : 8 5 1 -8 5 8 .
ed. New York, NY: Grune & Stratton: 1 968. 4 5 . Girard P, Sanchez 0, Leroyer C, e t al. Evaluation d u scan
26. Wells PS, Anderson DR, Rodger M, et al. Excluding pul ner spirale clans l'embolic pulmonaire study group. Deep
monary embolism at the bedside without diagnostic imag venous thrombosis in patients with acute pulmonary embo
ing: management of patients with suspected pulmonary lism; prevalence, risk factors, and clinical significance. Chest.
embolism presenting ro the emergency department by using 2005; 1 28 : 1 5 93- 1 600.
a simple clinical model and D-dimer. Ann Intern Med. 46. Aguilar D, Goldhaber SZ. Clinical uses of low-molecular
200 1 ; 1 3 5 :98-1 07. weight heparins. Chest. 1 999; 1 1 5 : 1 4 1 8- 1 423.
27. Wicki J, Perneger TV, Junod AF, et al. Assessing clinical prob 47. Matisse Investigators. Subcutaneous fondaparinux versus
ability of pulmonary embolisms in the emergency ward: a intravenous unfractionated heparin in the initial treatment of
simple score. Arch Intern Med. 200 1 ; 1 6 1 : 92-97. pulmonary embolism. N Engl] Med. 2003;349 : 1 695-1 702.
2 8 . Le Gal G, Righini M, Roy P-M, et al. Prediction of pulmo 48. Paikin JS, Eikelboom JW, Cairns JA, et al. New antithrom
nary embolism in the emergency department. The revised botic agents-insights from clinical trials. Nat Rev Cardiol.
Geneva score. Ann Intern Med. 2006; 1 44 : 1 6 5- 1 7 1 . 2 0 1 0;7:498-509.
C HAPT E R 1 7 I P L E U RAL E F F U S I O N D U E TO P U L M O NARY E M B O L I ZATI O N 29 5
49. Rostand RA , Feldman RL, Block ER. Massive hemorhorax S3. Armstrong AW, Spooner K, Sanders JW Lemierre's syndrome.
complicating heparin anticoagulation for pulmonary embolus. Curr Infect Dis Rep. 2000;2: 1 68- 1 73 .
South Medj 1 977;70: 1 1 28- 1 1 30. S4. Gowan RT, Mehran RJ, Cardinal P, e t a l . Thoracic complica
S O . Sexauer WP, Quezado Z, Lippmann ML, et al. Pleural effu tions of Lemierre syndrome. Can Respir J 2000;7:48 1-48S .
sions in right-sided endocarditis: characteristics and parho S S . Maitre B, Habibi A, Roudot-Thoraval F, er al. Acute
physiology. South Med] 1 992;8S : l 1 76- l 1 80 . chest syndrome in adults with sickle cell disease. Chest.
S I . Huang RM, Naidich DP, Lubat E, et a l . Septic pulmonary 2000; 1 1 7: 1 386- 1 392.
emboli: CT-radiographic correlation. A]R Am J Roentgenol.
1 9 89; 1 S 3 : 4 1 -4 S .
S2. Alifano M, Venissac N, Guillot F, et a l . Lemierre's syn
drome with bilateral empyema rhoracis. Ann Thorac Surg.
2000;69 :930-93 1 .
P l e u ra l E ff u s i o n S e co n d a ry to D i se a ses
of t h e G a st ro i ntest i n a l Tra ct
Diseases of the gastrointestinal tract are sometimes pancreatitis (4) . In one series, the incidence of pleural
associated with pleural effusion. In this chapter, the effusion was 84% in 1 9 patients with severe pancreatitis
exudative pleural effusions resulting from pancreatic but only 8 .6% in 1 1 6 patients with mild pancreatitis (4) .
disease, intraabdominal abscesses, esophageal perfo The presence of a pleural effusion is also associated
ration, abdominal operations, diaphragmatic hernia, with the development of a pseudocyst. In the series of
variceal sclerotherapy, hepatic transplantation, and Lankisch et al. (3), 29% of 66 patients with pleural effu
disease of the biliary tract are discussed. Transudative sions had pancreatic pseudocysts compared with 6% of
pleural effusions that occur with cirrhosis and ascites 67 patients without pleural effusions.
are discussed in Chapter 9 . The exudative pleural effusion accompanying
acute pancreatitis results primarily from the trans
diaphragmatic transfer of the exudative fluid arising
PA N C R EAT I C D I S EA S E
from acute pancreatic inflammation and from dia
Four different types o f nonmalignant pancreatic dis phragmatic inflammation (2) . Numerous lymphatic
ease can have an accompanying pleural effusion: acute networks j oin on the peritoneal and pleural aspects
pancreatitis, pancreatic abscess, chronic pancreatitis of the diaphragm ( 1 ) . Anatomically, the tail of the
with pseudocyst, and pancreatic ascites. pancreas is in direct contact with the diaphragm.
Hence, the exudate resulting from acute pancreatic
inflammation, which is rich in pancreatic enzymes,
Acute Pan creatitis
enters the lymphatic vessels on the peritoneal side
In older reports, the incidence of pleural effusion of the diaphragm and is conveyed to the pleural side
with acute pancreatitis was relatively low (3% to of the diaphragm. Because this fluid contains high
1 7%) ( 1 ,2) . More recent reports, however, have doc levels of pancreatic enzymes, the permeability of the
umented a much higher incidence of pleural effusion. lymphatic vessels is increased and fluid leaks from
Lankisch et al. (3) obtained computed tomography the pleural lymphatic vessels into the pleural space.
(CT) scans of the chest within 72 hours of admis The high enzymatic content of the pancreatic exudate
sion in 1 33 consecutive patients with their first attack may also cause partial or complete obstruction of the
of acute pancreatitis and reported that 50% of the pleural lymphatic vessels that leads to more pleural
patients had a pleural effusion. The effusions were fluid accumulation ( 1 ) . Of course, the diaphragm
bilateral in 77%, left sided in 1 6%, and right sided itself may be inflamed from the adjacent inflamma
in 8% . The thickness of the fluid on the CT scan tory process, and this inflammation may increase the
was less than 1 0 mm in 32 patients, 1 to 2 cm in permeability of the capillaries in the diaphragmatic
1 8 patients, and greater than 2 cm in 1 6 patients with pleura. This mechanism cannot be entirely respon
the CT scans (3) . sible for the pleural fluid accumulation, because the
The presence of a pleural effusion in patients with pleural fluid amylase concentration is almost always
acute pancreatitis is an indication of more severe higher than the simultaneous serum amylase.
296
C HAPTE R 1 8 I P LE U RAL E F F U S I O N S E C O N DARY TO D I S EASES OF TH E GASTRO I NTEST I N A L TRACT 297
Chest symptoms usually dominate the clinical pic establishing the presence of pseudocyst and the site
ture of the patient with chronic pancreatic disease of the ductal disruption ( 1 8) . Endoscopic retrograde
and a pleural effusion ( 1 4) . These patients report cholangiopancreatography (ERCP) also plays an
chest pain and shortness of breath. In one series of important role in the evaluation and management
1 0 1 patients from Japan, 42 complained of dyspnea of patients with pancreaticopleural fistula. ERCP is
and 29 complained of chest and back pain, whereas useful in delineating the ductal structure, the pseu
only 23 complained of upper abdominal pain ( 1 4) . docyst, and the fistulous connection to the pleura
Th e explanation fo r the lack o f abdominal symptoms through the sinus tract ( 1 2) . The greatest utility for
is that the pancreaticopleural fistula decompresses the ERCP is in defining the precise anatomic relationship
pseudocyst. Weight loss is common in patients with preoperatively so that a direct and expeditious surgi
chronic pancreatic pleural effusions ( 1 2) . cal procedure can be planned. Recently, some patients
Pleural effusion i s usually large, sometimes occu have been successfully treated by placing stents in the
pying the entire hemithorax. In most cases the effu pancreatic duct at the time the ERCP is performed
sion is unilateral and left sided, but approximately ( 1 9,20) . In one series ( 1 8), five of six patients were
20% are unilateral and right sided. Fifteen percent are successfully treated with a stent.
bilateral ( 1 2, 1 4) . If a therapeutic thoracentesis is per The initial therapy of a patient with a pancreatic
formed, the pleural effusion reaccumulates rapidly. pseudocyst and a pleural effusion should probably
In one patient, more than 1 3 L of pleural fluid was be nonoperative. The theory behind conservative
removed during three separate thoracenteses over a therapy is that if pancreatic secretions are minimized,
short period ( 1 5 ) . Because chest symptoms dominate the pseudocyst will regress and the sinus tract will
the clinical picture and some patients have no his close. Accordingly, a nasogastric tube is inserted and
tory of prior pancreatic disease, the diagnosis is easily the patient is given intravenous hyperalimentation.
missed unless the pleural fluid amylase is measured. It is probable that the patients are benefited if they
The diagnosis of a chronic pancreatic pleural effu are given somatostatin or octreotide, a synthetic ana
sion should be suspected in any individual with a log of somatostatin (2 1 ) . Somatostatin has numerous
large pleural effusion who appears to be chronically inhibitory actions on gastrointestinal functions, one
ill or has a history of pancreatic disease or abdomi of which is its inhibitory effect on pancreatic exocrine
nal trauma ( 1 6) . Many patients have no history of secretion ( 1 2) . It has been shown that somatostatin
pancreatic disease ( 1 2) . The best screening test for decreases the output from an external pancreatic
chronic pancreatic pleural effusion is to measure fistula by more than 80% (22) . Some authors rec
the pleural fluid amylase. The pleural fluid amylase ommend serial thoracentesis and subsequent tube
is usually markedly elevated ( > 1 ,000 U/L) ( 1 6) , thoracostomy if the effusions recur, but there is no
whereas the serum amylase may b e normal o r mildly evidence that such procedures are beneficial.
elevated ( 1 6) . An elevated pleural fluid amylase level If after 2 weeks, the patient remains symptomatic
is not diagnostic of pancreatic disease, as discussed in and the pleural fluid continues to accumulate, surgi
Chapter 7. cal intervention should be considered. Approximately
The other main diagnosis to consider in a patient 50% of patients require surgery. Surgery is more
with a chronic pleural effusion with a high amylase likely to be required in patients with more severe
level is malignant disease. Approximately 1 0% of pancreatic disease (2 1 ) . Before surgery, an endoscopic
patients with a malignant pleural effusion have an retrograde pancreatogram and an abdominal CT scan
elevated pleural fluid amylase level (6) . With malig should be performed to aid in planning the surgical
nancy, the cytology is frequently positive. In addition, procedure (23) . For example, if a leak from the duct
it is uncommon to have a malignant pleural effusion or pseudocyst is demonstrated in the distal portion of
with an amylase above 1 ,000 U/L. If there is diffi the gland, distal pancreatectomy will be curative. If a
culty in distinguishing between these two entities, direct pancreatic duct leak is found in the more prox
the differentiation can be made by obtaining amylase imal portion of the gland, without a pseudocyst, a
isoenzymes on the pleural fluid. With malignant effu direct anastomosis between the leak and a Roux-en-Y
sions, the amylase is of the salivary rather than the jej una! loop or a Whipple resection should be consid
pancreatic type ( 1 7) . ered. Conversely, if a large cyst is present within the
Diagnosis can usually b e established by CT of body of the gland, internal drainage should be per
the chest and abdomen, which frequently shows formed either into the stomach or with a Roux-en-Y
both the pseudocyst and the sinus tract ( 1 2) . Mag jej una! loop. Procedures that do not focus on removal
netic resonance imaging is complimentary to CT in of the disrupted portion of the gland or on drainage
C HAPTE R 1 8 I P LE U RAL E F F U S I O N S E C O N DARY TO D I S EASES OF TH E GASTRO I NTEST I N A L TRACT 29 9
of the pseudocyst usually fail. I f the preoperative defect in his diaphragm, he will develop a large pleural
ERCP is unsuccessful, pancreatography may be per effusion as a result of the Bow of Buid from the perito
formed at the time of surgery ( 1 2) . neal to the pleural cavity in the same way that pleural
An alternative approach t o the patient with a effusions develop secondary to ascites due to cirrhosis
pancreatic pseudocyst is to drain the pseudocyst per (see Chapter 9) . Approximately 20% of patients with
cutaneously. Under CT guidance, a catheter is intro pancreatic ascites have a pleural effusion (26) .
duced through the anterior abdominal wall, and then Most patients with pancreatic ascites and pleural
through the anterior and then the posterior wall of the effusion are initially thought to have cirrhosis and
stomach into the pseudocyst cavity. Side holes are cut ascites. The diagnosis is easily established if amylase
into the part of the catheter that lies in the pseudocyst determinations are made on the peritoneal and pleu
and that which lies in the stomach. Maintenance of ral Buid in such patients (26) . Most patients have a
this drainage for 1 5 to 20 days is thought to create a protein level above 3 . 0 g/dL in their ascitic Buid. The
fistulous tract between the pseudocyst and stomach treatment for pancreatic ascites is the same as for pan
akin to surgical marsupialization. In one series, 20 of creatic pleural effusion, except that serial paracenteses
26 patients (77%) were cured by this procedure (24) . rather than serial thoracenteses are performed (26) .
To my knowledge, there are no randomized studies
comparing the results with surgery and with percuta
SU BPHRENIC ABSCESS
neous drainage of pseudocysts.
The prognosis of patients with pancreaticopleural Subphrenic abscess continues t o b e a significant clinical
fistula appears to be favorable ( 1 2) . In the series of problem despite the development of potent antibiotics.
96 patients reviewed by Rockey and Cello ( 1 2) , the
overall mortality rate was 5 % , and 3 patients died
I n c i d e nce
of unrelated illnesses during their follow-up period.
Both patients who died as a direct result of the pan In most large medical centers, between 6 and 15 sub
creatic process were managed conservatively and died phrenic abscesses are seen each year (27-29) . Subphrenic
of sepsis. abscesses are discussed in this chapter because a pleural
In patients with chronic pleural effusions second effusion is present in approximately 80% of cases.
ary to pancreatic disease, the pleural surfaces may
become thickened, and in several patients, decorti
Pathogenesis
cations have been performed (25 ) . However, because
the pleural thickening gradually improves spontane Approximately 80% of subphrenic abscesses follow
ously, decortication should be delayed for at least intraabdominal surgical procedures (30,3 1 ) . Splenec
6 months following definitive treatment of the pan tomy is likely to be complicated by a left subphrenic
creatic disease to ascertain whether the pleural disease abscess (3 1 ) , as is gastrectomy. Deck and Berne (32)
will resolve spontaneously. noted a high incidence of subphrenic abscess after
One rare complication of pancreatic pleural effu exploratory laparotomy for trauma; in their study,
sion is the development of a bronchopleural fistula. 59% of subphrenic abscesses occurred after such an
Kaye ( 1 ) reported one such patient in whom the operation. Overall, approximately 1 % of abdominal
development of the bronchopleural fistula was her operations are complicated by subphrenic abscess
alded by the expectoration of copious quantities of (29) . Sanders (29) reviewed the incidence of sub
clear yellow Buid. In this situation, chest tubes should phrenic abscesses following 1 , 566 abdominal surgi
be inserted immediately to drain the pleural space cal procedures at the Radcliffe Infirmary in 1 965 and
and to protect the lung from the Buid with its high found 1 5 patients with subphrenic abscess. Sanders
enzymatic content. also reviewed the cases of 23 patients with pleural
effusion following intraabdominal surgical proce
dures during the same period. He found that 12 of
Pancreatic Ascites
the 23 patients had definite subphrenic abscesses, and
Some patients with pancreatic disease develop asci he believed that another 5 patients possibly had sub
tes characterized by high amylase and protein levels phrenic abscesses.
(2 1 ,26) . The genesis of the ascites is through leakage of Subphrenic abscess may also occur without
Buid from a pseudocyst directly into the peritoneal cav antecedent abdominal surgical procedures. It may
ity or a sinus tract from the pseudocyst into the perito result from processes such as gastric, duodenal, or
neal cavity. If such a patient should happen to have a appendiceal perforation; diverticulitis; cholecystitis;
3 00 PLE U RAL D I S EASES
pancreauus; or trauma (30) . In such patients, the but may be large, occupying more than 50% of the
diagnosis of subphrenic abscess is frequently not hemithorax (27,28,3 1 -3 3 ).
considered. In one series of 22 patients in whom Most patients with postoperative subphrenic
abscesses occurred without antecedent abdominal abscesses have fever, leukocytosis, and abdominal pain
operations, the diagnosis was established before the (27,28,29) , but frequently no localizing signs or symp
patient's death in only 4 1 % (30) . toms are present. The symptoms and signs ofsubphrenic
The pathogenesis of the pleural effusion associated abscess are variable. In a series of 60 patients, 3 7% had
with subphrenic abscess is probably related to inflam no abdominal pain, 2 1 % had no abdominal tender
mation of the diaphragm. Although Carter and Brewer ness, 1 5% had no temperature elevation greater than
(27) proposed that the pleural effusion arose from 39°C, and 8% had no leukocytosis above 1 O,OOO/mm3
the transdiaphragmatic transfer of abscess material (28) . The interval between the surgical procedure and
by the lymphatic vessels, this hypothesis is unlikely the development of the subphrenic abscess is usually
because fluid from these pleural effusions is only rarely 1 to 3 weeks but can be as long as 5 months (30, 3 1 ) .
culture positive. If the pleural effusion arose from the Examination o f the pleural fluid from patients
transdiaphragmatic transport of abscess material, with subphrenic abscesses usually reveals an exudate
bacteria as well as leukocytes should be transported. with predominantly polymorphonuclear leukocytes.
The diaphragmatic inflammation resulting from the Although the pleural fluid WBC may approach or
adjacent abscess probably increases the permeability even exceed 50,000/mm3, the pleural fluid pH and
of the capillaries in the diaphragmatic pleura and glucose level remain above 7.20 and 60 mg/dL,
causes pleural fluid to accumulate. respectively. It is distinctly uncommon for the pleural
fluid to become infected (27) . However, empyemas
C l i n ical M a n ifestati o n s have resulted from contamination of the pleural space
when the abscesses were drained percutaneously (34) .
The clinical picture of a patient with subphrenic
abscess can be dominated by either chest or abdomi
D i a g nosis
nal symptoms. In the series of 1 2 5 cases of Carter and
Brewer (27) , chest findings dominated the clinical The diagnosis of subphrenic abscess should be con
picture in 44% of patients. The main chest symptom sidered in any patient who develops a pleural effusion
is pleuritic chest pain. Radiographic abnormalities several days or more after an abdominal surgical proce
include pleural effusion, basal pneumonitis, com dure or in any other patient who has an undiagnosed
pression atelectasis, and an elevated diaphragm on the exudative pleural effusion containing predominantly
affected side. Pleural effusions occur in 60% to 80% polymorphonuclear leukocytes. The chest radio
of patients and are usually small to moderate in size, graphs from such a patient are shown in Figure 1 8. 1 .
A B
inserted the catheters under CT guidance and left 0 . 5 cm can be readily appreciated. Abdominal ultra
the catheters in place for a mean of 6 days (39) . sound studies can also identify Buid-filled intrahe
Mortality rates among patients with subphrenic patic lesions, but because CT scanning provides more
abscesses remain high, ranging from 20% to 45% (27, precise anatomic information, it is the procedure of
28,30,33) . Because much of the mortality is due choice (43 ) . Not all Buid-filled intrahepatic lesions
to delayed diagnosis or lack of a diagnosis before are pyogenic abscesses; cysts, hematomas, heman
autopsy, the possibility of a subphrenic abscess must giomas, and amebic abscesses can produce identical
be considered in every patient with an exudative findings on ultrasound studies and CT scans. The
pleural effusion containing predominantly polymor definitive diagnosis can be established by percutane
phonuclear leukocytes. In such patients, heavily pen ous aspiration guided by CT scanning or ultrasound.
etrated abdominal radiographs should be examined
for extravisceral gas, and one should consider obtain
ing an abdominal CT scan. Treatment
Esophageal rupture should always b e considered in With esophageal perforation secondary to esophago
the differential diagnosis of pleural effusions, because scopic examination, the endoscopist usually does not
if this entity is not rapidly treated, the mortality rate realize that the esophagus has been perforated (54) .
rapidly increases. However, patients usually report persistent chest or
epigastric pain within several hours of the procedure
( 5 5 ) , and such complaints should serve as indications
I n c i d e nce
for an emergency contrast study of the esophagus.
Esophageal perforation is uncommon. Michel et al. Patients with spontaneous rupture of the esophagus
(49) reported only 85 cases at the Massachusetts usually have a history of vomiting, followed by chest
General Hospital over a 2 1 -year period, whereas pain, and they frequently describe a sensation of tear
Reeder et al. ( 5 0) found 4 1 cases at The University ing or bursting in the lower part of the chest or the
of Chicago Hospitals over a 1 4-year period. Ryom epigastrium (56) . The chest pain is characteristically
et al. (5 1 ) reported that there were 2 8 6 cases of excruciating and is often unrelieved by opiates. Small
esophageal perforation in Denmark between 1 9 97 amounts of hematemesis are present in more than
and 200 5 . 50% of these patients (56) . Dyspnea is frequently a
prominent symptom. The presence of subcutaneous
emphysema that first appears in the suprasternal notch
Path ogenesis a n d Path ophysiologic
suggests esophageal perforation, but this appears late in
Mech a n i s m s
the course of perforation. Abbott et al. (56) , in a series
Esophageal perforation most commonly arises as a of 47 patients, found that only four (9%) had subcuta
complication of esophagoscopic examination. In one neous emphysema within the first 4 hours. The clinical
series of 1 08 cases, 67% occurred as a complication picture may be much less dramatic than that described.
of esophagoscopy (52) . Esophageal perforation is Chandrasekhara and Levitan (57) described a patient
particularly common with esophagoscopy when one with a ruptured esophagus and symptoms that were
has attempted to remove a foreign body or to dilate present for 5 days, who had only mild distress.
an esophageal stricture (52) . Overall, between 0 . 1 5 % In patients with esophageal perforation, the chest
and 0.70% of all esophagoscopic examinations are radiograph reveals a pleural effusion in approximately
complicated by esophageal perforation (49) . The 60% and a pneumothorax in approximately 25%
insertion of a Blakemore tube for esophageal varices (49). Most patients with spontaneous rupture have
can also be complicated by esophageal rupture; this a pleural effusion (58). The pleural effusion is usu
mechanism accounted for 1 1 % of all esophageal per ally left sided, but it may be right sided or bilateral.
forations in one series (49). Frequently, the diagnosis Other radiographic findings may include widening of
of esophageal perforation is missed in patients with the mediastinum and air visible within the medias
Blakemore tubes because they are so ill with multiple tinal compartments. The chest CT scan is useful in
problems (49) . suggesting the diagnosis of esophageal perforation
Esophageal perforations may also arise from for because it demonstrates periesophageal air tracks sug
eign bodies themselves, carcinomas, gastric intubation, gestive of esophageal perforation in most cases (59).
chest trauma, and chest operations. Finally, esophageal
rupture may occur as a complication of vomiting
D i a g nosis
(Boerhaave syndrome) . Spontaneous rupture almost
always involves the lower esophagus, just above the Because the mortality rate approaches 60% (58) when
diaphragm. treatment is delayed for more than 24 hours, the diag
The clinical symptoms of esophageal perfora nosis of esophageal rupture should be entertained any
tion are due to contamination of the mediastinum time one sees a patient with an exudative pleural effu
by oropharyngeal contents, which produces an acute sion, particularly when the patient appears acutely ill.
mediastinitis. When the mediastinal pleura ruptures, Examination of the pleural fluid is helpful in suggest
a pleural effusion develops, frequently complicated ing the diagnosis of esophageal perforation because
by a pneumothorax. Most of the morbidity from it is characterized by (a) a high amylase level, (b) a
esophageal perforation is due to the infection of the low pH, (c) the presence of squamous epithelial cells,
mediastinum and the pleural space by the oropharyn (d) ingested food particles, and (e) multiple patho
geal bacterial flora (53). gens on smear or culture.
3 04 PLE U RAL D I S EASES
An elevated pleural fluid amylase level appears to be Another characteristic of the pleural fluid from
the best indication of esophageal rupture. In the experi patients with esophageal rupture is that smear and
mental model, the pleural fluid amylase level is elevated cultures frequently reveal multiple organisms. The
within 2 hours of esophageal rupture (53). In one clini diagnosis of esophageal rupture should be considered
cal series, all seven patients with esophageal rupture had in any patient with a polymicrobial empyema, par
elevated pleural fluid amylase levels (56) . The origin of ticularly when the daily pleural fluid output is high.
the amylase is salivary rather than pancreatic because the The diagnosis of esophageal perforation is estab
saliva, with its high amylase content, enters the pleural lished when esophageal disruption is confirmed
space through the defect in the esophagus (60) . To my by contrast studies of the esophagus. The contrast
knowledge, in only two reported cases have pleural fluid agent of choice is probably meglumine and ioxaglate
amylase levels been within normal limits with esopha sodium (Hexabrix, 320 mg/mL) (68) . Barium has a
geal perforation and pleural effusion (6 1 ,62) . One of greater radiographic densiry, better mucosa! adher
these patients had Sjogren's syndrome and essentially ence, and causes minimal irritation to the tracheo
no production of saliva (62) . The other had a chronic bronchial tree, but it is not absorbed once it leaks into
perforation due to esophageal carcinoma (6 1 ) . the mediastinum or pleura and produces a marked
Th e pleural fluid p H i s usually decreased with inflammatory reaction in the pleura. When water
esophageal rupture (63,64) . In fact, Dye and Laforet soluble agents, such as Hexabrix or meglumine and
(64) concluded that a pleural fluid pH below 6.0 was diatrizoate sodium (Gastrografin) , are injected in the
highly suggestive of esophageal rupture and attributed pleural space, they are almost completely absorbed
the low pleural fluid pH to the leakage of acidic gas after 24 hours and neither creates much of an inflam
tric juice through the esophageal tear. Both of these matory response (68) . Hexabrix is considered the
conclusions appear to be wrong. Patients with severe agent of choice because Gastrografin creates marked
infections of the pleural space and an intact esophagus bronchospasm when it is aspirated and Hexabrix does
frequently have a pleural fluid pH below 6.0. not create such a reaction (68) . The contrast studies
Good et al. (65) demonstrated, in an experimental are positive in approximately 8 5 % of patients (49) .
model, that the pleural fluid pH falls j ust as rapidly If the perforation is small or has already closed spon
after esophageal perforation when the esophagogas taneously, the esophagogram may not be diagnostic.
tric junction is ligated. These authors concluded that It has been suggested that contrast studies of the
leukocyte metabolism was the major contributor to esophagus be done in the decubitus position when
the low pleural fluid pH with esophageal rupture. perforation is suspected (69) . In this position, the con
Nevertheless, the presence of a pleural fluid pH below trast material fills the whole length of the esophagus
7.00 increases the likelihood that the patient has a and thereby allows the actual site of the perforation
ruptured esophagus. and its interconnecting cavities to be demonstrated
Another useful test in diagnosing esophageal per in almost all patients (69) .
foration is examination of the Wright's stain of the If the esophageal perforation is not demonstrated
pleural fluid for squamous epithelial cells. Eriksen (66) by the contrast study of the esophagus, chest CT scan
demonstrated the presence of squamous epithelial cells may facilitate the diagnosis (70) . White et al. (70) per
in the pleural fluid from all 1 4 patients with esopha formed chest CT scans on 1 2 patients with esopha
geal perforation. Again, as with amylase, the squa geal perforation. They found esophageal thickening in
mous epithelial cells enter the pleural space through 9 patients, periesophageal fluid in 1 1 patients, extralu
the esophageal perforation. Obviously, the demonstra minal air in 1 1 patients, and pleural effusion in 9.
tion of food particles in pleural fluid is diagnostic of The site of the perforation was visible on CT scan in
esophageal perforation. two patients. The finding that most commonly pointed
The pleural fluid glucose level with esophageal rup to esophageal rupture was extraluminal air (70).
ture is usually decreased. If, however, the patient has
recently ingested food with a high glucose level, the glu
Treatment
cose level in the pleural fluid may exceed 500 mg/dL
(67) . The only two other conditions associated with The treatment of choice for esophageal rupture is explo
such a high pleural fluid glucose level are diabetes, ration of the mediastinum with primary repair of the
when the serum glucose level is very high, and instances esophageal tear and drainage of the pleural space and
where intravenous fluids with high glucose levels enter mediastinum (50,7 1 ,72) . Parenteral antibiotics should
the pleural space. be given to treat the mediastinitis and pleural infection.
C HAPTE R 1 8 I P LE U RAL E F F U S I O N S E C O N DARY TO D I S EASES OF TH E GASTRO I NTEST I N A L TRACT 305
Although conservative treatment consisting o f antibiot 1 0 mm in thickness on the decubitus film, a diag
ics and nasogastric suction is adequate in some patients nostic thoracentesis should be performed to rule out
with esophageal perforation (72) , in patients with pleural pleural infection. Although pulmonary embolization
effusion or pneumothorax complicating esophageal per and subphrenic abscess can cause pleural effusion
foration, mediastinal exploration should be performed. postoperatively, most effusions occurring within the
It is important to perform the mediastinal exploration first 72 hours of abdominal surgery are not due to
as soon as possible after the diagnosis is established these factors and resolve spontaneously.
because a delay of even 1 2 hours increases the mortal Patients with advanced epithelial ovarian cancer
ity rate (58) . However, primary repair can be attempted sometimes undergo removal of the peritoneum from
if the perforation has been present for more than the diaphragm or diaphragmatic resection to render the
24 hours. In one series, the mortality rate for 1 6 patients patients more likely to respond to chemotherapy (80) .
operated upon with primary repair 2 to 17 days after In a series (80) of 59 patients who underwent one of
the perforation was only 1 3 % (73) . If primary repair is these two procedures, the incidence of pleural effusion
not possible because the damaged tissue cannot hold the in the 40 patients who did not have a pleural effusion
sutures, the patient can be managed with T-tube intu preoperatively and who were not treated with tube
bation of the esophageal defect (74) . Operative repair thoracostomy was 60%. The effusions tend to resolve
of the esophageal perforation can be performed tho with time and drainage is indicated only if the patient
racoscopically if the patient is relatively stable and has becomes symptomatic (80) .
only mild mediastinal inflammation (75) . Nonoperative
management of iatrogenic perforations is frequently suc
D I A P H RAG M AT I C H E R N IA
cessful, provided the mediastinal and pleural spaces are
not soiled with ingested food or bacteria (76,77) . Hernias through the diaphragm are important in the
differential diagnosis of pleural effusions from two
viewpoints. First, they may mimic a pleural effusion.
A B D O M I N A L S U RG I CA L P R O C E D U R E S
Second, pleural effusions are usually present in patients
Th e incidence o f small pleural effusions after abdomi with a strangulated diaphragmatic hernia (8 1 ) .
nal operations is high. George and I (78) reported Diaphragmatic hernia should b e considered
a series of 200 patients who had bilateral decubitus whenever an apparent pleural effusion has an atypi
chest radiographs 48 to 72 hours following abdominal cal shape or location (Fig. 1 8 .3) . Air in the herniated
surgical procedures. Pleural effusions were identified intestine is usually the clue to this diagnosis. Occa
in 97 patients (49% ) . In a more recent series, Nielsen sionally, an upper gastrointestinal series and a small
et al. (79) reported that 89 of 1 2 8 patients (69%) bowel follow-through study in conjunction with a
undergoing upper abdominal surgery had pleural barium enema are necessary for accurate diagnosis.
effusions in the first 4 days postoperatively. Most The possibility of a strangulated diaphragmatic
of the pleural effusions are small; only 21 patients hernia should always be considered in patients with
(22%) in our series had pleural fluid that measured a left pleural effusion and signs of an acute abdomi
more than 1 0 mm in thickness on the decubitus nal catastrophe (82,83) . At least 90% of strangulated
films (78) . Larger left-sided pleural effusions are par diaphragmatic hernias are traumatic in origin, and at
ticularly common after splenectomy. Postoperative least 9 5 % are on the left side because the liver protects
pleural effusions are more common in patients under the right diaphragm. Strangulation can occur months
going upper abdominal surgical procedures (78), to years after the original injury, which is usually an
in patients with postoperative atelectasis (78,79) , automobile accident. Strangulation typically occurs
and in those with free abdominal fluid at the time suddenly and progresses rapidly. Left shoulder pain
of operation (78) . In our series (78 ) , a thoracentesis is generally present due to diaphragmatic irritation.
was performed on 20 patients, and in 1 6 of these, Serosanguineous exudative pleural fluid with pre
the pleural fluid was an exudate. The pleural effusions dominantly polymorphonuclear leukocytes is almost
in all but a single patient, who had a staphylococ always present. The diagnosis is usually suggested by
cal pleural infection, resolved spontaneously without air-fluid levels in the viscera strangulated in the left
any specific therapy. In summary, pleural effusions pleural space. Contrast studies of the gastrointestinal
frequently occur after abdominal surgical procedures tract are sometimes necessary to make the diagnosis.
and are usually related to diaphragmatic irritation or Immediate surgical treatment is imperative to prevent
atelectasis. If the pleural effusion measures more than gangrene of the strangulated viscera (8 1-83) .
3 06 PLE U RAL D I S EASES
the pleural fluid having a much lower bilirubin than one had subdiaphragmatic pathology, including 4 with
would anticipate (93) . In one patient who developed hematomas, 1 with a biloma, and 2 with abscesses
a pleural effusion as a complication of percutaneous (97) . Accordingly, patients with enlarging pleural effu
biliary drainage, the pleural fluid bilirubin was only sions after liver transplantation should be evaluated for
2 . 1 mg/dL (93) . However, the pleural fluid bilirubin subdiaphragmatic pathology.
has exceeded 25 mg/dL in some patients (93) . There is little data on characteristics of the pleural
The diagnosis of a bilious pleural effusion should fluid after liver transplantation. In one series ( 1 00),
be suspected in any patient with an obstructed biliary thoracentesis was done in 37 patient post lung trans
system. It is important to remember that the pleural plantation and the fluid was exudative in 1 6 (43%)
fluid may not appear to be bile, although the ratio and the bacterial cultures were positive in 7 ( 1 9%) .
of the pleural fluid to serum bilirubin is greater than The pleural effusion that occurs after liver trans
1 .0 (93) . The appropriate treatment for this condi plantation can be largely prevented if a fibrin seal
tion is the reestablishment of the biliary drainage. ant is sprayed on the undersurface of the diaphragm
Most patients who have a bilious pleural effusion around the insertion of the liver ligaments at the time
after trauma require decortication and diaphragmatic of transplantation. When Uetsuji et al. ( 1 0 1 ) used
repair (9 1 ) . The incidence of empyema with bilious the fibrin sealant in 25 liver transplant patients, none
pleural effusions approaches 50%, and one should developed a pleural effusion postoperatively.
constantly be aware of this complication.
P LE U RA L E F F U S I O N S AS
P L E U RA L E F F U S I O N S A FT E R LIVE R CO M P L I CAT I O N S OF TH E RA PY
TRA N S P LA N TATI O N F O R TU M O RS O F TH E LIV E R
Most patients who undergo an orthotopic liver trans Several different therapies fo r hepatic tumor are
plantation develop a pleural effusion postoperatively. associated with the development of pleural effu
Golfieri et al. (96) reviewed the chest x-rays of 300 sion. When hepatic tumors are treated surgically,
consecutive patients who had undergone 333 liver there is a high incidence of pleural effusion. In
transplants over a 1 1 -year period. They reported that one study ( 1 02) , the incidence of pleural effu
68% of the patients had a pleural effusion and the sion with an abdominal approach was 43% of
effusion occupied more than 25% of the hemithorax in 2 8 , whereas the incidence was 73% of 70 with
2 1 patients (7%). The effusion was unilateral on the a thoracoabdominal approach. With the tho
right side in 1 53 patients and bilateral in 53. The effu racoabdominal approach, 1 7% of the patients
sion lasted for more than 1 week in 5 9 patients. Only required thoracentesis and a mean of three thora
37 of the patients underwent a thoracentesis. Other centeses was done on each patient ( 1 02) . Pleural
authors have also reported a high prevalence of pleural effusions are also common after magnetic resonance
effusion following liver transplantation (97,98) . imaging-guided laser-induced thermotherapy. In
The pleural effusion after liver transplantation one study of 899 patients in which 2, 1 32 procedures
may be large. In one series of liver transplants in 48 were performed, 1 7 1 of the procedures ( 8 . 1 %) were
children, effusions that were large enough to cause complicated by pleural effusion and thoracentesis
clinically detectable respiratory compromise occurred was required in 1 6 (0.8%) ( 1 03) . Taj ima et al. ( 1 04)
in 23 ( 1 9 right sided and 4 left sided) (99) . Fifteen reported that the incidence of pleural effusion fol
of the patients in this latter series were treated with lowing transcatheter hepatic chemoembolization
chest tubes (99) . through the inferior phrenic vein was 41 % in 44
The pathogenesis of the pleural effusions after liver patients who underwent this procedure.
transplantation is not definitely known. It has been
suggested that the effusion is due to injury or irritation
of the right hemidiaphragm caused by the extensive REFERENCES
right upper quadrant dissection and retraction. The I. Kaye M O . Pleuropulmonary complications of pancreatitis.
natural history of a pleural effusion after transplanta Thorax. 1 968;23 :297-306.
tion is that it increases in size over the first 3 postop 2 . Gumaste V, Singh V, Dave P. Significance of pleural effu
sion in patients with acute pancreatitis. Am J Gastroenterol.
erative days and then gradually resolves over a period
1 992;87: 8 7 1 -874.
ranging from several weeks to several months (97) . In 3 . Lankisch PG, Groge M, Becher R. Pleural effusions: a new
one series, the effusion increased in size after the first negative prognostic parameter for acute pancreatitis. Am J
3 days in 10 of 42 patients. Seven of these 10 patients Gastroenterol. 1 994; 8 9 : 1 849- 1 8 5 1 .
308 PLE U RAL D I S EASES
4. Heller SJ, Noordhoek E, Tenner SM, et al. Pleural effu 27. Career R, Brewer LA. Subphrenic abscess: a thoracoabdominal
sion as a predictor of severity in acute pancreatitis. Pancreas. clinical complex. Am ] Surg. 1 964; 1 0 8 : 1 6 5- 1 74.
1 997; 1 5 :222-225 . 2 8 . DeCosse ]], Poulin TL, Fox PS, et al. Subphrenic abscess. Surg
5 . Roseman O M , Kowlessar O D , Sleisenger MH. Pulmonary Gynecol Obstet. 1 974; 1 3 8 : 84 1 -846.
manifestations of pancreatitis. N Engl J Med. 1 960;263: 29. Sanders RC. Post-operative pleural effusion and subphrenic
294-296. abscess. Clin Radio!. 1 970;2 l :308-3 1 2 .
6. Light RW, Ball WC. Glucose and amylase in pleural effusions. 3 0 . Sherman NJ, Davis JR, Jesseph J E . Subphrenic abscess: a con
JAMA. 1 973;225 :257-260. tinuing hazard. Am ] Surg. 1 969; 1 1 7 : 1 1 7- 1 23 .
7. Makela A, Kuusi T, Nuutinen P, et al. Phospholipase A2 activ 3 1 . Connell TR, Stephens DH, Carlson H C , et al. Upper abdom
ity in body fluids and pancreatic tissue in patients with acute inal abscess: a continuing and deadly problem. AJR Am J
necrotising pancreatitis. EurJ Surg. 1 999; 1 6 5 :3 5-42. Roentgenol. 1 980; 1 34:759-76 5 .
8 . Miller TA, Lindenauer SM, Frey CF, et al. Pancreatic abscess. 3 2 . Deck K B , Berne TV. Selective management of subphrenic
Arch Surg. 1 974; 1 08 : 545-5 5 1 . abscesses. Arch Surg. 1 979; 1 1 4 : 1 1 65-1 1 6 8 .
9 . Kolmannskog F, Kolbensrvedt A, Aakhus T. Computed 33. van der Sluis RF. Subphrenic abscess. Surg Gynecol Obstet.
tomography in inflammatory mass lesions following acute 1 984; 1 5 8 :427-43 5 .
pancreatitis. J Comput Assist Tomogr 1 9 8 1 ; 5 : 1 69- 1 72 . 3 4 . Samelson SL, Ferguson M K . Empyema following percutane
1 0. Tombroff M, Loicq A , De Koster J - P, e t al. Pleural effusion ous catheter drainage of upper abdominal abscess. Chest. 1 992;
with pancreaticopleural fistula. Br MedJ 1 973; 1 :330-33 1 . 1 02 : 1 6 1 2- 1 6 1 4 .
1 1 . She tty AN . Pseudocysts o f the pancreas: an overview. South 3 5 . Perera MR, Kirk A , Noone P. Presentation, diagnosis and
Med] 1 9 8 0;73 : 1 239-1 242. management of liver abscess. Lancet. l 980;3:629-632 .
12. Rockey DC, Cello JP. Pancreaticopleural fistula. Report of 7 36. Alexander E S , Proto AV, Clark RA. C T differentiation o f
patients and review ofthe literature. Medicine. l 990;69:332-344. subphrenic abscess and pleural effusion. AJR Am J Roentgenol.
1 3 . Anderson WJ, Skinner DB, Zuidema GD, et al. Chronic pancre 1 983; 1 4 5 :47-5 1 .
atic pleural effusions. Surg Gynecol Obstet. 1 973; 1 37:827-830. 37. Brook I , Frazier EH. Microbiology o f subphrenic abscesses: a
14. Uchiyama T, Suzuki T, Adachi A, et al. Pancreatic pleural effu 1 4-year experience. Am Surg. 1 999;65: 1 049- 1 053.
sion: case report and review of 1 1 3 cases in Japan. Am J Gas 3 8 . Bufalari A, Giustozzi G, Moggi L. Postoperative intraabdomi
troenterol. 1 992;87:387-39 1 . nal abscesses: percutaneous versus surgical treatment. Acta
1 5 . Miridjanian A, Ambruoso VN , Derby BM, et al. Massive Chir Belg. 1 996;96: 1 97-200.
bilateral hemorrhagic pleural effusions in chronic relapsing 39. Voros 0, Gouliamos A, Kotoulas G, et al. Percutaneous drain
pancreatitis. Arch Surg. l 969;98:62-66. age of intraabdominal abscesses using large lumen cubes under
16. Pottmeyer EW III, Frey CF, Matsuno S . Pancreaticopleural computed tomographic control. EurJ Surg. 1 996; 1 62:895-898.
fistulas. Arch Surg. 1 987; 1 22:648-654. 40. Chen SC, Lee YT, Yen CH, et al. Pyogenic liver abscess in the
1 7. Kramer MR, Saidana MJ, Cepero RJ, et al. High amylase elderly: clinical features, ourcomes and prognostic factors . Age
levels in neoplasm-related pleural effusion. Ann Intern Med. Ageing. 2009; 1 9 8 : 1 64- 1 72 .
1 989; 1 1 0: 5 67-569. 4 1 . C h u K-M, Fan S-T, Lai ECS, et a l . Pyogenic liver abscess. Arch
18. Khan AZ, Ching R, Morris-Stiff G, et al. Pleuropancreatic Surg. 1 996; 1 3 1 : 1 48- 1 5 3 .
fistulae: specialist center management. J Gastrointest Surg. 4 2 . Buchman T G , Zuidema GD. The role of computerized tomo
2009; 1 3 : 3 54-3 56. graphic scanning in the surgical management of pyogenic
19. Safadi BY, Marks JM. Pancreatic-pleural fistula: the role hepatic abscess. Surg Gynecol Obstet. 1 9 8 1 ; 1 5 3 : 1 -9 .
of ERCP in diagnosis and treatment. Gastrointest Endosc. 43. Ruiz-Hernandez ]], Leon-Mazorra M, Conde-Marcel A , e t al.
2000;5 1 :2 1 3-2 1 5 . EurJ Gastroenterol Hepatol. 2007; 1 9 : 8 53-8 5 8 .
20. Chebli JM, Gaburri PD, Meirelles D e Souza AF, et al. Internal 4 4 . Sarr M G , Zuidema GD. Splenic abscess: presentation, diag
pancreatic fistulas: proposal of a management algorithm based nosis and treatment. Surgery. 1 9 82;92:480-4 8 5 .
on a case series analysis. ] Clin Gastroenterol. 2004;3 8:795-800. 4 5 . Ho HS, Wisner DH. Splenic abscess in the intensive care unit.
2 1 . Parekh D, Segal I. Pancreatic ascites and effusion. Risk factors Arch Surg. 1 993; 1 28 : 842-848.
for failure of conservative therapy and the role of octreotide. 46. Johnson JF, Raff MJ, Barnwell PA, et al. Splenic abscess
Arch Surg. 1 992; 1 27:707-7 1 2 . complicating infectious endocarditis. Arch Intern Med.
2 2 . Pederzoli P, Bassi C, Falconi M, e t al. Conservative treatment 1 983; 1 43:905-9 1 2 .
of external pancreatic fisculae with parenteral nutrition along 4 7 . Tikkakoski T, Siniluoto T, Paivansalo M, e t al. Splenic abscess.
or in combination with continuous intravenous infusion of Imaging and intervention. Acta Radio!. 1 992;33 : 5 6 1-565.
somatostatin, glucagon or calcitonin. Surg Gynecol Obstet. 4 8 . Green BT. Splenic abscess: report of six cases and review of the
1 986; 1 63 :428-432. literature. Am Surg. 200 1 ;67:80-8 5 .
23. Krasnow AZ, Collier BO, Isitman AT, et al. The value of 4 9 . Michel L, Grillo HC, Malt RA . Operative and nonoperative man
preoperative imaging techniques in patients with chronic pan agement of esophageal perforations. Ann Surg. 1 98 1 ; 1 94:5 7-63.
creatic pleural effusions. Int] Pancreatol. 1 9 87;2:269-276. 50. Reeder LB, DeFilippi VJ, Ferguson MK. Current results of ther
24. Lang EK, Paolini RM , Pottmeyer A. The efficacy of pal apy for esophageal perforation. Am ] Surg. 1 995; 1 69:6 1 5-6 17.
liative and definitive percutaneous versus surgical drainage of 5 1 . Ryom P, Ravn JB, Penninga L, et al. Aetiology, treatment and
pancreatic abscesses and pseudocysts: a prospective study of mortality after oesophageal perforation in Denmark. Dan Med
8 5 patients. South Med] 1 9 9 1 ;84:5 5-64. Bull. 20 1 1 ; 5 8 :A4267.
25. Shapiro DH, Anagnostopoulos CE, Dineen JP. Decortication 52. Keszler P, Buzna E. Surgical and conservative management of
and pleurectomy for the pleuropulmonary complications of esophageal perforation. Chest. 1 98 1 ; 80: 1 58-162.
pancreatitis. Ann Thorac Surg. 1 970;9 :76-80 . 5 3 . Maulitz RM , Good JT Jr, Kaplan RL, et al. The pleuropul
26. Lipsett PA, Cameron J L . Internal pancreatic fistula. Am J Surg. monary consequences of esophageal rupture: an experimental
1 992; 1 63:2 1 6-220. model. Am Rev Respir Dis. 1 979; 1 20 :363-367.
C HAPTE R 1 8 I P LE U RAL E F F U S I O N S E C O N DARY TO D I S EASES OF TH E GASTRO I NTEST I N A L TRACT 309
5 4 . Quintana R , Bartley TD, Wheat M W Jr. Esophageal perfora 78. Light RW, George RB. Incidence and significance of pleural
tion: analysis of 10 cases. Ann Thorac Surg. 1 970; 1 0 :45-5 3 . effusion afrer abdominal surgery. Chest. 1 976;69:62 1 -626.
5 5 . Skinner DB, Little A G , DeMeester TR. Management of 79. Nielsen PH, Jensen SB, Olsen AD . Postoperative pleural
esophageal perforation. Am ] Surg. 1 9 80; 1 3 9 :760-764. effusion following upper abdominal surgery. Chest. 1 9 89;
56. Abbott OA, Mansour KA, Logan WD, et al. Atraumatic 96: 1 1 33-1 1 3 5 .
so-called "spontaneous" rupture of the esophagus. J Thorac 8 0 . Eisenhauer EL, D'Angelica M I , Abu-Rustum NR, e t al. Inci
Cardiovasc Surg. 1 970; 5 9 : 67-83 . dence and management of pleural effusions afrer diaphragm
57. Chandrasekhara R , Levitan R . Spontaneous rupture of the peritonectomy or resection for advanced Mullerian cancer.
esophagus. Arch Intern Med. 1 970; 1 26: 1 008- 1 009. Gynecol Oncol. 2006; 1 03 : 8 7 1 -877.
58. Finley RJ, Pearson FG, Weisel RD, et al. The management 8 1 . Niwa T, Nakamura A, Kato T, et al. An adult case of Boch
of non-malignant intrathoracic esophageal perforations. Ann dalek hernia complicated with hemothorax. Respiration. 2003;
Thorac Surg. 1 9 80;30: 575-5 8 1 . 70:644-646.
5 9 . Ghanem N , Altehoefer C , Springer 0 , e t al. Radiological find 82. Keshishian JM, Cox SA. Diagnosis and management of
ings in Boerhaave's syndrome. Emerg Radio/. 2003 ; 1 0 : 8- 1 3 . strangulated diaphragmatic hernias. Surg Gynecol Obstet.
6 0 . Sherr HP, Light RW, Merson M H , e t al. Origin o f pleural 1 962; 1 1 5 :626-632.
fluid amylase in esophageal rupture. Ann Intern Med. 1 972; 83. Atonchick JM, Epstein DM, Gefter WB, et al. Chronic trau
76:98 5-9 86. matic diaphragmatic hernia: the significance of pleural effusion.
6 1 . Faling LJ, Pugatch RD, Robbins AH . Case report: the diag Radiology. 1 98 8 ; 1 68:675-678.
nosis of unsuspected esophageal perforation by computed 84. Casaccia G, Crescenzi F, Palamides S, et al. Pleural effusion
romography. Am ] Med Sci. 1 9 8 1 ;28 1 :3 1 -34. requiring drainage in congenital diaphragmatic hernia: incidence,
62. Rudin JS, Ellrodt AG, Phillips EH. Low pleural fluid amylase aetiology and treatment. Pediatr Surg Int. 2006;22:585-5 88.
associated with spontaneous rupture of the esophagus. Arch 8 5 . Saks BJ, Kilby AE, Dietrich PA, et al. Pleural and mediastinal
Intern Med. 1 983; 1 43 : 1 034- 1 03 5 . changes following endoscopic injection sclerotherapy of
63. Good J T Jr, Taryle DA, Maulitz RM, e t al. The diagnostic esophageal varices. Radiology. 1 9 83; 1 49 :639-642.
value of pleural fluid pH. Chest. 1 980;78 : 5 5-5 9 . 86. Bacon BR, Bailey-Newton RS, Connors AF Jr. Pleural effusions
6 4 . Dye RA , Laforet E G . Esophageal rupture: diagnosis b y pleural after endoscopic variceal sclerotherapy. Gastroenterology. 1 9 8 5 ;
fluid pH. Chest. 1 974;66:454-456. 8 8 : 1 9 1 0- 1 9 1 4 .
6 5 . Good JT Jr, Antony VB, Reller LB, et al. The pathogenesis of 8 7 . Parikh SS, Amarapurkar D N , Dhawan P S , et al. Development
the low pleural fluid pH in esophageal rupture. Am Rev Respir of pleural effusion after sclerotherapy with absolute alcohol.
Dis. 1 9 83; 1 27:702-704. Gastrointest Endovasc. 1 993;39 :404-405.
66. Eriksen KR. Oesophagopleural fistula diagnosed by micro 8 8 . Edling JE, Bacon BR. Pleuropulmonary complications of
scopic examination of pleural fluid. Acta Chir Scand. 1 964; endoscopic variceal sclerotherapy. Chest. 1 99 1 ;99: 1 252- 1 257.
1 2 8:77 1 -777. 89. Sethy PK, Kochhar R, Behera D, et al. Pleuropulmonary com
67. Almoosa KF, Wardell N, Javaheri S. Elevated glucose in pleu plication of esophageal variceal sclerotherapy with absolute
ral effusion: an early clue ro esophageal perforation. Chest. alcohol. J Gastroenterol Hepatol. 2003; 1 8 : 9 1 0-9 1 4 .
2007; 1 3 1 : 1 567- 1 569. 90. Arai H, Abe T, Takayama H, e t a l . Respirarory effects of
68. Ginai AZ. Experimental evaluation ofvarious available contrast balloon occluded retrograde transvenous obliteration of gas
agents for use in the gastrointestinal tract in case of suspected tric varices-a prospective controlled study. J Gastroenterol
leakage: effects on pleura. Br J Radio/. 1 986;59:8 87-894. Hepatol. 20 1 1 ;26: 1 3 8 9- 1 394.
69. Demeester TR. Perforation of the esophagus. Ann Thorac Surg. 9 1 . Ivatury RR, O'Shea J, Rohman M . Post-traumatic thoraco
l 986;42:23 1 -232. biliary fistula. J Trauma. l 984;24:43 8-44 1 .
70. White CS, Templeron PA, Attar S . Esophageal perforation: 92. Delco F, Domenigheti G , Kauzlaric D , et al. Spontaneous
CT findings. AJR A m j Roentgenol. 1 993; 1 60:767-770. biliothorax (thoracobilia) following cholecystopleural fistula
7 1 . Lawrence DR, Ohri SK, Moxon RE, et al. Primary esopha presenting as an acute respirarory insufficiency. Chest. 1 994;
geal repair for Boerhaave's syndrome. Ann Thorac Surg. 1 999; 1 06 : 9 6 1 -963.
67: 8 1 8-820. 93. Strange C, Allen ML, Freedland PN, et al. Biliopleural fistula
72. Bufkin BL, Miller JI Jr, Mansour KA. Esophageal perforation: as a complication of percutaneous biliary drainage: experi
emphasis on management. Ann ThoracSurg. 1 996;6 1 : 1 447-1 45 1 . mental evidence for pleural infl a mmation. Am Rev Respir Dis.
73. Jougon J , McBride T, Delcambre F, e t al. Primary esophageal 1 9 8 8 ; 1 37:959-96 1 .
repair for Boerhaave's syndrome whatever the free interval 94. Dasmahapatra HK, Pepper JR. Bronchopleurobiliary fistula:
berween perforation and treatment. Eur J Cardiothorac Surg. a complication of intrahepatic biliary scent migration. Chest.
2004;25 :475-479. 1 9 88;94: 874-875 .
74. Naylor AR, Walker WS, Dark ], et al. T tube intubation in the 9 5 . Cunningham LW, Grohman M, Paz HL, e t al. Cholecysto
management of seriously ill patients with oesophagopleural pleural fistula with cholelithiasis presenting as a right pleural
fistulae. Br J Surg. 1 990;77:40-42. effusion. Chest. 1 990;97:75 1 -752.
75. Cho JS, Kim YD , Kim JW, et al. Thoracoscopic primary 96. Golfieri R, Giampalma E, Morselli Labate AM, et al. Pul
esophageal repair in patients with Boerhaave's syndrome. Ann monary complications of liver transplantation: radiologi
Thorac Surg. 20 1 1 ;9 1 : 1 5 52- 1 5 5 5 . cal appearance and statistical evaluation of risk facrors in
7 6 . Hasan S, Jilaihawi AN, Prakash D . Conservative management 300 cases. Eur Radio!. 2000; 1 0: 1 1 69-1 1 83.
of iatrogenic oesophageal perforations-a viable option. EurJ 97. Spizarny DL, Gross BH, McLoud T. Enlarging pleural effusion
Cardiothorac Surg. 2005;28:7- 1 0 . after liver transplantation. / Thorac Imaging. 1 99 3 ; 8 : 8 5-87.
7 7 . Minnich DJ, Yu P, Bryant A S , e t al. Management o f tho 98. Afessa B, Gay PC, Plevak DJ, et al. Pulmonary complica
racic esophageal perforations. Eur J Cardiothorac Surg. 20 1 1 ; tions of orthotopic liver transplantation. Mayo Clin Proc.
40:93 1 -937. 1 993;68 :427-434.
310 PLE U RAL D I S EASES
99. Bilik R, Yellen M , Superina RA. Surgical complications of pleural effusion after liver cancer surgery. Hepatogastroen
in children after liver rransplanrarion. Pediatr Surg. 1 992; terology. 1 997;44: 1 397- 1 400.
27: 1 3 7 1 - 1 37 5 . 1 03 . Vogl TJ, Straub R, Eichler K, et al. Malignanr liver tumors
1 00. Bozbas SS, Eyuboglu FO, Ozrurk Ergur F, et al. Pulmonary rreated with MR imaging-guided laser-induced thermother
complications and mortality after liver transplant. Exp Clin apy: experience with complications in 899 patients (2,520
Transplant. 2008;6:264-270. lesions) . Radiology. 2002;225:367-377.
1 0 1 . Uetsuji S, Komada Y, Kwon AH, et al. Prevenrion of pleu 1 04. Tajima T, Honda H, Kuroiwa T, et al. Pulmonary compli
ral effusion after hepatectomy using fibrin sealanr. Int Surg. cations after hepatic artery chemoembolization or infusion
1 994;79 : 1 35-137. via the inferior phrenic artery for primary liver cancer. J Vase
1 02. Kise Y, Takayama T, Yamamoto J, et al. Comparison between lnterv Radio/. 2002; 1 3 : 8 93-900.
thoracoabdominal and abdominal approaches in occurrence
P l e u ra l E ff u s i o n S e co n d a ry to D i se a ses
of t h e H e a rt
In this chapter, the pleural effusions that occur after although most effusions are small (5) . In a recent
coronary artery bypass graft (CABG) surgery, those study of 349 patients, the incidence of pleural effu
that occur after cardiac injury (Dressler's syndrome) , sion that occupied more than 25% of the hemitho
those that occur concomitantly with pericardia! dis rax 30 days after CABG was 9.7% (4) . In this study,
ease, and those that occur after heart transplantation the incidence of larger effusions was higher in the
are discussed. patients who received internal mammary grafts
( 1 0.9%) than it was in patients who received only
saphenous vein grafts (4. 5 %) (4) . Hurlbut et al. (6)
PO ST-CO R O N ARY A RT E RY BYPA S S
reported that 4% of 1 00 patients who had under
S U R G E RY
gone CABG developed moderate-to-large effusions.
More than 500,000 CABG procedures are now per If 1 0% of all the patients undergoing CABG develop
formed annually in the United States ( 1 ) . Because a moderate-to-large pleural effusion, then the exuda
pleural effusions complicate many of these proce tive pleural effusion following CABG is one of the
dures, pleural effusions that occur after CABG are most common types of exudative pleural effusion.
one of the most common types of effusion.
occurring early in the postoperative period are fre from Nashville, the prevalence of effusion at 30 days
quently bloody. In one series, the mean pleural fluid postoperatively that occupied more than 2 5 % of the
red blood cell count in 45 patients with large pleural hemithorax was only 3% in the off-pump group ( 1 5)
effusions within the first 30 days of surgery exceeded whereas in a previous study from the same hospital it
2,000,000/mm3 (7) , which is equivalent to a hema had been 1 0% in the on-pump group (4) .
tocrit of 20%.
The etiology of the effusions that occur more than
C l i n ical M a n ifestations
30 days after CABG is not known. The fluid is an exu
date with predominantly lymphocytes (7) . Because Dyspnea is the only symptom that most patients with
the fluid is an exudate, the effusion is probably not pleural effusions experience after CABG (4) . Pleuritic
due to congestive heart failure. The presence of the chest pain, chest wall tenderness, fever, pneumoni
lymphocytes suggests an immunologic basis. Pleural tis, and pericarditis are all unusual. In one series of
biopsies obtained within the first few months of sur 29 patients with large pleural effusions, 7 5 . 9% com
gery demonstrate an intense lymphocytic pleuritis (9) . plained of dyspnea, 1 0 .3% complained of chest pain,
Immunohistochemical staining demonstrates that the and only 1 (3.4%) complained of fever (4) .
lymphocytes in the pleural tissue are both T lympho The pleural effusions that occur after CABG sur
cytes and B lymphocytes with a predominance of gery tend to be unilateral on the left side. In the study
B lymphocytes (9) . The effusions have been attributed using ultrasound in which 42 of 47 patients had
to the post-cardiac injury syndrome (PCIS) ( 1 0) . pleural effusions on the seventh postoperative day,
This explanation, however, i s unsatisfactory because 1 7 (40%) of the effusions were unilateral on the left,
patients with PCIS usually have fever, chest pain, peri 24 (57%) were bilateral, and 1 (2%) was unilateral
carditis, and pneumonitis in addition to the pleural on the right (3) . By the 30th postoperative day, there
effusion. Patients with the late pleural effusions after were 27 patients with effusions, and 1 8 (67%) of
CABG usually do not have fever, chest pain, pericar these were unilateral left sided, 8 (30%) were bilateral,
ditis, or pneumonitis (4) . Possibly, the late pleural and 1 (4%) was unilateral right sided (3) .
effusion after CABG is a variant of or a limited variety In studies of patients who have larger pleural effu
of the PCIS (7) . sions, the effusions are usually left sided, or if they
The administration of topical hypothermia are bilateral, they are larger on the left. In the study
through iced slush during CABG surgery appears to by Sadikot et al. (7) of 7 1 patients with post-CABG
be associated with a higher prevalence of pleural effu pleural effusions who underwent thoracentesis, 42
sion. In one study of patients receiving only saphe of the effusions (59%) were unilateral left sided, 1 8
nous venous grafts, the prevalence of pleural effusion (25%) were bilateral and usually larger on the left,
was 50% in 50 patients receiving topical hypother and 1 1 ( 1 5%) were unilateral on the right.
mia, but only 1 8 % in 50 patients not receiving As mentioned in the preceding text, the larger pleu
topical hypothermia ( 1 1 ) . In a second study of 5 0 5 ral effusions that occur after CABG can be divided into
nonrandomized, consecutive patients undergoing those that occur within the first 30 days of the surgery
CABG surgery, 60% of the 1 9 1 patients who received and those that occur more than 30 days after surgery
topical hypothermia had a pleural effusion, whereas (7, 1 6) . The late effusions do not appear to evolve from
only 25% of the 3 1 4 patients who did not receive the early effusions. The characteristics of the pleural
topical hypothermia had a pleural effusion ( 1 2) . The fluid in the two situations are quite different. The pleu
explanation for the association between iced slush ral fluid with the early effusions is bloody, with a mean
and the presence of pleural effusion is not known, but red blood cell count of approximately 2,000,000/mm3
it has been speculated that cold injury to the phrenic (7, 1 6) . The pleural fluid is frequently eosinophilic,
nerve may cause atelectasis (5) . with a mean eosinophil percentage of greater than
It has been hypothesized that the development 40% (7) . The pleural fluid eosinophilia is probably
of the pleural effusion post-CABG is due, at least due to the blood in the pleural space. Patients with
in part, to the patients being on cardiopulmonary eosinophilic pleural effusions post-CABG also tend
bypass. This is not definitely the case, however, to have peripheral eosinophilia ( 1 7) . There is a signifi
because in two small series ( 1 3 , 1 4) the prevalence of cant correlation between the percentage of eosinophils
pleural effusion was actually higher in patients who in the pleural fluid and the serum, although the per
had off-pump coronary artery bypass surgery than in centage of eosinophils in the serum is lower ( 1 7) . In
those who had on-pump surgery. However, in a series these patients, the eosinophilia is correlated with the
C H APTE R 1 9 I PLE U RAL E F F U S I O N S E C O N DARY TO D I S EASES O F TH E H EART 313
levels of interleukin-5 and eotaxin-3 in the pleural As with the early effusion, the diagnosis of conges
fluid, which are higher than the corresponding levels tive heart failure is eliminated if the patient has an
in the serum ( 1 7) . The mean pleural fluid lactate exudative pleural effusion and the diagnosis of chy
dehydrogenase (LDH) with the bloody effusions is lothorax is excluded if the patient's pleural fluid is
approximately twice the upper limit of normal for clear or has a low triglyceride level. With a lymphocyte
serum (7) . It is likely that much of the pleural fluid predominant pleural effusion, one must exclude tuber
LDH is LDH- 1 , which is the LDH from the red culosis. Because the adenosine deaminase (ADA) level
blood cells. The pleural fluid protein is in the exuda is less than 40 IU/L in patients with pleural effusions
tive range, and the pleural fluid glucose level is not after CABG ( 1 9) and is above this level in patients
reduced (7) . with tuberculous pleuritis, demonstration of an ADA
In contrast to the bloody exudates that were dis below 40 IU/L virtually excludes the diagnosis of
cussed in the preceding text, the pleural fluid that tuberculous pleuritis. Patients with constrictive peri
occurs more than 30 days after CABG is a clear yellow carditis will usually have other signs and symptoms
lymphocyte-predominant exudate. The mean lym such as bilateral pedal edema and ascites.
phocyte percentage for 26 lace effusions in one series
was 61 %, whereas the mean eosinophil percentage
Treatm e n t
was only 2% (7) . The pleural fluid LDH tends to
be lower with the late effusions than with the early Most patients with smaller pleural effusions post
effusions and averages about the upper limit of nor CABG require no treatment as the effusion gradually
mal for serum (7) . As with the early effusions, the disappears (5). When a patient is identified with a large
pleural fluid protein is in the exudative range and pleural effusion after CABG (occupying more than
the pleural fluid glucose level is not reduced (7) . 25% of the hemithorax) , a thoracentesis should be
A small percent of patients will develop dyspnea from performed to exclude the other diagnoses in the differ
a pleural effusion more than 90 days post-CABG ential outlined earlier. Because most of these patients
surgery ( 1 8) . Most effusions occurring this long after are dyspneic (4) and the therapy of choice for these
surgery are transudates due to heart failure ( 1 8) . effusions is a therapeutic thoracentesis, it is recom
mended that the initial thoracentesis be a therapeutic
thoracentesis (5) .
D i a g nosis
If the other diagnostic possibilities are excluded
The diagnosis of pleural effusion secondary to CABG and the fluid recurs, a second and then a third thera
is one of exclusion. In the days immediately after peutic thoracentesis are indicated. Many patients are
CABG, the main diagnoses to exclude are congestive also given nonsteroidal anti-inflammatory agents
heart failure, pulmonary embolus, parapneumonic (NSAIDs) or oral prednisone, but there are no
effusion, and chylothorax. Congestive heart failure controlled studies documenting the efficacy of this
is excluded if the patient has an exudative pleural approach. There is one study that evaluated the
effusion. Chylothorax is excluded if the fluid is clear effectiveness of diclofenac, an NSAID, in prevent
yellow or if the triglyceride levels are low. Pulmonary ing pleural effusion in the immediate postoperative
embolus is more difficult to exclude, and a computed period (20) . Niva et al. (20) randomized patients to
tomography (CT) angiography is necessary in some receive 50 mg diclofenac (22 patients) or placebo
cases (see Chapter 1 7) . However, the pleural effu ( 1 9 patients) orally every 8 hours in the postoperative
sion with pulmonary embolus usually occupies less period. They reported chat the control group had a
than 25% of the hemithorax and disappears spon higher incidence of pleural effusion (42. 1 % ) at dis
taneously within a couple of weeks. Patients with charge than did the diclofenac-treated group (22.7%)
parapneumonic effusions are usually febrile, and the (20) . In a second study, lmazio et al. (20a) studied
pleural fluid differential white blood cell (WBC) whether colchicine would reduce the incidence of
count reveals predominantly neutrophils and a very effusions post cardiac surgery in a double-blind ran
low percentage of eosinophils. domized study of 360 patients. They reported that
The differential diagnosis is somewhat different the incidence of pleural effusion was significantly less
for the late pleural effusion occurring after CABG, ( 1 2.2%) in the group that received colchicine starting
and the main diagnoses to consider are congestive on the third postoperative day and continuing for a
heart failure, chylothorax, tuberculosis, malignancy, month than it was in the group that received placebo
constrictive pericarditis, and pulmonary embolus. (22. 8%) (2 1 ) .
314 PLE U RAL D I S EASES
Some patients have been managed successfully the 22 patients ( 1 8%) had effusions before surgery.
with chemical pleurodesis. Before one becomes too Nine patients required thoracentesis to relieve dys
aggressive in managing this condition, it is important pnea and the median time from placement to thora
to realize that most patients will do well with no more centesis was 23 days. All the pleural fluids examined
than a couple of thoracenteses. In our prospective were blood-tinged exudates (22) .
study, we followed 30 patients with pleural effusions
occupying more than 2 5 % of the hemithorax for
PO ST-CA R D IAC I N J U RY { D R E S S L E R'S}
12 months. During this period, 8 (27%) received no
SYN D RO M E {PCI S}
invasive treatment for the pleural effusion, 16 (53%)
received a single thoracentesis, 2 (7%) received two PCIS is characterized by the onset of fever, chest pain,
thoracenteses, and 4 ( 1 3%) received three or more pleuropericarditis, and parenchymal infiltrates in the
thoracenteses. Only one patient was still receiving weeks following injury to the pericardium or myocar
periodic thoracenteses 1 2 months after CABG (4) . dium (23,24) . There is no universal agreement on the
Twenty-two of the 25 patients who underwent tho definition of the PCIS (22) . Mott et al. (25) defined
racentesis reported that their dyspnea was alleviated noncomplicated PCIS as the presence of a tempera
with the thoracentesis (4) . ture greater than 1 00 . 5°F, patient irritability, pericar
On occasion, the effusion persists despite several dia! friction rub, and a small pericardia! effusion with
therapeutic thoracenteses. We have subjected eight or without pleural effusion following cardiac trauma
such patients to thoracoscopy in recent years. At tho (25 ) . They defined complicated PCIS as a noncom
racoscopy, several patients had thin sheets of fibrous plicated PCIS plus the need for hospital readmission
tissue that coated the lung and prevented it from with or without the need for pericardiocentesis or
expanding (9) . It is likely that this sheet of fibrous thoracentesis (2 5 ) . I prefer the proposed definition of
tissue "trapped" the lung and prevented it from reex Imazio et al. (26) who define the PCIS as the presence
panding. After the fibrous tissue coating the visceral of at least two of the following: fever without alter
pleura was removed, the lung expanded and the effu native explanation, pleuritic chest pain, pericardia!
sion did not recur (9) . However, because most had friction rub, new or worsening pleural effusion, and
a mechanical or a chemical pleurodesis at the same new or worsening pericardia! effusion. This syndrome
time, one cannot be certain that the decortication was has been described following myocardial infarction,
responsible for the effusion not recurring. Recurrent cardiac surgery, blunt chest trauma, percutaneous
pleural effusions post-CABG surgery have also been left ventricular puncture, pacemaker implantation,
managed successfully with video-assisted thoracic angioplasty, and repair of pectus excavatum (27) .
surgery with talc insuffiation (2 1 ) . No mention was
made in this latter paper about membranes encasing I n c i d e nce
the visceral pleural (2 1 ) .
The incidence of the PCIS was thought by Dressler
I n view o f the series mentioned i n the preced
to be 3% to 4% after an acute myocardial infarction
ing text, thoracoscopy is recommended for an effu
(23 ) . Subsequent studies have demonstrated that the
sion after CABG that continues to recur for several
incidence is probably less than 1 % (28), but the inci
months despite several therapeutic thoracenteses. At
dence is much higher in patients with large transmural
thoracoscopy, any fibrous tissue coating the visceral
infarctions in which the pericardium is involved (29) .
pleura should be removed and the parietal pleura
In one series, 1 5% of patients with an acute myocar
should be abraded to create a pleurodesis.
dial infarction and pericarditis developed the post
myocardial infarction syndrome during the follow-up
P l e u ra l Effu s i o n After Ve ntricu l a r Assi st
period (29) . The incidence of the syndrome is much
Device Placement
higher following surgical procedures involving the
Patients who receive ventricular assist devices are pericardium than after an acute myocardial infarction
being used more frequently as bridges to cardiac (24,30) . Engle et al. (24) reported that 30% of 257
transplantation or to facilitate patient transfers. It children undergoing cardiac operations developed
appears that most placements of this device are asso the syndrome. Miller et al. (3 1 ) reported an incidence
ciated with the development of a pleural effusion of 1 7. 8 % in 944 patients undergoing cardiac surgery
(22) . Guha et al. (22) reviewed the placement of at Johns Hopkins Hospital during a 1 -year period. In
22 of these devices and reported that every patient a recent study (32), the incidence of PCIS was 1 5 %
had a pleural effusion after placement (22) . Six of in 360 patients undergoing cardiac surgery.
C H APTE R 1 9 I PLE U RAL E F F U S I O N S E C O N DARY TO D I S EASES O F TH E H EART 315
congestive heart failure, pulmonary embolism, and 1 33 consecutively discharged patients with pericar
pneumonia. Congestive heart failure as a cause of the dia! effusion. Thirty-five of the patients (26%) had
pleural effusion is excluded by the demonstration of a roentgenographically demonstrable pleural effusion
an exudative pleural fluid. A pulmonary CT angio and no other lung disease. Twenty-one of the patients
gram or a perfusion lung scan should be obtained to had inflammatory pericardia! disease without conges
exclude the diagnosis of pulmonary embolization. tive heart failure, and 1 5 of them had only a left-sided
It is important not to mistakenly diagnose pulmonary pleural effusion, 3 had more fluid on the left than on
embolism rather than the PCIS because anticoagu the right, and in 3 , the effusions were of the same size
lation is contraindicated in the PCIS (23) . Patients on both sides. Of the five patients with inflammatory
with the syndrome are at risk for developing hemo pericarditis and congestive heart failure, the effusions
pericardium. One report suggested that the diagnosis were equal bilaterally in two, greater on the right side
of the syndrome could be established by demonstrat in two, and left sided in one. Two of the three patients
ing a high titer of antimyocardial antibodies and a low with constrictive pericarditis had a unilateral left
complement level in the pleural fluid (38) . However, sided effusion. Sun et al. (42) performed CT scans
a subsequent publication (39) and our own unpub on 74 patients with pericardia! effusions and reported
lished observations have failed to confirm that levels that 52 (70%) had a pleural effusion. The incidence
of antimyocardial antibodies in the pleural fluid are of effusion was comparable whether the patient
elevated in patients with the PCIS . had benign or malignant pericardia! disease (42) .
Tomaselli et al. (43) reviewed 30 cases of constrictive
pericarditis and found that a pleural effusion was pres
Treatment
ent in 1 8 (60%) . In 12 of the 1 8 patients, the effusion
This syndrome usually responds to treatment with was bilateral and approximately symmetric. Three
antiinflammatory agents such as aspirin or indometh effusions were left sided, and three were right sided.
acin (33) . In the more severe forms of the syndrome, We described one patient with constrictive pericardi
corticosteroids may be necessary (33 ) . However, the tis who had a large unilateral right-sided pleural effu
prophylactic administration of intravenous methyl sion, which we attributed to the transdiaphragmatic
prednisolone at a standard anti-inflammatory dose transfer of his ascitic fluid (44) .
in children after cardiac surgery with cardiopulmo The mechanism responsible for the pleural effu
nary bypass neither prevents nor attenuates the PCIS sion associated with pericardia! disease is not clear.
(25 ) . A recent placebo-controlled double-blind ran The obvious explanation is that the pulmonary and
domized study (40) of 360 patients reported that the systemic capillary pressures are elevated secondary
administration of colchicine on the third postopera to the pericardia! disease, resulting in a transudative
tive day and continuing for 30 days was associated pleural effusion. However, if this were the sole expla
with a reduction in the incidence of pericardia! effu nation, one would not expect most of the effusions to
sion from 22.8% to 1 2.8 % and a reduction in the be left sided in patients with inflammatory pericardia!
incidence of pleural effusion from 26. 5 % to 1 2 .2% disease and also that some patients with constrictive
It is important to establish the diagnosis of the pericarditis would have exudative pleural effusions
PCIS in patients who have undergone CAB Gs because (43 ) . It is probably that in patients with pericardia!
the pericarditis may cause graft occlusion. Urschel effusions, some of the fluid passes directly from the
et al. (36) reported that graft occlusion occurred in pericardia! space into the pleural space. Gibson and
1 2 of 1 4 patients (86%) who developed the syndrome Segal (45) showed that 20% of protein injected into
after CABGs and who were treated symptomatically. the pericardium entered the pleural space within
When 3 1 subsequent patients were treated with pred an hour of the injection. The pericardia! inflammation
nisone, 30 mg/day for a week and tapering doses for itself may increase the ease by which fluid passes from
5 weeks thereafter, in addition to aspirin 600 mg q.i.d., the pericardia! to the pleural space (4 1 ) . In addition
only 5 ( 1 6%) of the grafts became occluded (36) . the pericardia! inflammation may also cause inflam
mation of the pleura covering the pericardium which
could lead to a pleural effusion.
P E R I CA R D I A L D I S EA S E
The characteristics of the pleural fluid seen in
A substantial percentage o f patients with pericardia! conjunction with pericardia! disease are not well
disease develop a pleural effusion, which is usually left described. Tomaselli et al. (43) reported that the fluid
sided. Weiss and Spodick (4 1 ) reviewed the charts of was exudative in three patients and transudative in
C H APTE R 1 9 I PLE U RAL E F F U S I O N S E C O N DARY TO D I S EASES O F TH E H EART 317
one patient with constrictive pericarditis. We described heart failure, and a pleural effusion post-heart trans
one patient with a large right-sided pleural effusion plant similar to the effusions post-CABG surgery. If
with constrictive pericarditis who had a pleural Buid the effusion is very small, it can probably be ignored
protein of 4 g/dL (44) . In another recent report of because these effusions are very common (47) . If the
two patients with constrictive pericarditis secondary effusion occupies less than 2 5 % of the hemithorax
to bromocriptine therapy, the pleural Buid in one had but represents more than just blunting of the cos
a protein level of 4 g/dL (simultaneous serum level of tophrenic angle, a diagnostic thoracentesis should
6.4 g/dL) (46) . I would suspect that the pleural Buid be performed if the patient is complaining of short
with inflammatory pericardia! disease is also exudative. ness of breath or if the patient is not feeling up to
Obviously, the treatment of choice for the pleural par. It is important to remember that these patients
effusion secondary to pericardia! disease is to treat the are immunosuppressed and accordingly are more
pericardia! disease. likely to have infections. Moreover, patients with
pleural infections may not be febrile because of the
immunosuppression.
PO ST-H EART TRA N S P LA N TAT I O N
Pleural effusions are common after heart transplan
P U L M O NARY V E I N STE N O S I S A FTE R
tation. Misra et al. (47) retrospectively reviewed the
CATH ETE R A B LATI O N O F ATR I A L
charts and imaging studies of 72 patients who had
F I B R I L LATI O N
undergone orthotropic heart transplants over a 6-year
period and reported that 6 1 patients (85%) had a A procedure that is being used more and more
pleural effusion postoperatively while only 1 9 (26%) commonly for the treatment of chronic atrial fibrilla
had effusions preoperatively. Most of the effusions tion is circumferential pulmonary-vein ablation (50) .
were small and bilateral, but 1 6% of the patients A known complication of this procedure is pulmonary
had effusions that occupied more than 2 5 % of the vein stenosis (5 1 ) , which occurs in approximately 3%
hemithorax (47) . All of these effusions were attrib of patients who undergo this procedure (52) . Present
uted to the transplantation procedure. The effusions ing symptoms with pulmonary vein stenosis include
were largest at median hospital day 6 . 5 . Most of shortness of breath, cough, and hemoptysis (53). In
them resolved within the first year after transplanta one series, 5 of 1 8 patients (28%) had a pleural effu
tion (47) . Pleural Buid was examined in four of the sion whereas 7 (39%) had a parenchymal infiltrate
patients and was exudative in two (47) . (53) . The pleural Buid was described in one case, and it
A second study (48) reviewed all the pulmonary was exudative with a pleural Buid protein of 4.7 g/dL
complications in 1 57 patients that received 1 5 9 (serum 7. 1 g/dL) and a pleural Buid LDH level of
transplants. Ten of the patients (6.3%) developed 1 ,308 U/L (serum 302 U/L) (54) . Pathologic exami
pleural effusions that warranted a diagnostic tho nation of the areas of infiltrate reveal hemorrhagic
racentesis and/or specific therapeutic intervention infarctions. The pathogenesis of the infiltrates and the
with diuretics, antibiotics, or chest tubes (48) . The effusion is probably ischemia from the lack of perfu
etiology of the effusion was found to be parapneu sion of the lung drained by the occluded vein. The
monic in four cases, chylothorax in one, hemothorax perfusion lung scan shows decreased perfusion to the
secondary to Aspergillus pneumonia in one, sepsis areas drained by the stenotic vein. Treatment is stent
in one, and probable transudates in three cases (48) . ing or balloon angioplasty of the stenotic vein (5 5 ) .
There was one case report of a pleural effusion due
to post-transplant lymphoproliferative disorder (49) .
F O N TA N P R O C E D U R E
At autopsy, the patient had pleural-based tumor
nodules compatible with post-transplant lympho With the Fontan procedure, the right ventricle is
proliferative disorder (49) . bypassed by an anastomosis between the superior vena
The management of patients with pleural effusions cava and the inferior vena cava and, the right atrium,
depends on the size of the effusion and the patient's or the pulmonary artery (56) . The procedure is typi
clinical picture. If the effusion occupies more than cally performed for tricuspid atresia or univentricular
2 5 % of the hemithorax, a diagnostic thoracentesis heart. Pleural effusion is a significant problem after
should be performed immediately in an attempt to the Fontan procedure. Persistent pleural drainage is
ascertain the etiology of the effusion. The main consid the primary cause of prolonged postoperative hospital
erations are pleural infection, chylothorax, congestive stay in patients who have had a Fontan procedure (57) .
318 PLE U RAL D I S EASES
Zellers et al. (58) analyzed pleural fluid formation It appears that creation of a fenestration at the
after this procedure on 46 patients. They reported time of surgery will reduce the amount of pleural
that the median amount of pleural drainage was fluid drainage. Lemler et al. (63) randomly assigned
3 ,220 mL, with a range of 1 5 5 to 3 1 ,000 mL. Most 49 patients to receive or not receive a fenestration
patients had pleural drainage from both sides. Gupta at the time of the Fontan procedure. The fenestra
et al. (59) reviewed the pleural fluid drainage in 1 00 tion consisted of a single 3- to 6-mm communica
consecutive patients who underwent the Fontan oper tion between the Fontan channel and the pulmonary
ation and reported that the median duration of chest venous atrium (63) . Performance of the fenestra
tube drainage was 1 0 days and the median volume tion was associated with a reduction in the median
of drainage was 1 4 . 7 ml/kg/day. chest tube drainage from 4 , 5 8 8 to 1 ,734 mL and
The pathogenesis of the formation of the large a decreased hospital stay from 23 to 1 2 days (63) .
amounts of pleural fluid postoperatively in these The disadvantage to the creation of a fenestration
patients is not definitely known. It is probably related is that it creates a right-to-left shunt which can lead
to the increased systemic venous pressure. It is unclear, to increased hypoxemia (63) . Not all studies have
however, whether increased pleural fluid transudation found that fenestration is associated with decreased
from the parietal pleura, lymphatic leakage into the amounts of pleural fluid drainage. In a study in which
pleural space (60) , or hormonal changes are responsi the patients were not randomized, Atik et al. reported
ble for the large accumulations of pleural fluid. Spicer that pleural drainage was less in patients who did not
et al. (6 1 ) analyzed the factors that were related to receive fenestration (64) .
the development of pleural effusions after the Fontan It has been hypothesized that alterations in the
procedure in 71 children. They found that patients hormones that regulate fluid and electrolyte homeo
with significant aortopulmonary collateral vessels stasis may also play a role (65) . Indeed in one study,
evidenced by angiographic opacification of the pul patients who developed effusions following surgery
monary arteries or veins had more prolonged pleural had an elevated serum renin and angiotensin com
drainage. They believed that the aortopulmonary col pared with those who did not (65) . In one study,
lateral vessels contributed to volume loading of the patients who were given enalapril 5 ,ug/kg intrave
systemic ventricle and to elevation of the pulmonary nously within 1 hour of surgery and every 1 2 hours
artery, and right atrial and caval pressures, all of which thereafter until they were able to tolerate enteral feed
increase the rate of formation of the pleural fluid. ing had significantly less total pleural fluid drainage
When 1 3 patients were subjected to preoperative ( 1 0.6 mL/kg) than did patients who did not receive
embolization of these vessels, the median duration of enalapril ( 1 9.6 mL/kg) (66) . However, the admin
the effusion postoperatively was only 6 . 5 days (6 1 ) . istration of captopril had no effect on pleural fluid
Gupta e t al. (59) performed a multivariate analysis formation in a subsequent study (57) .
for significant risk factors for pleural effusions last The pleural fluid with the Fontan procedure is an
ing more than 2 weeks or more than 20 mL/kg/ day exudate in almost every case (67) . In one series of 1 5
in 1 00 children undergoing the Fontan procedure. patients, the mean pleural fluid LDH was 1 , 575 IU/L
They found that a lower preoperative oxygen satura suggesting a very inflammatory state (67) . The pleural
tion, a smaller conduit size, and a longer duration of fluid triglycerides were elevated in 4 of the 1 5 patients
cardiopulmonary bypass were associated with the lon indicating that these patients had chylothoraces (67) .
ger and larger amounts of fluid drainage. The authors The optimal treatment for the patient who has
were unable to demonstrate a relationship between prolonged pleural drainage after a Fontan procedure
the presence of aortopulmonary collateral vessels and remains to be determined. The intrapleural adminis
the amount of pleural fluid drainage (59) . Yun et tration of tetracycline at the end of the surgical pro
al. (62) studied 85 patients undergoing the Fontan cedure had no effect on the amount or the duration
procedure and reported that prolonged pleural drain of the fluid drainage (58). One study by Cava et al.
age was associated with a low pulmonary vascular demonstrated that the duration of pleural drainage
compliance and cardiopulmonary bypass time. The was shorter and the total volume was less when the
Fontan procedure is also performed in adults and per patients postoperatively were placed on a regimen
sistent pleural effusion is a problem in adults as it is consisting of diuretics (hydrochlorothiazide and spi
in children. Burkhart et al. (62a) reported that 36 of ronolactone) , fluid restriction, captopril, a low-fat
1 2 1 adults who underwent the Fontan procedure had diet, and a minimum of 0.5 L of supplemental oxygen
pleural effusions that persisted more than 2 weeks. by nasal cannula than when they were not placed on
C H APTE R 1 9 I PLE U RAL E F F U S I O N S E C O N DARY TO D I S EASES O F TH E H EART 319
such a protocol (68 ). In patients who have mark 1 4 . Montes FR, Maldonado JD, Paez S , et al. Off-pump versus
on-pump coronary artery bypass surgery and postoperative
edly prolonged pleural fluid drainage, consideration
pulmonary dysfunction. J Cardiothorac Vase Anesth. 2004;
should be given to the implantation of a pleuroperi 1 8 :69 8-703.
toneal shunt (69), the insertion of an indwelling cath 15. Mohamed KH, Johnson T, Rodriguez RM , et al. Pleural
eter such as the Pleurx Pleural Catheter (Care Fusion, effusions post coronary artery bypass grafting performed with
San Diego, California) or an attempt at pleurodesis or without a bypass pump. Am J Respir Crit Care Med. 200 1 ;
1 63:A90 1 .
with a sclerosing agent such as doxycycline (70) .
1 6 . Light RW, Rogers JT, Cheng D-S, et al. Large pleural effu
Some patients have been managed successfully with sions occurring after cotonary artery bypass grafting. Ann
medium-chain triglycerides or with pleurectomy Intern Med. 1 999; 1 30:89 1-896.
and ligation of the thoracic duct (60) . Caverly and 1 7. Kalomenidis I, Stathopoulos GT, Barnette R, et al. Eotaxin-3
coworkers administered octreotide to 1 0 patients and interleukin-5 pleural fluid levels are associated with pleural
fluid eosinophilia in pose-coronary artery bypass grafting
with persistent effusions after the Fontan procedure
pleural effusions. Chest. 2005 ; 1 27:2094-2 1 00 .
and reported resolution of the effusion with octreo 1 8 . Peng M C , H o u CJ, Li JY, e t al. Prevalence o fsymptomatic large
tide in 6 of the patients (7 1 ) . pleural effusions first diagnosed more than 30 days after coro
nary artery bypass graft surgery. Respirology. 2007; 1 2 : 1 22-1 26.
19. Lee YCG, Rogers JT, Rodriguez RM , et al. Adenosine deami
REFERENCES nase levels in nontuberculous lymphocytic pleural effusions.
Chest. 200 1 ; 1 20:356-36 1 .
I. American Heart Association. Heart Disease and Stroke 20. Niva M , Biancari F, Valkama J , et al. Effects of diclofenac in
Statistics-1999 Update. Dallas, TX: American Heart the prevention of pericardial effusion after coronary artery
Association; 1 99 8 . bypass surgery. A prospective, randomized study. J Cardiovasc
2 . Peng M-J, Vargas FS, Cukier A , e t al. Postoperative Surg (Torino). 2002;43 :449-453.
pleural changes after cotonary revascularization. Chest. 20a. Imazio M, Brucato A, Rovere ME, et al. Colchicine prevents
1 992; 1 0 1 :327-330. early postoperative pericardia! and pleural effusions. Am
3 . Vargas FS, Cukier A, Hueb W, et al. Relationship between Heartf. 2 0 1 1 ; 1 62 : 527-532.
pleural effusion and pericardia! involvement after myocardial 2 1 . Paull DE, Delahanty TJ, Weber FJ, et al. Thoracoscopic
revascularizacion. Chest. 1 994; 1 0 5 : 1 748- 1 752. talc pleurodesis for recurrent, symptomatic pleural effusion
4. Light RW, Rogers JT, Moyers JP, et al. Prevalence and clinical following cardiac operations. Surg Laparosc Endosc Percutan
course of pleural effusions after coronary artery and cardiac Tech. 2003; 1 3 :339-344.
surgery. Am ] Respir Crit Care Med. 2002; 1 66: 1 563- 1 566. 22. Guha A, Munjampalli S, Bandi V, et al. Pleural effusion
5 . Light RW. Pleural effusions after coronary artery bypass graft after ventricular assist device placement: prevalence and
surgery. Curr Opin Pulm Med. 2002;8:308-3 1 1 . pleural fluid characteristics. Chest. 2008; 1 34:3 82-386.
6 . Hurlbut D , Myers ML, Lefcoe M , et al. Pleuroplumonary 23. Dressler W The post-myocardial infarction syndrome. Arch
morbidity: internal thoracic artery versus saphenous vein Intern Med. 1 9 59; I 03:28-42.
graft. Ann Thorac Surg. I 990;50:959-964. 24. Engle MA, Zabriskie JB, Senterfic LB, et al. Post-pericardiot
7. Sadikoc RT, Rogers JT, Cheng D-S, et al. Pleural fluid omy syndrome: a new look at an old conclition. Mod Concepts
characteristics of patients with symptomatic pleural effu Cardiovasc Dis. 1 975;44:59-64.
sion post coronary artery bypass surgery. Arch Intern Med. 2 5 . Mott AR, Fraser CD Jr, Kusnoor AV, et al. The effect of short
2000; 1 60:2665-2668 . term prophylactic methylprednisolone on the incidence and
8 . Daganou M, Dimopoulou I, Michalopoulos N, e t al. Respira severity of postpericardiotomy syndrome in children under
tory complications after coronary artery bypass surgery with going cardiac surgery with cardiopulmonary bypass. J Am
unilateral or bilateral internal mammary artery grafting. Chest. Coll Cardiol. 200 1 ;37: 1 700- 1 706.
1 998; 1 1 3 : 1 285-1289. 26. Imazio M, Brucato A, Ferrazzi P, et al. Postpericardiotomy
9 . Lee YC, Vaz MA, Ely KA , et al. Symptomatic persistent syndrome: a proposal for diagnostic criteria. J Cardiovasc Med
pose-coronary artery bypass graft pleural effusions requiring (Hagerstown). In press.
operative treatment: clinical and histologic features. Chest. 27. Berberich T, Haecker FM, Kehrer B, et al. Postpericardi
200 1 ; 1 1 9:795-800. otomy syndrome after minimally invasive repair of pectus
10. Kim YK, Mohsenifar Z, Koerner SK. Lymphocytic pleu excavatum. J Pediatr Surg. 2004;39:el-e3 .
ral effusion in postpericardiotomy syndrome. Am Heart j. 28. Liem KL, ten Veen JH, Lie KI, et al. Incidence and sig
1 9 8 8; 1 1 5 : 1 077- 1 079. nificance of heart-muscle antibodies in patients with acute
1 1 . Allen BS, Buckberg GD, Rosenkranz ER, et al. Topical cardiac myocardial infarction and unstable angina. Acta Med Scand.
hypothermia in patients with coronary disease. An unneces 1 979;206:473-475 .
sary adjunct to cardioplegic protection and cause of pulmo 2 9 . Toole J C , Silverman M E . Pericarditis o f acute myocardial
nary morbidity. J Thorac Cardiovasc Surg. 1 992; I 04:626-63 1 . infarction. Chest. 1 975 ;67:647-653.
1 2 . Nikas DJ, Ramadan FM, Elefteriades JA. Topical hypother 30. McCabe JC, Ebert PA, Engle MA, et al. Circulating heart
mia: ineffective and deleterious as adjunct to cardioplegia for reactive antibodies in the post-pericardiotomy syndrome.
myocardial protection. Ann Thorac Surg. I 998;65:28-3 1 . J Surg Res. 1 973; 1 4 : 1 5 8- 1 64.
1 3 . Staton GW, Williams WH , Mahoney EM, et al. Pulmonary 3 1 . Miller RH, Horneffer PJ, Gardner TJ, et al. The epidemiol
outcomes of off-pump vs on-pump coronary artery bypass ogy of the postpericardiotomy syndrome: a common compli
surgery in a randomized trial. Chest. 2005 ; 1 27:892-90 1 . cation of cardiac surgery. Am Heartj. 1 98 8; 1 1 6 : 1 323- 1 329.
320 PLE U RAL D I S EASES
32. Imazio M , Brucato A, Rovere ME, et al. Contemporary fea 5 4 . Salamon F, Hirsch R , Tur-Kaspa R , e t al. Search fo r the com
tures, risk factors, and prognosis of the post-pericardioromy plication. N Engl] Med. 2006;354:9 57-963.
syndrome. Am ] Cardiol. 20 1 1 ; 1 08 : 1 1 83-1 1 87. 55. Packer DL, Keelan P, Munger TM, et al. Clinical presentation,
33. Imazio M . The post-pericardiotomy syndrome. Curr Opin investigation, and management of pulmonary vein stenosis
Pulm Med. 20 1 2; 1 8 :366-374. complicating ablation for atrial fibrillation. Circulation. 2005;
34. De Scheerder I, De Buyzere M, Robbrecht J, et al. Postopera 1 1 1 : 5 46-5 54.
tive immunologic response against contractile proteins after 56. Laks H , Milliken JC, Perloff JK, et al. Experience with the
coronary bypass surgery. Br Heart]. l 986;56:440-444. Fontan procedure. J 7horac Cardiovasc Surg. 1 984;88:
3 5 . Kossowsky WA, Epstein PJ, Levine RS. Post-myocardial 939-95 1 .
infarction syndrome: an early complication of acute myocar 57. Heragu N , Mahony L. Is captopril useful in decreasing pleu
dial infarction. Chest. 1 973;63 :3 5-40. ral drainage in children after modified Fontan operation? Am
36. Urschel HC Jr, Razzuk MA, Gardner M . Coronary artery ] Cardiol. 1 999;84: 1 1 09-1 1 1 2.
bypass occlusion secondary co post-cardiocomy syndrome. 58. Zellers TM, Driscoll DJ, Humes RA, et al. Glenn shunt: ef
Ann 7horac Surg. 1 976;22: 528-53 1 . fect on pleural drainage after modified Fontan operation. J
37. Stelzner TJ, King TE Jr, Antony VB, et al. The pleuropul 7horac Cardiovasc Surg. 1 989;98 :725-729.
monary manifestations of the postcardiac injury syndrome. 5 9 . Gupta A, Daggett C, Behera S, et al. Risk factors for persis
Chest. 1 983;84:383-387. tent pleural effusions after the extracardiac Fontan procedure.
3 8 . Kim S, Sahn SA. Postcardiac injury syndrome. An immuno J 7horac Cardiovasc Surg. 2004; 1 27: 1 664- 1 669.
logic pleural Buid analysis. Chest. 1 996; 1 09 : 5 70-572. 60. van de Wal HJ, Tanke RF, Roef MJ. The modified Senning
39. Shrivastava R, Venkatesh S, Pavlovich BB, et al. Immunologi operation for cavopulmonary connection with autologous
cal analysis of pleural Buid in post-cardiac injury syndrome. tissue. ] 7horac Cardiovasc Surg. 1 994; 1 0 8 :377-3 80.
Postgrad Med]. 2002;78 :362-363. 6 1 . Spicer RL, Uzark KC, Moore JW, et al. Aortopulmonary col
40. Imazio M , Brucato A, Rovere ME, et al. Colchicine prevents lateral vessels and prolonged pleural effusions after modified
early postoperative pericardia! and pleural effusions. Am Fontan procedures. Am Heart]. 1 996; 1 3 1 : 1 1 64-1 1 6 8 .
Heart]. 2 0 1 1 ; 1 62 : 527-5 32. 6 2 . Yun TJ, I m YM, Jung S H , et al. Pulmonary vascular compli
4 1 . Weiss JM, Spodick DH. Association of left pleural effusion ance and pleural effusion duration after the Fontan proce
with pericardia! disease. N Engl] Med. 1 9 83;308:696-697. dure. Int] Cardiol. 2009 ; 1 3 3 : 5 5-6 1 .
42. Sun JS, Park KJ, Kang DK. CT findings in patients with 62a. Burkhart HM, Dearani JA, Mair D D e t al. Th e modified
pericardia! effusion: differentiation of malignant and benign Fontan procedure: early and late results in 1 32 adult pa
disease. AJR A m j Roentgenol. 20 1 0; 1 94:W489-W494. tients. J 1hor Cardiovasc Surg. 2003; 1 2 5 : 1 252- 1 25 9 .
43. Tomaselli G, Gamsu G, Stulbarg MS. Constrictive pericar 63. Lemler MS, Scott WA, Leonard SR, e t a l . Fenestration
ditis presenting as pleural effusion of unknown origin. Arch improves clinical outcome of the Fontan procedure: a
Intern Med. 1 9 8 9 ; 1 49:20 1 -203. prospective, randomized study. Circulation. 2002; 1 0 5 :
44. Sadikot RT, Fredi JL, Light RW. A 43-year-old man with a 207-2 1 2 .
large recurrent right-sided pleural effusion. Chest. 2000; 1 1 7: 6 4 . Atik E, Ikari N M , Martins TC, e t al. Fontan operation and
1 1 9 1 - 1 1 94. the cavopulmonary technique: immediate and late results
4 5 . Gibson AT, Segal MB. A study of the routes by which pro according co the presence of atrial fenestration. Arq Bras
tein passes from the pericardia! cavity co the blood in rabbits. Cardiol. 2002;78 : 1 62- 1 66.
J Physiol. l 978;280:423-433 . 6 5 . Mainwaring RD, Lamberti JJ, Carter TL Jr, et al. Renin,
46. Champagne S, Coste E, Peyriere H, e t a l . Chronic constric angiotensin II, and the development of effusions following
tive pericarditis induced by long-term bromocriptine therapy: bidirectional Glenn and Fontan procedures. J Card Surg.
report of two cases. Ann Pharmacother. 1 999;33: 1 05 0- 1 054. 1 99 5 ; 1 0 : 1 1 1 - 1 1 8 .
47. Misra H, Dikensoy 0, Rodriguez RM , et al. Prevalence of 66. Thompson LD, McElhinney D B , Culbertson CB, e t al. Peri
pleural effusions post orthocopic heart transplantation. Respi operative administration of angiotensin-converting enzyme
rology. 2007; 1 2 : 8 87-890. inhibitors decreases the severity and duration of pleural
48. Lenner R, Padilla ML, Teirstein AS, et al. Pulmonary effusions following bidirectional cavopulmonary anastomo
complications in cardiac transplant recipients. Chest. sis. Cardiol Young. 200 1 ; 1 1 : 1 9 5-200.
200 1 ; 1 20:5 08-5 1 3 . 67. Brixey AG, Luo Y, Bernard Y, et al. Characteristics of pleural
4 9 . Lamba M , Jabi M , Padmore R, e t al. Isolated pleural PTLD after effusions in patients undergoing the Fontan or Glenn pro
cardiac transplantation. Cardiovasc Pathol. 2002; 1 1 :346-350. cedures for congenital cardiac defects. Am J Respir Crit Care
5 0 . Oral H, Pappone C, ChughA, et a!. Circumferential pulmonary Med. 2 0 1 1 ; 1 83:A2962.
vein ablation for chronic atrial fibrillation. N EnglJ Med. 2006; 68. Cava JR, Bevandic SM, Steltzer MM, et al. Strategy co reduce
354:934-94 1 . persistent chest tube drainage after the Fontan operation. Am
5 1 . Robbins IM, Colvin EV, Doyle TP, et al. Pulmonary vein ste j Cardiol. 2005;96 : 1 30- 1 33 .
nosis after catheter ablation of atrial fibrillation. Circulation. 69. Sade RM , Wiles H B . Pleuropericoneal shunt for persistent
1 998;98: 1 769- 1 77 5 . pleural drainage after Fontan procedure. J 7horac Cardiovasc
52. Saad E B , Rossillo A , Saad CP, et a l . Pulmonary vein stenosis Surg. 1 990; 1 00:62 1 -623.
after radiofrequency ablation of atrial fibrillation: functional 70. Hoff DS, Gremmels DB, Hall KM, et al. Dosage and ef
characterization, evolution, and influence of the ablation fectiveness of intrapleural doxycycline for pediatric post
strategy. Circulation. 2003; 1 08 : 3 1 02-3 1 07. cardiocomy pleural effusions. Pharmacotherapy. 2007;27:
5 3 . Saad EB, Marrouche NF, Saad CP, et al. Pulmonary 995-1 000.
vein stenosis after catheter ablation of atrial fibrillation: 7 1 . Caverly L, Rausch CM, Da Cruz E, et al. Octreotide treat
emergence of a new clinical syndrome. Ann Intern Med. ment of chylothorax in pediatric patients following cardio
2003; 1 3 8 :634-63 8 . thoracic surgery. Congenit Heart Dis. 2 0 1 0 ; 5 : 573-578.
P l e u ra l D i se a s e i n O b stet r i cs
a n d Gyn eco l ogy
In this chapter, the pleural effusions seen in the practice which, in turn, led to the pleural effusion. This does
of obstetrics and gynecology are discussed. The ovarian not appear to be the sole explanation, however,
hyperstimulation syndrome (OHSS) occurring before because the syndrome can still be produced in rabbits
pregnancy, fetal pleural effusions, postpartum pleural when the ovaries are exteriorized (5). This indicates
effusion, Meigs' syndrome, and finally, the pleural that there must be systemic effects involved in the
effusions secondary to endometriosis are addressed. fluid shifts into the peritoneal and pleural cavities. It
is now believed that the syndrome is precipitated by
an ovarian product, vasoactive peptide, or cytokine
OVA R I A N H Y P E RSTI M U LATI O N
that has been released into the peritoneal cavity by the
SYN D RO M E
ovary or that has gained access to the systemic circu
OHSS is a serious complication of ovulation induc lation directly from the corpus luteum or serosal ves
tion with human chorionic gonadotropin (hCG) and sels. Three likely candidates are vascular endothelial
occasionally clomiphene ( 1 ) . This syndrome is char growth factor (VEGF) (6,7) , interleukin 8 (IL-8) (7) ,
acterized by ovarian enlargement, ascites, pleural effu and interleukin 6 (IL-6) (6) . All three have been
sion, hypovolemia, hemoconcentration, and oliguria shown to be markedly elevated in the follicular fluid
(2) . A rare complication is the occurrence of throm and ascites. Antibodies to VEGF reduce the ascites
boembolism related to hemoconcentration (2) . The induced endothelial permeability to 44% of control
severe form with ascites or pleural effusion, or both, and antibodies to IL-8 reduce the ascites-induced
occurs in approximately 3% of patients undergo endothelial permeability to 34% of control (7) .
ing ovulation induction for in vitro fertilization, but Antibodies to IL-6 do not significantly affect ascites
radiologically evident pleural effusions develop only in induced endothelial permeability (7) . It has also been
approximately 1 % (3) . shown that hCG upregulates the VEGF expression of
granulosa cells in the OHSS, but not in control groups
(8) . The serum levels ofVEGF are elevated in patients
Path ogenesis
with OHSS (8) . The mean levels of VEGF in ascitic
OHSS has two primary components: (a) enlargement fluid with the OHSS (7) are higher than the mean level
of the ovaries accompanied by the formation of fol ofVEGF in any type of pleural effusion (7) . The mean
licular, luteal, and hemorrhagic ovarian cysts and level of VEGF in the pleural fluid with the OHSS is
edema of the stroma, and (b) an acute shift of fluid elevated but is only about 60% that in ascitic fluid (9) .
out of the intravascular space (3) . The syndrome is There are probably two factors responsible for the
more frequent in cycles resulting in pregnancy (4) . accumulation of pleural fluid with OHSS. In patients
The exact pathogenesis of this syndrome is not clear. with bilateral effusions, the probable mechanism is a
At one time, it was thought that the OHSS resulted generalized capillary leak syndrome. In patients with
from high local concentrations of estrogen in the ova large right-sided pleural effusions, it is probable that
ries causing altered capillary permeability and ascites, the fluid moves directly from the peritoneal space to the
32 1
322 PLE U RAL D I S EASES
pleural space. Evidence supporting this mechanism is One abnormality now diagnosed on occasion is fetal
the fact that the effusions are frequently large and right pleural effusion. Almost all neonatal pleural effusions
sided, many patients have ascites, and the observation are probably persistent fetal pleural effusions. The
in one patient that the pleural fluid IL-6 level was more prevalence of fetal pleural effusion is approximately 1
than 1 00 times higher than the simultaneously obtained in 1 0,000 deliveries ( 1 5) . The prevalence is approxi
serum level ( 1 0) . Obviously, the pleural fluid in this case mately twice as high in boys as in girls ( 1 6) . There is
was not due to a generalized capillary leak syndrome. one report of a patient who had three children, each
of whom had a fetal pleural effusion ( 1 7) .
Very early i n pregnancy, the incidence o f pleural
C l i n ical M a n ifestati o n s
effusion is higher. In one study of 965 pregnancies
Patients with OHSS initially develop abdominal dis evaluated by ultrasound between 7 and 1 0 weeks, the
comfort and distension, followed by nausea, vomiting, incidence of pleural effusion was 1 .2% ( 1 8) . These
and diarrhea. As the syndrome worsens, the patients early pregnancies complicated by effusion had poor
develop evidence of ascites and then hydrothorax outcomes with miscarriages in 86% ( 1 8) .
or breathing difficulties. In the most severe stages, I f the fetal pleural effusion is untreated, there i s a
the patients develop increased blood viscosity due to high mortality rate. In one series of untreated patients,
hemoconcentration, coagulation abnormalities, and the mortality rate was 37% for 54 untreated fetuses
diminishing renal function (3) . Respiratory symptoms ( 1 5 ) . The high perinatal mortality rate in cases of fetal
develop 7 to 14 days after the hCG injection (4) . pleural effusion is related to three factors: develop
The pleural effusions with OHSS are usually right ment of nonimmune hydrops, prematurity, and pul
sided. In the series of 33 patients with pleural effu monary hypoplasia ( 1 5 ) . lntrathoracic compression of
sions reported by Abramov et al. ( 1 1 ) , 1 7 effusions the developing lung produces pulmonary hypoplasia.
(52%) were right sided, 9 (27%) were bilateral, and This pulmonary hypoplasia can sometimes result in
7 (2 1 %) were unilateral left sided. At times, a pleu perinatal death. The mortality rate is also higher if there
ral effusion may be the sole manifestation of OHSS are associated chromosomal abnormalities or struc
( 1 2) . The pleural effusion can be a significant prob tural abnormalities. Ruano et al. ( 1 9) reported that the
lem, as evidenced by one patient who had 8 , 5 00 mL mortality of fetuses with isolated pleural effusions was
pleural fluid aspirated from her pleural space over 5 of 1 4 (36%) , of those with structural abnormalities
1 4 days ( 1 3) . The pleural fluid in patients with OHSS and no chromosomal abnormalities 1 9 of 1 9 ( 1 00%) ,
is an exudate. In the series of Abramov et al. ( 1 4) , the of those with chromosomal abnormalities 20 of 23
mean pleural fluid protein was 4 . 1 g/dL, whereas the (87%) . The most common chromosomal abnormality
mean serum protein was 4.4 g/dL. in the series of Ruano et al. ( 1 9) was Turner syndrome,
The incidence of the syndrome can be reduced if which occurred in 1 5 of the 23 patients (65%).
the serum estrogen levels and the number of ovarian
follicles are monitored. If the serum estrogen levels are
Pathog e n e s i s a n d C l i n ical M a n ifestati o n s
very high or if there are more than 1 5 ovarian follicles
with a high proportion of small and intermediate size The pathogenesis of fetal pleural effusion is not
follicles, hCG should be withheld (2) . known and is possibly multifactorial. There is some
evidence that the fetal pleural effusions are actually
chylothoraces. Benacerraf and Frigoletto (20,2 1 ) ana
Treatment
lyzed the pleural fluid from two fetuses and reported
The treatment of the OHSS is primarily supportive abundant lymphocytes in both, but one had almost
( 1 ) . Hemoconcentration should be treated with intra all T lymphocytes, whereas the other had a mixture
venous fluids, because hypovolemia can lead to renal of T and B lymphocytes. The presence of the lym
failure and even death. If the patient has a large pleural phocytes was suggestive of a chylous effusion. A ratio
effusion and is dyspneic, a therapeutic thoracentesis of the pleural fluid to serum lgG that exceeds 0.6 is
should be performed. Rarely is more than one thera very suggestive of a neonatal pleural effusion (22) .
peutic thoracentesis necessary ( 1 ) . In addition, most neonatal pleural effusions, which
in many cases are continuations of fetal pleural effu
sions, are chylothoraces (23) . Analysis of the pleural
F ETA L P LE U RA L E F F U S I O N
fluid for chylomicrons is not useful in the diagnosis
Th e ability to diagnose fetal abnormalities prena of fetal chylothorax because the fetuses are not eating
tally has been facilitated by diagnostic ultrasound. any lipids.
C HAPTE R 20 I P L E U RAL D I S EASE I N O B STETRICS A N D G Y N E CO LOGY 323
In a review of the literature for isolated fetal pleu to screen for bacterial or viral infection. At the same
ral effusion in 1 998, 204 cases were found (24) . The time, a diagnostic thoracentesis should be performed
fetal pleural effusions were bilateral in 7 4%, unilateral and the pleural fluid sent for culture, cell analysis,
right sided in 1 1 %, and unilateral left sided in 1 4% and biochemical study. The lung size and the fetal
(24) . Polyhydrarnnios was noted in 72% and hydrops lung distensibility are assessed through ultrasound
fetalis in 57% (24) . The reason for the polyhydram before and after the thoracentesis. Fetuses that have
nios is not clear, but it has been suggested that the less-than-normal lung expansion are then subjected
increased intrathoracic pressure may interfere with to surgical intervention.
normal fetal swallowing. There is a high frequency of The possible surgical interventions for fetal pleu
chromosomal abnormalities in fetuses with pleural ral effusion are pleuroamniotic shunt or repeated
effusions. In one series, the prevalence of chromosomal therapeutic thoracentesis. Rodeck et al. (28) reported
abnormalities was 50% in 1 52 fetuses with other sono their results with the implantation of pleuroamniotic
graphic abnormalities and 1 2% in 94 patients with shunts in eight human fetuses with pleural effusion.
isolated pleural effusion (25) . These shunts were established with double-pigtail
nylon catheters with external and internal diam
eters of 0 .2 1 and 0 . 1 5 mm, respectively. The shunts
Treatment
were introduced transamniotically under ultrasound
The optimal management for fetuses with pleural visualization through the fetal midthoracic wall into
effusions is controversial ( 1 5 ) . If the pleural effusions the effusion. All eight fetuses in this series had large
are not treated, some resolve spontaneously whereas pleural effusions with hydramnios. Twelve pleu
others deteriorate to generalized hydrops. In one roamniotic shunts were placed in these fetuses. One
review of 204 cases, spontaneous remission occurred shunt was noted to be free in the amniotic cavity
in 22% (24) . Characteristics of cases that resolved 1 week after insertion and a second shunt was inserted.
spontaneously included diagnosis made early in the The remaining 1 1 shunts functioned until delivery
second trimester, unilateral effusion, and no associ (median 2 . 5 weeks; range 1 - 1 4 weeks) . In six fetuses,
ated polyhydramnios or hydrops (24) . Some infants the pleural effusions almost completely resolved and
die from pulmonary hypoplasia after delivery, whereas the hydramnios disappeared. Three of the six fetuses
others survive. In a review of 82 cases in 1 993, the had hydrops that resolved after the insertion of the
mortality rates of those treated surgically and those shunt. All six infants survived, and five had no respi
treated conservatively were comparable ( 1 5 ) . Overall, ratory difficulty at birth. Fetal hydrothoraces did not
the mortality rate in cases not terminated by abortion resolve in two patients, both of whom had hydrops
was 36% ( 1 5) . However, in a second review in 1 99 8 , and died shortly after delivery. In a second study, Blott
43% of the fetuses with treatment had a good out et al. (29) inserted shunts into 1 1 fetuses between 24
come whereas only 22% without treatment had and 35 weeks of gestation and reported that the effu
a favorable outcome (24) . In a more recent review sions were successfully drained in all cases and that
(26) of 1 72 fetuses with isolated pleural effusion, the 8 of the 1 1 fetuses survived. It should be noted that
overall survival rate was 63% with various treatment the shunts become displaced intrathoracically in a
approaches. sizeable percentage of fetuses, but there appears to be
The following scheme is suggested for the manage no long-term pulmonary complications and they do
ment offetal pleural effusion. When a pleural effusion not need to be removed (30) .
is discovered in a fetus, a thoracentesis is performed An alternative approach to the management of the
immediately if there is fetal distress or if there is dia fetal pleural effusion is to perform serial thoracenteses.
phragmatic inversion or shift of the mediastinum to Benacerraf and Frigoletto (20,2 1 ) performed three
the contralateral side (27) . Otherwise the fetus is eval to five thoracenteses on two fetuses between 20 and
uated for other abnormalities. Then a repeat ultra 24 weeks of gestation with massive pleural effusions.
sound is obtained in 2 to 3 weeks. If the effusion has The effusion did not recur after the last thoracentesis,
decreased in size, then the fetus is followed with scans and normal babies resulted from both pregnancies.
every 2 to 3 weeks. If the effusion is stable, then serial There have been case reports of attempts to create
scans are performed and a thoracentesis is performed a pleurodesis in the fetus by the intrapleural injection
immediately before delivery to facilitate neonatal of the immunostimulant OK-432 (3 1 ) or maternal
resuscitation. If the effusion increases in size, then blood (32) . In general, pleurodesis is not recom
a diagnostic amniocentesis should be performed for mended because the long-term side effects from a fetal
chromosomal analysis and culture of amniotic fluid pleurodesis remain to be determined. In fetuses that
324 PLE U RAL D I S EASES
are known to have large effusions j ust before delivery, to that in nonpregnant women of the same age (39) . In
if a cesarean section is performed, a thoracentesis can general, I would guess that the leading cause of pleu
be performed while the fetoplacental circulation is ral effusion during pregnancy is pulmonary embolism.
still intact (33 ) . Performance of the thoracentesis in The risk of venous thromboembolism among pregnant
this manner preserves systemic oxygenation (33) . or postpartum women is 4.29 times greater compared
with nonpregnant women (40) . The risk of pulmo
Lo n g -Term Prog nosis
nary embolism is much higher (- 1 5 times) in the
postpartum period than during pregnancy (40) . Other
The prognosis of fetuses that have pleural effusion and common causes of pleural effusions include pneumo
that receive pleuroamniotic shunting appears to be nia with effusion and viral illnesses. The incidence of
good. Thompson et al. (34) studied 17 infants who had pleural effusions is low (-3%) in patients with severe
undergone pleuroamniotic shunting for a fetal pleural preeclampsia or the hemolysis, elevated liver enzymes,
effusion at a median age of 1 2 months. They reported and low platelet count (HELLP) syndrome (4 1 ) . If
that respiratory symptoms and respiratory function the patient has a transudative effusion, peripartum
were no different in these 17 infants than those in a cardiomyopathy, which complicates approximately 1 in
control group (34) . 300 pregnancies, should be considered.
The mechanism responsible for the pleural effusion is pseudo-Meigs' syndrome (52) . Nevertheless, I classify any
unknown. No therapeutic intervention is necessary patient with a pelvic neoplasm associated with ascites
in the absence of symptoms or signs of illness (42) . and pleural effusion, in whom surgical extirpation of
the tumor results in permanent disappearance of the
Del ayed Postpa rtu m ascites and pleural effusion, as having Meigs' syndrome.
in approximately 70% of patients, is left sided in Some patients have exacerbation of their symptoms
1 0%, and is bilateral in 20% (59) . The only symptom coincident with menses.
referable to pleural effusion is shortness of breath. The treatment of the massive ascites, pleural
Ascites may not be evident on physical examination. effusion, and endometriosis is difficult. Hormonal
The pleural fluid is usually an exudate. Although therapy (progestational agents, danazol, or luprolide
several authors have stated that the pleural fluid with acetate [Lupron] ) fails in at least 50% of cases. The
Meigs' syndrome is a transudate (56,6 1 ) , this opinion most common treatment is total abdominal hysterec
appears to be based on the gross appearance of the tomy and bilateral salpingo-oophorectomy, but this is
fluid rather than on its protein levels. Most pleural difficult owing to the pelvic endometriosis (66) .
fluids secondary to Meigs' syndrome have a protein
level above 3 . 0 g/dL (56, 6 1 -64) . The pleural fluid
usually has a low white blood cell (WBC) count REFERENCES
(fewer than l ,OOO/mm3) and is occasionally bloody
1 . Rozenberg S. Review of clinical course and treatment of ovar
(56,64) . At times, the level of CA 1 2 5 in the pleural ian hyperstimulation syndrome (OHSS) . Hum Reprod Update.
fluid is elevated and this should not be taken as an 2003;9:77-96.
indication of malignancy (52,65) . 2 . Rizk B , Aboulghar M . Modern management of ovarian hyper
stimulation syndrome. Hum Reprod. 1 9 9 1 ;6: 1 082- 1 087.
3 . Levin MF, Kaplan BR, Hutton LC. Thoracic manifestations
D i a g nosis a n d M a n agement of ovarian hyperstimulation syndrome. Can Assoc Radio! ].
l 995;46:23-26.
The diagnosis of Meigs' syndrome should be consid 4. Gregory WT, Patton PE. Isolated pleural effusion in severe
ered in all women who have pelvic masses, ascites, and ovarian hyperstimulation: a case report. Am J Obstet Gynecol.
pleural effusions. If in such patients cytologic exami 1 999; 1 80: 1 468-1471 .
5. Polishuk WZ, Schenker JG. Ovarian overstimulation syn
nation of the ascitic and pleural fluid is negative, an
drome. Fertil Steril. 1 969;20:24 1 -249.
exploratory laparotomy or at least a diagnostic lapa 6. Loree de Mola JR. Pathophysiology of unilateral pleural effu
roscopy should be performed with surgical removal sions in the ovarian hyperstimulation syndrome. Hum Reprod.
of the primary neoplasm. The diagnosis is confirmed 1 999; 1 4 : 272-273.
when the ascites and the pleural fluid resolve postop 7. Chen SU, Chou CH, Lin CW, et al. Signal mechanisms
of vascular endothelial growth factor and interleukin-8 in
eratively and do not recur. Postoperatively, the pleural ovarian hyperstimulation syndrome: dopamine targets their
fluid disappears from the chest rapidly and is usually common pathways. Hum Reprod. 2 0 1 0;25:757-767.
completely gone within 2 weeks (63) . 8 . Wang TH, Horng SG, Chang CL, et al. Human chorionic
gonadotropin-induced ovarian hyperstimulation syndrome is
associated with up-regulation of vascular endothelial growth
factor. J Clin Endocrinol Metab. 2002;87:3300-3308.
E N D O M ETRI O S I S
9 . Chen CD, Wu MY, Chen HF, et al. Prognostic importance
At times, severe endometriosis is complicated by mas of serial cytokine changes in ascites and pleural effusion in
women with severe ovarian hyperstimulation syndrome. Fertil
sive ascites (66) . In a substantial proportion of patients,
Steril. 1 999;72:286-292.
a pleural effusion is also present. Muneyyirci-Delale 10. Loree de Mola JR, Farredondo-Soberon F, Randle CP, et al.
et al. (66) presented 4 cases and reviewed the literature Markedly elevated cytokines in pleural effusion during the
with 23 additional cases and found that 8 of the 27 ovarian hyperstimulation syndrome: transudate or ascites?
patients (30%) also had a pleural effusion. The pleural Fertil Steril. 1 997;67:780-782.
1 1 . Abramov Y, Elchalal U, Schenker JG. Pulmonary manifesta
effusion was described as a bloody exudate and in one
tions of severe ovarian hyperstimulation syndrome: a multi
instance had an elevated pleural fluid level of CA 1 2 5 center study. Fertil Steril. 1 999;71 :645-65 1 .
(66) . Th e pleural effusion i s thought t o b e due t o ascitic 1 2 . Rabinerson D , Shalev J, Royburt M , et al. Severe unilateral
fluid gaining entrance to the pleural cavity through the hydrothorax as the only manifestation of the ovarian hyperstim
diaphragm, as in Meigs' syndrome. The pleural effusion ulation syndrome. Gynecol Obstet Invest. 2000;49 : 1 40- 1 42.
13. Yuen BH, McComb P, Sy L, et al. Plasma prolactin, human
is either right sided or bilateral (67) . chorionic gonadotropin, estradiol, testosterone, and proges
Most patients initially present with abdominal terone in the ovarian hyperstimulation syndrome. Am J Obstet
distension, pain, anorexia, and nausea. Because sig Gynecol. 1 979; 1 33:3 1 6-320.
nificant weight loss occurs in many of the patients, the 14. Abramov Y, Elchalal U, Schenker JG. Febrile morbidity in
usually presumptive diagnosis is malignancy. How severe and critical ovarian hyperstimulation syndrome: a mul
ticentre study. Hum Reprod. 1 99 8 ; 1 3 : 3 1 28-3 1 3 1 .
ever, a large proportion also has clinical manifestations 1 5 . Hagay Z , Reece A, Roberts A, e t al. Isolated fetal pleural
of endometriosis, such as progressive dysmenorrhea effusion: a prenatal management dilemma. Obstet Gynecol.
and cul-de-sac and uterosacral ligament nodularity. 1 993;8 l : 1 47- 1 52 .
C HAPTE R 20 I P L E U RAL D I S EASE I N O B STETRICS A N D G Y N E CO LOGY 327
1 6 . Eddleman KA, Levine AB, Chickara U, et al. Reliability of 38. Kocijancic K, Kocijancic I, Vidmar G . Sonography of
pleural fluid lymphocyte counts in the antenatal diagnosis of pleural space in healthy individuals. ] Clin Ultrasound.
congenital chylochorax. Obstet Gynecol. 1 9 9 1 ;78: 530-5 32. 2005;33:3 86-389.
1 7 . Chang YL, Lien R, Wang CJ, et al. Congenital chylochorax in 39. Light RW. Pleural diseases in pregnancy. Int Pleural News!.
three siblings. Am ] Obstet Gynecol. 2005; 1 92:2065-2066. 2006;4:6-7.
1 8 . Hashimoto K, Shimizu T, Fukuda M, et al. Pregnancy out 40. Heit JA, Kobbervig CE, James AH, et al. Trends in the
come of embryonic/fecal pleural effusion in the first trimester. incidence of venous chromboembolism during pregnancy or
] Ultrasound Med. 2003;22:50 1-505. postpartum: a 30-year population-based study. Ann Intern
19. Ruano R, Ramalho AS, Cardoso AK, et al. Prenatal diagnosis Med. 2005 ; 1 43:749-750.
and natural history of fetuses presenting with pleural effusion. 4 1 . Haddad B, Barron JR, Livingston JC, et al. HELLP (hemoly
Prenat Diagn. 2 0 1 1 ; 3 1 :496-499 . sis, elevated liver enzymes, and low platelet count) syndrome
20. Benacerraf B R , Frigolecco FD Jr. Mid-trimester fecal choracen versus severe preeclampsia: onset at < or = 28 weeks' gesta
cesis. ] Clin Ultrasound. 1 98 5 ; 1 3 :202-204. tion. Am ] Obstet Gynecol. 2000; 1 83 : 1 475-1 479.
2 1 . Benacerraf BR, Frigoletto FD Jr, Wilson M . Successful 42. Hughson WG, Friedman PJ, Feigin OS, et al. Postpartum
midtrimester choracentesis with analysis of the lymphocyte pleural effusion: a common radiologic finding. Ann Intern
population in the pleural effusion. Am ] Obstet Gynecol. 1 9 86; Med. 1 9 82;97: 8 5 6-8 5 8 .
1 5 5 :398-399. 4 3 . Udeshi UL, McHugo J M , Crawford J S . Postpartum pleural
22. Tsukimori K, Nakanami N, Fukushima K, et al. Pleural fluid/ effusion. Br J Obstet Gynaecol. 1 98 8 ; 9 5 : 894-897.
serum immunoglobulin ratio is a diagnostic marker for con 44. Wallis MG, McHugo JM, Carruthers DA, et al. The preva
genital chylochorax in utero . ] Perinat Med. 2006;34:3 1 3-3 1 7 . lence of pleural effusions in pre-eclampsia: an ultrasound
23. Chernick V, Reed MH. Pneumo thorax and chylochorax in the study. Br J Obstet Gynaecol. 1 989;96:43 1 -433.
neonatal period. J Pediatr. l 970;76: 624-632. 4 5 . Gourgoulianis Kl, Karantanas AH, Diminikou G, et al. Benign
24. Aubard Y, Derouineau [, Aubard V, et al. Primary fecal hydro postpartum pleural effusion. Eur Respir J 1 99 5 ; 8 : 1 748- 1 75 0 .
chorax: a literature review and proposed antenatal clinical 46. Kochenour N K , Branch O W, Rote NS, et a l . A new postpar
strategy. Fetal Diagn Ther. 1 99 8 ; 1 3 :325-333. tum syndrome associated with antiphospholipid antibodies.
25. Waller K, Chaichongwongwatthana S, Yamasmit W, et al. Obstet Gynecol. 1 987;69 :460-468.
Chromosomal abnormalities among 246 fetuses with pleural 47. Ayres MA, Sulak PJ. Pregnancy complicated by antiphospho
effusions detected on prenatal ultrasound examination: factors lipid antibodies. South MedJ 1 9 9 1 ;84:266-269.
associated with an increased risk of aneuploidy. Genet Med. 48. Garg R, Sanjay Das V, et al. Spontaneous pneumothorax: An
2005 ;7:4 1 7-42 1 . unusual complication of pregnancy-a case report and review
26. Deurloo KL, Devlieger R, Lopriore E , et al. Isolated fecal of literature. Ann Thorac Med. 2008;3 : 1 04-1 0 5 .
hydrochorax with hydrops: a systematic review of prenatal 49. Meigs J V, Cass JW. Fibroma o f the ovary with ascices and
treatment options. Prenat Diagn. 2007;27: 893-899. hydrochorax. Am ] Obstet Gynecol. 1 937;33 :249-267.
27. Yamamoto M, Insunza A, Carrillo J, et al. lntrathoracic pressure 50. Meigs JV. Pelvic tumors other than fibromas of the ovary with
in congenital chylochorax. Keystone for the rationale of chora ascices and hydrochorax. Obstet Gynecol. 1 954;3 :47 1-486.
coamniocic shunting? Fetal Diagn Ther. 2007;22: 1 69-1 7 1 . 5 1 . Yu J, Grimes DA. Ascites and pleural effusions associated with
28. Rodeck CH, Fisk NM, Fraser D [, et al. Long-term in utero drain endometriosis. Obstet Gynecol. 1 9 9 1 ;7 8 : 5 33-5 34.
age offetal hydrochorax. N Engl] Med. 1 988;3 1 9 : 1 1 3 5-1 1 3 8 . 52. Loizzi V, Cormio G, Resca L, et al. Pseudo-Meigs syndrome
29. Blott M, Nicolaides K H , Greenough A . Pleuroamniotic and elevated CA 1 2 5 associated with struma ovarii. Gynecol
shunting for decompression of fetal pleural effusions. Obstet Oncol. 2005;97:282-284.
Gynecol. 1 9 8 8 ; 1 02:28 8-290. 53. Abramov Y, Anteby SO, Fasouliotis SJ, et al. The role of
30. Sepulveda W, Galindo A, Sosa A, et al. lntrachoracic dislodge inflammatory cycokines in Meigs' syndrome. Obstet Gynecol.
ment of pleura-amniotic shunt. Three case reports with long 2002;99(5 suppl 1 ) : 9 1 7-9 1 9 .
term follow-up. Fetal Diagn Ther. 2005;20 : 1 02- 1 0 5 . 5 4 . Abramov Y, Anteby S O , Fasouliocis SJ, e t al. Markedly ele
3 1 . Chen M, Chen C P, Shih J C , e t al. Antenatal treatment of vated levels of vascular endothelial growth factor, fibroblast
chylochorax and cystic hygroma with OK-432 in nonimmune growth factor, and interleukin 6 in Meigs syndrome. Am J
hydrops fecalis. Fetal Diagn Ther. 2005;20:309-3 1 5 . Obstet Gynecol. 200 1 ; 1 84:354-3 5 5 .
32. Parra ] , Amenedo M , Muniz-Diaz E , e t al. A new success 5 5 . Samanth KK, Black WC I l l . Benign ovarian stromal tumors
ful therapy for fecal chylochorax by intrapleural injection of associated with free peritoneal fluid. Am J Obstet Gynecol.
maternal blood. Ultrasound Obstet Gynecol. 2003;22:290-294. 1 970; 1 07:538-54 5 .
33. Prontera W, Jaeggi ET, Pfizenmaier M, et al. Ex ucero intrapar 56. Meigs JV. Fibroma of the ovary with ascites and hydrothorax.
tum treatment (EXIT) of severe fecal hydrochorax. Arch Dis Meigs' syndrome. Am ] Obstet Gynecol. 1 9 54;67:962-9 87.
Child Fetal Neonatal Ed. 2002;86:F5 8-F60. 57. Kirschner PA. Porous diaphragm syndromes. Chest Surg Clin
34. Thompson PJ, Greenough A, Nicolaides KH. Respiratory N Am. 1 99 8 ; 8 :449-472.
function in infancy following pleura-amniotic shunting. Fetal 5 8 . Lemming R. Meigs' syndrome and pathogenesis of pleurisy
Diagn Ther. 1 993;8 :79-83. and polyserositis. Acta Med Scand. 1 960; 1 6 8 : 1 97-204.
3 5 . Rocha G . Pleural effusions in the neonate. Curr Opin Pulm 5 9 . Majzlin G, Stevens FL. Meigs' syndrome: case report and
Med. 2007; 1 3 :305-3 1 1 . review of literature. J Int Coll Surg. 1 9 64;42:625-630.
36. Shih YT, Su PH, Chen JY, et al. Common etiologies of neona 60. Patsner B . Meigs syndrome and "false positive" preoperative
tal pleural effusion. Pediatr Neonatol. 20 1 1 ;52:25 1 -25 5 . serum CA- 1 2 5 levels: analysis of ten cases. Eur J Gynaecol
37. Kocijancic I, Pusenjak S, Kocijancic K , e t al. Sonographic Oncol. 2000;2 1 :362-363.
detection of physiologic pleural fluid in normal pregnant 6 1 . O'Flanagan SJ, Tighe BF, Egan TJ, et al. Meigs' syndrome and
women. J Clin Ultrasound. 2005;33:63-66. pseudo-Meigs' syndrome. ] R Soc Med. 1 9 87;80 :252-253.
32 8 PLE U RAL D I S EASES
62. Hurlow RA, Greening WP, Krantz E. Ascites and hydrorhorax 66. Muneyyirci-Delale 0, Neil G, Serur E, e t al. Endomerriosis
in association with srroma ovarii. Br J Surg. 1 976;63 : 1 1 0- 1 1 2 . with massive ascites. Gynecol Oncol. l 998;69:42-46.
63. Jimerson SD. Pseudo-Meigs' syndrome: a n unusual case with 67. Bhojawala J, Heller OS, Cracchiolo B, et al. Endometrio
analysis of the effusions. Obstet Gynecol. 1 973;42:53 5-537. sis presenting as bloody pleural effusion and ascites-report
64. Neustadt JE, Levy RC. Hemorrhagic pleural effusion in of a case and review of the literature. Arch Gynecol Obstet.
Meigs' syndrome. JAMA. 1 968;204: 1 79- 1 80. 2000;264:39-4 1 .
6 5 . Timmerman 0, Moerman P, Vergote I. Meigs' syndrome with
elevated serum CA 1 25 levels: two case reports and review of
the literature. Gynecol Oncol. l 9 9 5 ; 5 9 :405--40 8 .
P l e u ra l D i se a s e D u e to Co l l a g e n
Va sc u l a r D i se a ses
329
33 0 PLE U RAL D I S EASES
D i a g nosis
These features were not seen in any of 1 0,000 other 63 (83%) had pleural fluid glucose levels below
pleural fluids due to diverse causes (8). 50 mg/dL (6) . The explanation for the low pleural
The pleural fluid from patients with rheumatoid fluid glucose in this condition is not known precisely.
pleuritis may contain "ragocytes" or RA cells. The If the serum level of glucose is increased in patients
term ragocyte was coined by Delbarre et al. ( 1 1 ) who with rheumatoid pleural effusions, little change is
described small, spherical, cytoplasmic inclusions in seen in the pleural fluid glucose levels ( 1 6- 1 8) , but
neutrophilic leukocytes and occasionally in mono similar results are obtained in patients with low
cytes in unstained wet films of the sediment obtained pleural fluid glucose levels from other diseases ( 1 9) .
from the synovial fluid of patients with various types I n contrast, when patients with rheumatoid pleu
of arthritis. These inclusions were reminiscent of ral effusions are given oral urea ( 1 8) or intravenous
raisin seeds; hence, the adoption of the prefix rago, d-xylose ( 1 7) loads, the pleural fluid and serum levels
which is derived from the Greek word for grape. of these substances equilibrate over several hours.
The inclusion bodies have been shown to represent Carr and McGuckin ( 1 8) have suggested that the
phagocytic vacuoles or phagosomes, which are of rheumatoid inflammatory process alters the normal
greater size than normal lysosomes of granular leu state of one or more enzymes that constitute the
kocytes ( 1 2) . The presence of these cells is not use carbohydrate transport mechanism of cellular mem
ful diagnostically because pleural effusions of other branes. This interpretation should be viewed with
etiologies, particularly those with a low glucose level, some caution. The relationship between the serum
contain these cells (8, 1 3) . and pleural fluid glucose levels is dictated not only
by the ease with which glucose passes from the serum
Concomitant Infection
into the pleural fluid but also by the rate at which
When a patient is seen with RA and a pleural effusion the pleural surfaces and fluid use the glucose. Because
characterized by a low glucose level ( <20 mg/dL) , the pleural fluid glucose level falls within 30 minutes
a low pH (<7.20), and a high LDH level, one must from 2,000 to 236 mg/dL after the intrapleural injec
rule out pleural infection, which can produce pleural tion of glucose in patients with rheumatoid pleuritis
fluid with the same characteristics (see Chapter 1 2) . ( 1 6) , there must be either rapid glucose uptake by the
Hindle and Yates ( 1 4) first reported a pyopneumo pleura or no great barrier to its diffusion.
thorax in a patient with a rheumatoid pleural effu The pleural surfaces with rheumatoid pleuritis
sion. At thoracotomy, it was found that a necrobiotic appear to be active metabolically, as manifested by
nodule in the visceral pleura had broken down, pro the high pleural fluid LDH and the low pleural fluid
ducing a bronchopleural fistula. Jones and Blodgett glucose levels, although the metabolic activity of
( 1 5) subsequently reported that 5 of 1 0 patients with rheumatoid pleural fluid is virtually nil even when
rheumatoid pleural effusion followed for a 5-year glucose is added ( 1 7) . The thickened pleura in rheu
period developed empyemas. These investigators matoid pleuritis probably limits the movement of
found that empyemas were more common in patients glucose into the pleural space, and because glucose
who had been treated with corticosteroids, and they consumption by the pleural surfaces is high, an equi
attributed the pleural infection to the creation of a librium is formed in which the pleural fluid glucose
bronchopleural fistula through necrobiotic subpleu level is much lower than the serum glucose level.
ral rheumatoid nodules.
Cholesterol levels
When a patient with an apparent rheumatoid
pleural effusion is seen, it is important to obtain both Another interesting characteristic of rheumatoid
aerobic and anaerobic cultures of the pleural fluid. In pleural effusions is their tendency to contain choles
addition, the pleural fluid should be centrifuged, and terol crystals or high levels of cholesterol. Ferguson
the sediment should be Gram stained because stains (5) first reported on two patients with rheumatoid
made in this manner are more sensitive than those pleural effusions in whom the pleural fluid contained
made on uncentrifuged pleural fluid. numerous cholesterol crystals. Subsequently, Naylor
(8) reported that 5 of 24 rheumatoid pleural fluids
Glucose levels
(2 1 %) contained cholesterol crystals. Some rheuma
The most striking characteristic of the rheumatoid toid pleural effusions contain high levels of cholesterol
pleural effusion is its low glucose content. In a review without cholesterol crystals (6) .
of76 patients with rheumatoid pleuritis, 48 (63%) had Lillington et al. (6) measured the lipid levels in
pleural fluid glucose levels below 20 mg/dL, whereas seven rheumatoid pleural effusions and found levels
332 PLE U RAL D I S EASES
above 1 ,000 mg/dL in four of the seven fluids. One be treated with nonsteroidal anti-inflammatory drugs
of the two patients whom I have seen with cholesterol such as aspirin or ibuprofen for 8 to 1 2 weeks initially.
crystals in the pleural fluid had rheumatoid pleuritis. If the pleural effusion persists and if the j oint symp
The cholesterol crystals impart a sheen to the fluid toms are not well controlled, then appropriate therapy
when viewed with the naked eye under proper light should be directed toward the rheumatologic prob
ing. High cholesterol levels make the pleural fluid lem. If the only symptomatic problem is the pleural
turbid. The significance of the presence of high levels disease, then the patient should have a therapeutic
of cholesterol or cholesterol crystals in the pleural thoracentesis and possibly an intrapleural injection
fluid is unknown. Cholesterol pleural effusions are of corticosteroids. There have been two reports con
discussed more extensively in Chapter 26. cerning the intrapleural injection of corticosteroids;
the first (22) had two patients, and the intrapleural
Biopsy
corticosteroids were ineffective; the second (26) had
Closed pleural biopsies have a limited role in the one patient who seemed to respond to one injection
diagnosis of rheumatoid pleuritis. Although a pleu of 1 20 mg of depomethylprednisolone.
ral biopsy specimen may reveal a rheumatoid nodule Decortication should be considered in patients
diagnostic of rheumatoid pleuritis in an occasional with thickened pleura who are symptomatic with
patient, the pleural biopsy usually reveals only chronic dyspnea. Computed tomographic examination is use
inflammation or fibrosis. Pleural biopsy is not recom ful in delineating the extent of the pleural thickening.
mended in the typical case of rheumatoid pleuritis. In patients with pleural effusions, the significance
In atypical cases, however, such as in patients without of the pleural thickening can be gauged by measur
arthritis or in those with a normal pleural fluid glucose ing the pleural pressure serially during a therapeutic
level, thoracoscopy, or pleural biopsy should be per thoracentesis (see Chapter 28). If the pleural pressure
formed to rule out malignant disease and tuberculosis. drops rapidly as pleural fluid is removed, the lung is
trapped by the pleural disease (27) , and decortication
Prog nosis a n d Treatment
should be considered. The decortication procedure is
difficult to perform in patients with rheumatoid pleu
The natural history of rheumatoid pleuritis is vari ritis because it is not easy to develop a plane between
able. In the series of Walker and Wright ( 1 ) , 1 3 of the lung and the fibrous peel. Therefore, air leaks
1 7 patients (76%) had spontaneous resolution of persist longer than usual after decortication (25) .
their pleural effusions within 3 months, although Nevertheless, decortication can substantially improve
1 of the 1 3 patients had a subsequent recurrence. One the quality of life of some patients with dense pleural
patient had a spontaneous resolution after 1 8 months fibrosis secondary to rheumatoid disease.
of observation, whereas another had a persistent effu As mentioned earlier, patients with rheumatoid
sion for more than 2 years. One patient developed pleural effusions have a high incidence of compli
progressive severe pleural thickening and eventu cated parapneumonic effusions. The management of
ally had to undergo a decortication. The last patient such patients is the same as for any patient with com
developed an empyema. plicated parapneumonic effusion (see Chapter 1 2) .
Little information is available in the literature on Th e incidence o f persistent bronchopleural fistula is
the efficacy of therapy in rheumatoid pleural disease. higher in the patient with rheumatoid disease, and
Some patients have appeared to respond to systemic more exploratory thoracotomies are required (25 ) .
corticosteroids ( 1 ) , whereas in others no beneficial
effects were observed (20-22) . The degree of activity
Sti l l's D i sease
in the pleural space and in the joints is not neces
sarily parallel. In one report, the administration of Pleural effusions sometimes occur with adult onset
methotrexate was associated with improvement in the Still's disease. Uppal et al. (28) reported 1 7.9% of 28
arthritis but also with the development of a pleural patients with adult onset Still's disease had a pleural
effusion (23 ) . The main goal of therapy should be to effusion. The pleural fluid with adult onset Still's disease
prevent the progressive pleural fibrosis that may neces is an exudate with a predominance of neutrophils (29) .
sitate a decortication in a small percentage of patients
( 1 ,4,2 1 ,24,25) . There are no controlled studies evalu
SYST E M I C L U P U S E RYTH E M ATO S U S
ating the efficacy of corticosteroids or nonsteroidal
anti-inflammatory drugs in the treatment of rheuma Both systemic and drug-induced lupus erythematosus
toid pleural effusion. It is recommended that patients (LE) may affect the pleura.
C HAPTER 2 1 I PLE U RAL D I S EAS E D U E TO COLLAG E N VASC U LAR D I S EASES 333
I n c i d e nce
Tzelepis (4 1 ) reviewed 77 cases of the shrinking lung be within one dilution of each other. However, the
syndrome from 1 965 to 2008 and reported that 65% pleural fluid ANA titers were above 1 :40 in 8 of the
had pleuritic chest pain. 74 patients ( 1 0.8%) who did not appear to have lupus
Of course, patients with SLE may have pleural and in 3 of the patients the titers were greater than
effusions for reasons other than SLE. Patients with the 1 60. The staining pattern in the patients with lupus
nephrotic syndrome may have hypoproteinemia and pleuritis tended to be homogeneous, whereas it tended
pleural effusions on this basis. In addition, patients to be speckled in the patients without lupus, but again
with SLE may have uremia, pericardia! effusions, pneu there was some overlap. The patients with lupus also
monia, pulmonary emboli, congestive heart failure, or tended to have higher titers for specific ANA to sin
other disorders that can produce pleural effusions. gle-stranded DNA (ssDNA) , double-stranded DNA
(dsDNA) , smooth muscle, and ribonucleoprotein (43) .
In another series (44), ANA titers were performed on
D i a g nosis
266 pleural fluids including 1 6 with SLE. The titers
The possibility of lupus pleuritis should be considered were � 1 : 1 60 in all patients with SLE but also exceeded
in any patient with an exudative pleural effusion of this level in 1 6 patients (6.4%) with pleural effusion
unknown etiology. due to other diseases (44) .
Another study evaluated the ANA titers of 1 26
Pleural Fluid Examina tion
pleural fluids including 7 due to SLE (45 ) . In this
The pleural fluid is usually a yellow or serosanguine study, the ANA tests were performed using a com
ous exudate. The differential white blood cell (WBC) mercially available kit that used an indirect immuno
count on the pleural fluid may reveal a preponderance fluorescent antibody method with a human epithelial
of polymorphonuclear leukocytes or mononuclear 2 (HEP-2) cell line. Although all the pleural fluids
cells (37) . from patients with SLE had titers greater than 1 :320,
Halla et al. (7) reported that measurement of the the pleural fluid ANA titer and pattern essentially
pleural fluid glucose, LD H, and pH levels were useful in mimicked the titer and pattern in the serum (45 ) .
distinguishing rheumatoid pleural effusions from lupus I n addition, the pleural fluid ANA titers were greater
effusions. They reported that patients with lupus pleu than 1 : 1 60 in 1 3 other patients, including 1 1 patients
ritis had a pleural fluid glucose level above 80 mg/dL, with malignant effusions, 1 with tuberculous pleuritis,
an LDH level less than two times the upper limit and 1 with an empyema due to amebiasis. In each
for serum, and a pH above 7.20, whereas patients instance, the pleural fluid ANA titer was within one
with rheumatoid pleuritis had a glucose level below dilution of the serum ANA titer (45 ) .
25 mg/dL, an LDH level more than two times the Th e demonstration o f L E cells i n pleural fluid is
upper limit for serum, and a pH below 7.20. These suggestive of lupus pleuritis (46) . In one recent study,
biochemical tests do not always distinguish between the LE cell test was positive in 8 of 1 1 patients who
lupus and rheumatoid pleuritis because an occasional were thought to have lupus pleuritis. Moreover, the LE
patient with lupus pleuritis will have a low pleural fluid cell test on the pleural fluid was negative in 1 3 patients
glucose, a high pleural fluid LDH, or a low pleural with serum and effusion ANA titers greater than 1 : 1 60
fluid pH (37) . In one report of a pleural effusion due to who were not thought to have SLE. However, because
procainamide, the pleural fluid was an exudate with a the LE cell tests on the pleural fluid always correlated
WBC count of 53,200/mm3, an LDH of 4,296 IU/L, with the LE test results on the serum, the same test on
a pH of 7. 1 95, and a glucose of 79 mg/dL (42) . the pleural fluid provided no additional information
Although in the past it was believed that the most (46). There have been false-positive reports with the
useful test for establishing the diagnosis of lupus LE cell test on the pleural fluid (47) .
pleuritis was the measurement of the ANA level in On the basis of the studies mentioned in the pre
the pleural fluid, this does not appear to be the case. ceding text, it appears that tests for ANA or LE cells
In one study of 82 patients, including 8 with known on the pleural fluid add very little to the information
SLE, the pleural fluid ANA levels were less useful than obtained from these tests on the serum. Accordingly,
had been reported previously. The pleural fluid ANA they are not recommended.
titers were increased to 1 :320 or above in six patients
Biopsy
with lupus pleuritis and were below 1 : 1 60 in two
patients with SLE and effusions due to other factors Pleural biopsy is useful in establishing the diagnosis
(43) . In the six patients with lupus pleuritis, the ANA of lupus pleuritis if immunofluorescence is combined
titers in the pleural fluid and in the serum tended to with light-microscopic examination. Chandrasekhar
C HAPTER 2 1 I PLE U RAL D I S EAS E D U E TO COLLAG E N VASC U LAR D I S EASES 335
had Sjogren's syndrome without other connective tis blood eosinophilia. The systemic vasculitis with the
sue disease (5 8 ). The pleural fluid has been described Churg-Strauss syndrome resembles that of periarteri
in two patients and was a lymphocyte-predominant tis nodosa, but severe renal disease is uncommon. The
exudate with normal pH and glucose level and a low classic histologic picture consists of a necrotizing vas
adenosine deaminase level (5 9,60) . culitis, eosinophilic tissue infiltration, and extravascu
lar granulomas, bur all three components are found in
a minority of cases. In recent years, it has been sug
Fam i l i a l Med iterra n e a n Feve r
gested that there is an association between the admin
Familial Mediterranean fever, also known as familial istration of leukotriene receptor antagonists and the
paroxysmal polyserositis, is a rare cause of paroxysmal development of the Churg-Strauss syndrome (70) .
attacks of fever and pleuritic chest pain, sometimes Pleural involvement occurs on occasion with the
with pleural effusion (6 1 ,62) . The hallmark of the Churg-Strauss syndrome. In Lanham et al.'s (68) review
disease is the recurrent, acute, self-limited febrile of the literature in 1 984, 1 8 of 6 1 patients (30%) in
episodes of peritonitis, pleuritis, synovitis, or an whom chest radiograph results were reported had a pleu
erysipelas-like syndrome. Familial Mediterranean ral effusion. However, in a recent review of 96 patients,
fever is an autosomal recessive disease that occurs pleural effusions were said to be rare (69) . High
almost exclusively in Armenians and Sephardic Jews resolution CT scans demonstrate ground glass opaci
who have their origin in the Mediterranean countries. ties and consolidations in the majority of patients (7 1 ) .
The initial attack usually occurs before age 20 and Th e pleural fluid findings with the Churg-Strauss
is typically dominated by peritoneal symptoms and syndrome are relatively unique. Erzurum et al. (72)
signs. The initial attack is characterized by pleuritic reported on one patient with bilateral effusions and
chest pain and fever in fewer than 1 0% of patients, but pleural fluid with an LDH of 2,856 IU/L, a pH of
approximately 40% have an attack of febrile pleurisy 7.08, a glucose level less than 10 mg/dL, and 1 0,400
during the course of their disease (6 1 ) . Chest radio WBC with 95% eosinophils. The only other disease
graphs during the acute pleuritic attacks reveal eleva with comparable pleural findings is paragonimiasis.
tion of the ipsilateral diaphragm and frequently small Patients with Churg-Strauss syndrome respond
pleural effusions (62) . The pleural fluid can contain well to treatment with steroids, although some
predominantly polymorphonuclear leukocytes (63) or patients benefit from the addition of immunosup
lymphocytes (64) . The radiographic abnormalities pressive agents. The vasculitic illness is usually of lim
and the symptoms are usually completely gone within ited duration, but relapses can occur, and they should
48 hours. Approximately 40% of the patients also be detected and treated early (68) .
have amyloidosis (65) . The attacks are recurrent,
with irregular intervals of days to months between
Weg e n e r's G ra n u l o m atos i s
the attacks. Because the administration of colchicine,
0 . 5 mg orally twice daily, decreases the frequency Wegener granulomatosis, characterized by necrotizing
of the attacks (66,67) , it is worthwhile to establish granulomatous vasculitis of the small vessels, typically
this diagnosis in patients with recurrent episodes of involves the upper and lower respiratory tracts and
polyserositis. produces glomerulonephritis (20) . Radiologically,
the most common patterns in the lung are solitary
or multiple nodular densities, either poorly defined
C h u rg-Stra uss Syn d ro m e
or sharply circumscribed (67) . An associated small
Churg-Strauss syndrome is a disorder characterized pleural effusion is frequently seen (73,74) .
by hypereosinophilia and systemic vasculitis occur In one series of 1 1 patients, 6 (55%) had small
ring in individuals with asthma and allergic rhinitis pleural effusions (74), whereas in another series of
(68) . The American College of Rheumatology has 1 8 patients, 4 (22%) had pleural effusions (73 ) . The
proposed six criteria for the Churg-Strauss syndrome, pleural fluid in patients with Wegener's granulomato
with four being necessary for the diagnosis with an sis has not been well characterized, but it is probably
8 5 % sensitivity and 99 .7% specificity; the six crite an exudate. Because effective treatment for this disease
ria are asthma, eosinophilia greater than 1 0%, para is now available (20) , it is important to consider this
nasal sinusitis, pulmonary infiltrate, histologic proof diagnosis in patients with parenchymal infiltrates and
of vasculitis, and mononeuritis multiplex (69). Typi a pleural effusion. Measurement of serum antineu
cally, this disease begins with allergic rhinitis, with the trophil cytoplasmic antibodies (ANCA) is useful in
subsequent development of asthma and peripheral the diagnosis of Wegener's granulomatosis. There are
C HAPTER 2 1 I PLE U RAL D I S EAS E D U E TO COLLAG E N VASC U LAR D I S EASES 337
two major patterns of ANCA immunofluorescence 3 . Faurschou P, Francis D, Faarup P. Thoracoscopic, histological,
and clinical findings in nine case of rheumatoid pleural effu
c-ANCA shows granular staining in the cytoplasm
sion. Thorax. 1 98 5 ;40:37 1 -375.
that is accentuated in the cleft between the neutro 4. Feagler JR, Sorensen GD, Rosenfeld MG, et al. Rheumatoid
phil nuclear lobes and p-ANCA shows perinuclear pleural effusion. Arch Pathol. 1 97 1 ;92:257-266.
accentuation around the periphery of the neutrophil 5. Ferguson GC. Cholesterol pleural effusion in rheumatoid
nucleus. Most patients with Wegener's granulomato lung disease. Thorax. 1 966;2 1 : 577-5 82.
6. Lillington GA, Carr DT, Mayne JG. Rheumatoid pleurisy
sis are c-ANCA positive (75 ) .
with effusion. Arch Intern Med. 1 97 1 ; 1 28 :764-768 .
7. Halla JT, Schronhenloher RE, Volanakis J E . Immune com
plexes and other laboratory features of pleural effusions. Ann
Eosi n o p h i l i a-Mya l g i a Syn d ro m e
Intern Med. 1 980;92:748-752.
In the late 1 980s, an epidemic of the eosinophilia 8 . Naylor B . The pathognomonic cytologic picture of rheuma
toid pleuritis. Acta Cytol. I 990;34:465-473.
myalgia syndrome was linked to the dietary ingestion
9 . Pritikin JD, Jensen WA, Yenokida GG, et al. Respiratory fail
of contaminated 1-tryptophan. The clinical manifes ure due to a massive rheumatoid pleural effusion. J Rheumatol.
tations of the eosinophilia-myalgia syndrome include 1 990; 1 7:673-67 5 .
myalgias, arthralgias, skin rashes, muscle pain, edema, 1 0 . Bagga S . Rheumatoid lung disease a s seen on PET/CT scan.
fatigue, neuropathy, and marked peripheral eosino Clin Nucl Med 2007; 32:753-754.
1 1 . Delbarre F, Kahan A, Amor B, et al. La ragocyte synovial: son
philia (76) . A similar syndrome appears sporadically
interet pour le diagnosic des maladies rheumatismales. Presse
without exposure to contaminated 1-tryptophan (77) . Mid. 1 964;72 : 2 1 29-2 1 32 .
More than half of the patients with the eosinophilia 1 2 . Sahn SA. Immunologic diseases of the pleura. Clin Chest Med.
myalgia syndrome have respiratory complaints, with 1 9 85;6: 1 03- 1 1 2 .
dyspnea occurring most frequently. 1 3 . Faurschou P. Decreased glucose i n RA-cell-positive pleural
effusion: correlation of pleural glucose, lactic dehydrogenase
Pleural effusions can occur with the eosinophilia and protein concentration to the presence of RA-cells. Eur J
myalgia syndrome. In one large series of 1 , 53 1 patients, Respir Dis. 1 984;6 5 : 272-277.
7 1 8 patients had chest radiographs and pleural effu 14. Hindle W, Yates DAH . Pyopneumothorax complicating rheu
sions were present in 1 2% (78) . The pleural effusions matoid lung disease. Ann Rheum Dis. 1 965;24:57-60.
15. Jones FL, Blodgett RC. Empyema in rheumatoid pleuropul
are usually bilateral and are sterile eosinophilic exu
monary disease. Ann Intern Med. 1 97 1 ;74:665-67 1 .
dates (77,79) . Although some patients have improved 1 6 . Ball GV, Whitfield CL. Studies o n rheumaroid disease pleural
with the discontinuation ofl-tryptophan or corticoste fluid. Arthritis Rheum. 1 966;9: 846.
roid therapy, the response is often incomplete and the 17. Dodson WH , Hollingsworth JW. Pleural effusion in rheuma
disease may be chronic and progressive. toid arthritis. N Engl] Med. 1 966;275 : 1 337-1 342.
1 8 . Carr DT, McGuckin WE Pleural fluid glucose. Am Rev Respir
Dis. 1 968;97: 302-305.
M i sce l l a neous D iseases 1 9 . Russakoff AH, LeMaistre CA, Dewlett HJ. An evaluation of
the pleural fluid glucose determination. Am Rev Respir Dis.
Occasionally, patients with other collagen vascular dis 1 962;85 :220-223.
eases such as scleroderma (80), temporal arteritis (8 1 ) , 20. Hunninghake GW, Fauci AS. Pulmonary involvement in the
collagen vascular diseases. Am Rev Respir Dis. 1 979; 1 1 9 :471-503.
ankylosing spondylitis, polyarteritis nodosa, Behs;et's
2 1 . Mays EE. Rheumatoid pleuritis: observations in eight cases
syndrome (82) , Kawasaki's disease (83), eosinophilic and suggestions for making the diagnosis in patients without
pneumonia (84) , or dermatomyositis have a pleural the "typical findings." Dis Chest. I 968;53:202-2 1 4 .
effusion, but it appears that the pleural effusions in 2 2 . Russell M L , Gladman DD, Mintz S . Rheumatoid pleu
such patients usually result from complications of the ral effusion: lack of response to intrapleural corticosteroid.
J Rheumatol. 1 986; 1 3 : 4 1 2-4 1 5 .
disease, such as heart failure, pneumonia, or pulmonary
23. Abu-Shakra M, Nicol P, Urowitz M B . Accelerated nodulosis,
embolism, rather than from the primary disease. How pleural effusion, and pericardia! tamponade during metho
ever, there has been one case report of a patient with trexate therapy. ] Rheumatol. 1 994;2 1 :934-937.
Behs;et's syndrome who presented with life-threatening 24. Brunk JR, Drash EC, Swineford 0 . Rheumatoid pleuritis
chylothorax and chylopericardium (82) . successfully treated with decortication. Report of a case and
review of the literature. Am J Med Sci. 1 966;25 1 : 545-5 5 1 .
2 5 . Yarbrough JW, Sealy WC, Miller JA. Thoracic surgical prob
lems associated with rheumatoid arthritis. J Thorac Cardiovasc
REFERENCES
Surg. 1 975;68 :347-3 54.
26. Chapman PT, O'Donnell JL, Moller PW Rheumatoid pleural
I. Walker WC, Wright V. Rheumatoid pleuritis. Ann Rheum Dis. effusion: response to intrapleural corticosteroid. Rheumatology.
1 967;26:467-474. 1 992; 1 9 :478-480.
2. Horler AR, Thompson M . The pleural and pulmonary 27. Light RW, Jenkinson SG, Minh V, et al. Observations on pleu
complications of rheumatoid arthritis. Ann Intern Med. ral pressures as fluid is withdrawn during thoracentesis. Am
1 9 59; 5 I : 1 1 79- 1 203. Rev Respir Dis. 1 980; 1 2 1 :799-804.
33 8 PLE U RAL D I S EASES
2 8 . Uppal SS, Al-Mucairi M, Hayat S, et al. Ten years of clini 5 1 . Hunder GG, McDuffie FC, Hepper NGG. Pleural fluid
cal experience with adult onset Still's disease: is the outcome complement in systemic lupus erythematosus and rheumatoid
improving? Clin Rheumatol. 2006;25 : 1 055-1 060. arthritis. Ann Intern Med. 1 972;76:357-362.
29. Ferreiro L, Alvarez-Dobano JM, Valdes L. Systemic diseases 52. Breuer GS, Deeb M, Fisher D, et al. Therapeutic options
and the pleural. Arch Bronchopneumol. 20 l 1 ;47:36 1 -370. for refractory massive pleural effusion in systemic lupus ery
30. Harvey AM , Shulman LE, Tumulty PA, et al. Systemic lupus thematosus: a case study and review of the literature. Semin
erythemarosus: review of the literature and clinical analysis of Arthritis Rheum. 2005;34:744-749 .
1 3 8 cases. Medicine. 1 954;33:29 1 -437. 5 3 . Sharma S, Smith R , Al-Hameed F. Fibrothorax and severe
3 1 . Winslow WA, Ploss LN, Loitman B . Pleuritis in systemic lung restriction secondary to lupus pleuritis and its successful
lupus erythematosus: its importance as an early manifestation treatment by pleurectomy. Can Respir J 2002;9:33 5-337.
in diagnosis. Ann Intern Med. 1 95 8 ;49:70-8 8 . 54. Prakash UB. Respiratory complications in mixed connective
32. Alarcon-Segovia D, Alarcon DG. Pleuro-pulmonary manifesta tissue disease. Clin Chest Med. 1 99 8 ; 1 9 :733-746.
tions of systemic lupus etythematosus. Dis Chest. 1 9 6 1 ;39:7- 1 7. 55. Ferreiro L, Alvarez-Dobano JM, Valdes L. Systemic diseases
33. Man BL, Mok CC. Serositis related ro systemic lupus erythe and the pleural. Arch Bronchopneumol. 20 1 1 ;47:3 6 1 -370.
marosus: prevalence and outcome. Lupus. 2005; 1 4 : 822-826. 56. Strimlan CV, Rosenow EC, Divertie MG, et al. Pulmo
34. Blomgren SE, Condemi JJ, Vaughan JH. Procainamide nary manifestations of Sjiigren's syndrome. Chest. 1 976;
induced lupus erythemarosus. Am J Med. l 972; 52:338-348 . 70:354-36 1 .
3 5 . Purnell D C , Baggenstoss AH, Olsen AM . Pulmonary lesions 57. Lohrmann C , Uhl M , Warnatz K, et al. High-resolution
in disseminated lupus erythemarosus. Ann Intern Med. 1 95 5 ; CT imaging of the lung for patients with primary Sjogren's
42: 6 1 9-628 . syndrome. Eur] Radio/. 2004;52: 1 37-143.
36. Gueft B, Laufer A. Further cyrochemical studies in systemic 5 8 . Teshigawara K, Kakizaki S, Horiya M, et al. Primary Sjogren's
lupus erythematosus. Arch Pathol. 1 9 54;57:20 1 -226. syndrome complicated by bilateral pleural effusion. Respirology.
37. Good JT Jr, King TE, Antony VB, et al. Lupus pleuritis: 2008; 1 3 : 1 55-1 5 8 .
clinical features and pleural fluid characteristics with special 5 9 . Alvarez-Sala R , Sanchez-Tori! F, Garcia-Martinez J, e t al.
reference to pleural fluid antinuclear antibodies. Chest. 1 983; Primary Sjogren syndrome and pleural effusion. Chest. 1 9 89;
84:7 1 4-7 1 8 . 96: 1 440- 1 44 1 .
3 8 . Mitroo S, Gelber AC, Hitchon CA, e t al. Clinical and sero 60. Ogihara T, Nakatani A, Ito H, et al. Sjiigren's syndrome with
logic facrors associated with lupus pleuritis. ] Rheumatol. 20 1 O; pleural effusion. Intern Med. 1 99 5 ;34:8 1 1 -8 1 4 .
37:747-753. 6 1 . Sohar E, Gafni J, Pras M, e t al. Familial Mediterranean fever.
3 9 . Gould DM, Dayes ML. Roentgenologic findings in systemic Am ] Med. 1 967;43 :227-2 5 3 .
lupus erythematosus. ] Chronic Dis. 1 9 5 5 ;2: 1 36- 1 4 5 . 62. Ehrenfeld E N , Eliakim M, Rachmilewitz M . Recurrent
4 0 . Harpey J - P. Lupus-like syndromes induced b y drugs. Ann polyserositis (familial Mediterranean fever; periodic disease) .
Allergy. 1 974;33:256-26 1 . Am ] Med. 1 9 6 1 ;3 1 : 1 07- 1 23 .
4 1 . Toya S P, Tzelepis GE. Association of the shrinking lung 63. Merker H-J, Hersko C, Shibolet S . Serosal exudates in familial
syndrome in systemic lupus erythemarosus with pleurisy: a Mediterranean fever. Am ] Clin Pathol. 1 967;48 :23-29.
systematic review. Semin Arthritis Rheum. 2009;39:30-37. 64. Katsenos S, Mermigkis C, Psathakis K, et al. Unilateral lym
42. Smith PR, Nacht RI . Drug-induced lupus pleuritis mimicking phocytic pleuritis as a manifestation of familial Mediterranean
pleural space infection. Chest. 1 992; 1 0 1 :268-269 . fever. Chest. 2008 ; 1 33:999- 1 00 1 .
43. Khare V, Baethge B , Lang S, e t al. Antinuclear antibodies in 6 5 . Odabas AR, Cetinkaya R , Selcuk Y, e t al. Familial Mediter
pleural fluid. Chest. 1 994; 1 06:866-87 1 . ranean fever. South Med] 2002;9 5 : 1 400- 1 403.
44. Poree! JM, Ordi-Ros J , Esquerda A, et al. Antinuclear anti 66. Zemer D, Revach M, Pras M, et al. A controlled trial of col
body testing in pleural fluid for the diagnosis oflupus pleuritis. chicine in preventing attacks of familial Mediterranean fever.
Lupus. 2007; 1 6 :25-27. N Engl] Med. 1 974;29 1 :932-934.
45. Wang DY, Yang PC, Yu WL, et al. Serial antinuclear anti 67. Dinarello CA, Wolff SM, Goldfinger SE, et al. Colchicine
bodies titre in pleural and pericardia! fluid. Eur Respir J 2000; therapy for familial Mediterranean fever. A double-blind trial.
1 5 : 1 1 06-1 1 1 0 . N Engl] Med. 1 974;29 1 :934-937.
46. Wang DY, Yang P C , Yu WL, e t al. Comparison of different diag 68. Lanham JG, Elkon KB, Pusey CD, et al. Systemic vasculi
nostic methods for lupus pleuritis and pericarditis: a prospec tis with asthma and eosinophilia: a clinical approach to the
tive three-year study. ] Formos Med Assoc. 2000;99:375-380. Churg-Strauss syndrome. Medicine. l 984;63 :65-8 1 .
47. Chao TY, Huang SH, Chu CC. Lupus erythematosus cells in 69. Guillevin L, Cohen P, Gayraud M , et al. Churg-Strauss syn
pleural effusions: diagnostic of systemic lupus erythematosus? drome. Clinical study and long-term follow-up of 96 patients.
Acta Cytol. 1 997;4 1 : 1 23 1 - 1 233. Medicine {Baltimore}. 1 999;78:26-37.
4 8 . Chandrasekhar AJ, Robinson J, Barr L. Antibody deposition 70. Wechsler ME, Garpestad E, Flier SR, et al. Pulmonary infil
in the pleura: a finding in drug-induced lupus. J Allergy Clin trates, eosinophilia, and cardiomyopathy following corticoste
Immunol. 1 978;6 1 :399-402. roid withdrawal in patients with asthma receiving zafirlukast.
49. Pertschuk LP, Moccia LF, Rosen Y, et al. Acute pulmonary JAMA. 1 998 ;279:45 5-457.
complications in systemic lupus erythemarosus. Immuno 7 1 . Szczeklik W, Sokolowska B, Mastalerz L, et al. Pulmonary
fluorescence and light microscopic study. Am J Clin Pathol. findings in Churg-Strauss syndrome in chest X-rays and high
1 977;6 8 : 5 5 3-5 57. resolution computed tomography at the time of initial diagno
5 0 . Andrews BS, Arora NS, Shadforth MF, et al. The role of sis. Clin Rheumatol. 20 1 0;29: 1 1 27- 1 1 34 .
immune complexes in the pathogenesis of pleural effusions. 7 2 . Erzurum S E , Underwood GA, Hamilos D L , et a l . Pleural effu
Am Rev Respir Dis. 1 9 8 1 ; 1 24: 1 1 5- 1 20. sion in Churg-Strauss syndrome. Chest. 1 9 89;9 5 : 1 357- 1 3 5 9 .
C HAPTER 2 1 I PLE U RAL D I S EAS E D U E TO COLLAG E N VASC U LAR D I S EASES 339
7 3 . Fauci AS, Wolff S M . Wegener's granulomacosis: studies in 8 0 . Thompson AE, Pope J E . A study o f the frequency o f peri
eighteen patients and a review of the literature. Medicine. cardia! and pleural effusions in scleroderma. Br J Rheumatol.
1 973;52: 535-5 6 1 . 1 998;37: 1 320- 1 323.
7 4. Gonzales L, Van Ordstrand HS. Wegener's granulomatosis: 8 1 . Marie I, Heliot P, Muir JF, et al. Pleural effusion revealing
review of 1 1 cases. Radiology. 1 973 ; 1 08 :295-300. giant cell arteritis. Eur] Intern Med. 2004; 1 5 : 1 25-1 27.
75. Homer RJ. Antineutrophil cytoplasmic antibodies as markers for 82. Coplu L, Emri S, Selcuk ZT, et al. Life threatening chylous
systemic autoimmune disease. Clin ChestMed. 1 998; 1 9:627-639. pleural and pericardia! effusion in a patient with Behs:et's
76. Martin RW, Duffy J, Engel AG, et al. The clinical spectrum of syndrome. Ihorax. 1 992;47:64-6 5 .
the eosinophilia-myalgia syndrome associated with 1-tryptophan 83. Hamada H, Terai M, Honda T, e t al. Marked pleural and
ingestion. Ann Intern Med. 1 990; 1 1 3 : 1 24- 1 34. pericardia! effusion with elevated vascular endothelial growth
77. Killen Jw. Swift GL, White RJ. Eosinophilic fasciitis with pulmo factor production: an uncommon complication of Kawasaki
nary and pleural involvement. Postgrad MedJ 2000;76:36-37. disease. Pediatr Int. 2005 ;47: 1 1 2-1 1 4 .
78. Swygert LA, Maes EF, Sewell LE, et al. Eosinophilia 8 4 . Samman YS, Wali SO, Abdelaal MA, e t al. Chronic eosino
myalgia syndrome. Results of national surveillance. JAMA. philic pneumonia presenting with recurrent massive bilateral
1 990;264: 1 698-1 703. pleural effusion: case report. Chest. 200 1 ; 1 1 9 : 968-970.
79. Strumpf IJ, Drucker RD, Ander KH, et al. Acute eosinophilic
pulmonary disease associated with the ingestion of l-tryptophan
containing products. Chest. 1 9 9 1 ;99:8- 1 3 .
P l e u ra l E ff u s i o n D u e to D r u g R e a ct i o n s
Adverse reactions to drugs produce only a small per 1 month of initiating therapy with the drug. The
centage of all pleural effusions. Because the pleural symptoms with the acute presentation include dys
disease in most cases rapidly resolves when the drug pnea, nonproductive cough, and fever. The chest
is discontinued; however, it is important to consider radiograph is usually abnormal. In one series of 3 3 5
the possibility of drug-induced pleural disease in all patients, 1 86 (56%) had infiltrates, 65 ( 1 9%) had
patients with pleural effusions. The lupus-like syn infiltrates and effusion, 14 (4%) had only an effusion,
dromes associated with various drugs are described and 70 (2 1 %) had a normal chest radiograph (5) .
in Chapter 2 1 . In this chapter, the pleural diseases Most patients with acute pleuropulmonary reac
resulting from the administration of nitrofurantoin, tions to nitrofurantoin have both peripheral eosino
dantrolene, ergot alkaloids, dasatinib, amiodarone, philia ( > 3 5 0/mm3) and lymphopenia (< l ,OOO/mm3)
interleukin 2 (IL-2), procarbazine, methotrexate, (6) . The only reported pleural fluid analysis showed
interleukin 1 1 (IL- 1 1 ) , all-trans-retinoic acid, val 1 7% eosinophils (6) .
proic acid and clozapine are discussed. These are the The chronic syndrome occurs when the patient
only drugs convincingly incriminated in the produc has been taking nitrofurantoin from 2 months to
tion of pleural disease other than drugs that produce 5 years and is much less frequent than the acute syn
the lupus-like syndrome. At the end of the chapter, drome. The presentation is insidious, with the grad
there is a summary of other drugs that have been ual onset of dyspnea on exertion and a nonproductive
implicated but not proved to cause pleural effusions. cough (4) . Patients with the chronic syndrome always
have abnormal chest radiographs; diffuse bibasilar
N ITRO F U RANTO I N infiltrates are the most common abnormality (5) .
Pleural effusions, which are less common with the
Nitrofurantoin (Furadantin) is widely used in the treat
chronic form, occur in fewer than 1 0% of patients.
ment of urinary tract infections. Israel and Diamond
No patients with the chronic syndrome have had a
( 1 ) first reported that the administration of nitrofu
pleural effusion without an infiltrate (4) .
rantoin could be associated with the development of
The diagnosis of nitrofurantoin pleuropulmonary
an acute febrile illness with pulmonary infiltrates and
reaction should be suspected in all patients with a
pleural effusion. A subsequent review of the literature
pleural effusion who are taking nitrofurantoin. If the
and the records of the company that produces nitrofu
drug is discontinued, the patient with the acute syn
rantoin in 1 969 revealed that approximately 200 cases
drome usually improves clinically within 1 to 4 days,
of this syndrome had been reported (2) . Now, there
and the chest radiograph becomes normal within a
have been more than 2,000 cases reported (3) . It is
week (5). Symptoms and signs with the chronic syn
thought that nitrofurantoin injures the lung through
drome resolve much more slowly (5).
the production of oxygen radicals (4) .
Pulmonary reactions to nitrofurantoin may
DANTRO L E N E
develop in two distinct patterns characterized by the
length of treatment before the development of the Dantrolene sodium (Dantrium) is a long-acting skel
syndrome (4) . The acute presentation occurs within etal muscle relaxant used in treating patients with
340
C H A PT E R 22 I P L E U RAL E F F U S I O N D U E TO D R U G REACT I O N S 34 1
spastic neurologic disorders. The chemical structure It is not clear what these authors meant by the term
of dantrolene is similar to that of nitrofurantoin (7, 8 ) . pleurisy, but apparently all the patients had pleural effu
The chronic administration of dantrolene can lead to sions or pleural thickening ( 1 1 ) . The pleurisy developed
an eosinophilic pleural effusion (7, 8 ) . At one of my 1 month to 3 years after methysergide therapy was ini
previous institutions, the Veterans Affairs Medical tiated, and in five patients, it was bilateral. Only 1 of
Center in Long Beach, we saw more than 1 0 instances the 1 3 patients had concomitant retroperitoneal fibro
of pleural effusions due to dantrolene during a 1 0-year sis. Although no description of the pleural fluid was
period. This institution had a large population of included in this report, in another report of a patient
patients with spinal cord injury for whom dantrolene with bilateral pleural effusions, the fluid was bloody
was frequently prescribed. In one report, four patients on one side and clear on the other ( 1 2) . When methy
developed an eosinophilic pleural effusion 2 months sergide was discontinued, the patients' symptoms
to 3 years after the initial administration of dantrolene and signs improved. At follow-up 6 months or more
(7) . In one case, the eosinophilic effusion developed after discontinuation of the drug, pleural fibrosis was
12 years after dantrolene therapy was initiated (8) . not detectable or was slight in seven patients, moder
The pleural effusions in all patients were unilateral, ate in three patients, and severe in two patients. The
and no associated pulmonary infiltrates were seen. In two patients with severe fibrosis were those who had
the reported series of four patients, one had a peri continued to take methysergide for the longest period
cardia! friction rub and another had a pericardia! rub ( 1 8 and 36 months) after the onset of this pleurisy
with a pericardia! effusion (7) . Two of the patients ( 1 1 ) . Therefore, the occurrence of a pleural effusion
were febrile, and two had pleuritic chest pain. or pleural thickening in a patient taking methysergide
All reported patients have had at least 5% eosino is a strong indication for the prompt discontinuation
phils in their peripheral blood (7) . The pleural fluid of the drug.
is an exudate with normal glucose and amylase levels. Rinne ( 1 3) reviewed the chest radiographs of 1 23
The differential white blood cell count on the pleu patients taking bromocriptine for Parkinson's disease
ral fluid has revealed at least 3 5 % eosinophils in all and found that 7 patients (6%) had pleural effusions,
cases. When dantrolene is discontinued, the patients pleural thickening, and pulmonary infiltrates. As of
improve symptomatically within days, but it takes 1 9 8 8 , there had been a total of 23 patients reported
several months for the pleural effusions to resolve who developed pleuropulmonary disease while taking
completely. Administration of corticosteroids may bromocriptine ( 1 4) . All the patients were men, and
accelerate the resolution of the eosinophilic effusion most of them have had a history oflong-term cigarette
(9) . The mechanism by which dantrolene produces smoking. The prevalence of symptomatic pleuropul
the eosinophilic pleural effusion is unknown. monary disease among individuals taking bromocrip
tine is 2% to 5% ( 1 4) . Patients had taken the drug for
6 months to 4 years before symptoms developed. The
E RG O T A L KA LO I D S
chest radiograph reveals unilateral or bilateral pleural
Ergot alkaloid drugs such as bromocriptine, ergota thickening or effusion with or without pulmonary
mine, dihydroergotamine, nicergoline, pergolide, and infiltrates. An occasional patient has only pulmonary
dopergine are sometimes used in the long-term treat infiltrates. Analysis of the pleural fluid reveals an exu
ment of Parkinson's disease. Ergot derivatives have date with predominantly lymphocytes and frequently
also been used in the treatment and prophylaxis of eosinophils ( 1 4, 1 9) . The erythrocyte sedimentation
migraine and cluster headaches (methysergide, ergot rate (ESR) is sometimes markedly elevated in these
amine) . The long-term administration of any of these patients ( 1 8) . There is some suggestion that patients
drugs can lead to pleuropulmonary changes ( 1 0- 1 8) . taking bromocriptine-type drugs are more likely to
It has been suggested that the pleural changes are develop pleural disease if they have a history of expo
due to increased serotonin levels with a subsequent sure to asbestos ( 1 8,20) .
increase in fibroblast activity ( 1 9) . As of 2002, there had been a total of 87 patients
Methysergide (Sansert) is a serotonin antagonist reported who developed one or more symptoms of
used to treat migraine headaches. The association of serosal fibrosis after taking pergolide, including 3 8
methysergide administration with the development cases with pleural effusion and 2 1 cases with pleural
of retroperitoneal fibrosis and fibrosing mediastinitis fibrosis (2 1 ) . The mean age of the patients was 64 years,
is well established ( 1 0) . One report described 1 3 cases and approximately three fourths of the patients were
of "pleurisy'' secondary to methysergide treatment ( 1 1 ) . men (2 1 ) . Most patients had been taking the drug for
342 PLE U RAL D I S EASES
more than 1 year and had not recovered at the time transudative. Seven of the nine effusions were lym
that they were reported (2 1 ) . phocyte predominant and one was a chylothorax (26) .
Th e natural history o f pleuropulmonary disease Pleural biopsy has shown lymphocytic infiltration (25 ) .
during treatment with ergot alkaloids is unclear. The mechanisms responsible for the development
The disease progresses only in some of the patients of the pleural effusion is unknown. It is thought that
who continue taking the drug ( 1 4) . On discontinu it is immune mediated given the predominance of
ation, most patients improve, but complete resolu lymphocytes in the pleural effusion, the high inci
tion of the process is rare. It is recommended that dence of patients who have lymphocytosis, and the
annual chest radiographs be obtained in patients who lymphocytic infiltrates on the pleural biopsy.
are taking ergot alkaloids on a long-term basis. If the The optimal management of a patient with a
radiograph reveals pleural or parenchymal infiltrates, dasatinib-related pleural effusion has yet to be defined.
strong consideration should be given to stopping the If dasatinib is discontinued, the effusion usually
ergot alkaloids and using an alternative drug for the resolves (22) . If the patient is given corticosteroids,
treatment of the Parkinson's disease. it will resolve more rapidly. Dasatinib can be rein
stituted after the effusion resolves. Since effusion are
least common with the 1 00 mg/day dose, this is rec
DASATI N I B
ommended when dasatinib is reinstituted. An alterna
Dasatinib is a tyrosine-kinase inhibitor approved for tive approach after the patient has been off dasatinib
the treatment of BCR-ABL positive chronic myeloid is to institute therapy with nilotinib, which is another
leukemia (CML) and Philadelphia chromosome pos tyrosine kinase inhibitor (27) .
itive acute lymphoblastic leukemia (ALL) after ima
tinib failure. The incidence of dasatinib-associated
A M I O DA RO N E
pleural effusions is approximately 20% to 40% (22) ,
which is probably the highest incidence of pleural Amiodarone is an antiarrhythmic drug that may pro
effusion associated with any drug. Administration of duce severe and potentially lethal pulmonary toxic
dasatinib once daily is associated with fewer pleural ity. The current incidence of pulmonary toxicity in
effusions than is administration twice daily and is patients receiving amiodarone is 5% to 1 0% (28),
equally effective (23) . In one study (23), 3 1 9 patients and 5 % to 1 0% of those with pulmonary toxicity
were randomized to receive dasatinib 1 40 mg qd or die of pulmonary fibrosis. The pulmonary toxicity is
70 mg bid. The incidence of pleural effusion was 39% characterized by the insidious onset of nonproductive
in the bid group and 20% in the qd group (23) . In a cough, dyspnea, weight loss, and, occasionally, fever.
subsequent study (24) , 662 patients were randomized The chest radiograph reveals parenchymal infiltrates,
to dasatinib 1 00 mg qd, 50 mg bid, 1 40 mg qd, or which are predominantly interstitial (28) . The toxicity
70 mg bid. In this series (24) , the incidence of pleural rarely begins before 2 months of therapy, and it rarely
effusions was lower in the 1 00 mg qd group ( 1 4%) occurs in patients receiving less than 400 mg/day.
than in the 1 40 mg qd group (26%), the 70 mg bid Pleural effusions occur as a manifestation of amio
group (25%), or the 5 0 mg bid group (23%). The darone toxicity, but they are uncommon (29-3 1 ) .
mean time to the development of the pleural effusion Gonzalez-Rothi e t al. (29) reviewed 1 1 cases o f pleural
was about 200 days, and it tended to be longer in disease attributed to amiodarone and found that all 1 1
the groups that received 1 00 mg per day (24) . Pleural had concomitant parenchymal involvement. Subse
effusion was cited as the reason for discontinuing the quently, a case has been reported in which there was no
drug in less than 5% of patients. Peripheral lympho parenchymal involvement (30) . The pleural fluid is an
cytosis was more common in the patients that devel exudate (29-3 1 ) and may have predominantly lympho
oped pleural effusion than in those that did not (24) . cytes (3 1 ) , macrophages (30) , or polymorphonuclear
Most of the pleural effusions were symptomatic (24) . leukocytes (29) . Pleural fluid eosinophilia has not been
The characteristics of the pleural fluid secondary reported with amiodarone toxicity. The pleural abnor
to dasatinib therapy have not been well described. malities resolve when the amiodarone is discontinued.
Bergeron et al. (25) performed thoracentesis on six
patients and reported that the fluid was exudative in
I NT E R L E U KI N -2
all and five were lymphocytic and one was a chylotho
rax. Quintas-Cardama et al. (26) reported that seven Recombinant IL-2 is sometimes used in the treat
of nine pleural fluids were exudative while two were ment of malignancy, most commonly melanoma
C H A PT E R 22 I P L E U RAL E F F U S I O N D U E TO D R U G REACT I O N S 343
or renal cell carcinoma. The administration of IL-2 pleural effusion or a worsening of an existing pleural
is accompanied by multiple acute but generally effusion, but none required thoracentesis (3 5 ) . In a
reversible toxic effects, including fever, chills, leth subsequent study in which diuretics were given, no
argy, diarrhea, anemia, thrombocytopenia, eosino patient developed a pleural effusion (34) .
philia, confusion, and diffuse erythroderma, among
others (32) .
A L L-TRA N S - R ETI N O I C AC I D
One of the primary side effects of IL-2 adminis
tration is the development of pulmonary infiltrates This drug is used in the treatment of acute promyelo
and pleural effusion (32,33) . Vogelzang et al. (32) cytic leukemia. Up to one fourth of patients treated
reviewed the chest radiographs of 54 patients who with this drug develop a life-threatening reaction
were receiving high-dose IL-2 with or without lym that resembles the capillary leak syndrome which is
phokine-activated killer cell therapy for advanced attributed to IL-2 (36,37) . Symptoms develop 1 to
cancer and reported that 28 (52%) had a pleu 22 days after the initiation of treatment and include
ral effusion (32) . Other abnormalities on the chest fever, fluid retention, dyspnea, multiple organ failure,
radiograph included pulmonary edema in 41 % and and hemorrhagic and thrombotic manifestations.
focal infiltrates in 22%. The abnormalities were more In one series, 1 1 of 1 5 patients with the syndrome
frequent in patients receiving bolus rather than con had pleural effusion 7 of which were right sided and
stant intravenous therapy (32) . These pulmonary 2 each bilateral and left sided (36) . Many patients
reactions were clinically significant in that 1 9 patients also have lung parenchymal ground glass opacities,
(35%) either developed dyspnea at rest or required consolidation, or nodules (36) . The characteristics of
intubation. The pleural effusions tend to resolve, the fluid have not been reported (38) .
but they persisted in 1 7% of patients 4 weeks fol The seriousness of this syndrome is emphasized by
lowing therapy. In a second study, 26 of 54 patients the observation that 9 of 1 4 8 patients treated with
(48%) developed a pleural effusion after IL-2 therapy all-trans-retinoic acid died. Early administration of
(33) . In this series, 80% of the patients had either high-dose corticosteroids usually leads to prompt
alveolar edema or interstitial edema. Two of the symptomatic improvement, although full radio
patients without parenchymal infiltrates had a pleural graphic recovery may require several days and resid
effusion (33) . ual pleural thickening may remain in the occasional
The pathogenesis of the pleural effusion with IL-2 patient (36,37) .
therapy is probably related to the generalized capil
lary leak syndrome that sometimes occurs afrer IL-2
P R O CARBAZ I N E
therapy. It is likely that the pleural fluid originates
from the leaky capillaries in the lung. Therefore, the Procarbazine hydrochloride (Matulane) , a methyl
pleural fluid would be expected to be an exudate, but hydrazine derivative, is effective in the treatment of
to my knowledge, there is no published description Hodgkin disease and other lymphomas. Two detailed
of the pleural fluid characteristics. It is unclear as to case reports have described pleuropulmonary reac
why the pleural effusion persists so much longer than tions consisting of chills, cough, dyspnea, and bilat
does the pulmonary edema. eral pulmonary infiltrates with pleural effusions
occurring after treatment with procarbazine (39,40) .
In both instances, rechallenge with procarbazine
I NT E R LE U KI N - 1 1
again produced the infiltrates and pleural effusions.
Recombinant human IL- 1 1 is used to prevent severe Both patients had peripheral eosinophilia. When
thrombocytopenia in patients receiving myelosup the drug was discontinued, the patients' symptoms
pressive chemotherapy in patients with nonmyeloid and radiologic changes resolved within several days
malignancies. The administration of IL- 1 1 to normal (39,40) . This syndrome appears to be identical to that
individuals is associated with an increase in the associated with nitrofurantoin.
plasma volume of approximately 20% due to sodium
retention (34) . This increase in plasma volume is
M ETH OTR EXATE
thought to be responsible for most side effects from
IL- 1 1 , which include dyspnea, edema, and pleural One report exists of pleural effusion occurring after
effusions (34) . In one study of patients with stage methotrexate therapy for trophoblastic tumors
IV breast cancer, 7 of 40 patients ( 1 8%) developed a (4 1 ) . Walden et al. (4 1 ) treated 3 1 7 patients with
344 PLE U RAL D I S EASES
developed bilateral exudative pleural effusions Minoxidil. Siddiqui et al. (62) reported a case of
on the second day of her seventh or eighth a patient who developed a unilateral exuda
course. The effusion resolved within 2 weeks, tive pleural effusion with a pleural fluid pro
but recurred when the intravenous immuno tein level of 7 . 5 g/dL while taking minoxidil.
globulin was readministered. However, when After a therapeutic thoracentesis, the effusion
a different preparation of immunoglobulin recurred. The effusion gradually resolved after
was given, there was no recurrence of the pleu minoxidil was discontinued (62) . The patient
ral effusion ( 5 5 ) . This observation suggests had no renal disease. Minoxidil had previ
that the pleural effusion represents a reaction ously been implicated in pleural effusions in
to a component of the first preparation and patients with renal disease, but it was unclear
not of the immunoglobulin itself (38) . whether the effusions were due to uremic
Interferons. There is one case report of a 54-year-old pleuritis or fluid overload (62) .
man who developed a moderate right pleu Mitomycin. Small pleural effusions are frequently
ral effusion 9 days after starting therapy with present in patients who have interstitial infil
recombinant interferon-a-2a for chronic hep trates secondary to mitomycin (63) .
atitis (56). Although there was a slight increase Propylthiouracil. Middleton et al. ( 64) have reported
in serum antinuclear antibody titer, the one patient who developed left pleuritic chest
patient's symptoms and signs did not satisfy pain and an eosinophilic pleural effusion 3 weeks
the criteria of drug-induced lupus. The effusion after starting propylthiouracil (PTU) . A thora
gradually resolved upon discontinuation of the centesis 5 weeks after starting therapy revealed
interferon. 1 6% eosinophils. The patient continued tak
Isotretinoin. There is one report of a 49-year-old ing the PTU for 2 more weeks, and the effu
woman who developed an eosinophilic pleu sion enlarged and the eosinophils increased to
ral effusion 7 months after starting isotreti 45%. The effusion then resolved after PTU was
noin therapy for systemic sclerosis (38) . When discontinued (64) . Sen et al. (65) reported a
isotretinoin was discontinued, the chest radio patient who developed an eosinophilic pleural
graph became normal within 3 months. There effusion 1 1 years after starting PTU which went
is another case report of a similar case (57), and away with steroid therapy and discontinuation
the manufacturer of isotretinoin has on file of the drug. The effusion recurred when PTU
three other cases of pleural effusion occurring therapy was reinstituted (65).
in patients who took isotretinoin for acne. Simvastatin. Simvastatin is a member of the statin
Mesa/amine. Trisolini et al. (58) reported a family of drugs used to treat hypercholesterol
36-year-old patient who developed fever and emia. There is one case report of a patient who
bilateral areas of parenchymal infiltration with developed a moderate-sized right-sided pleu
eosinophilic pleural effusions several weeks ral effusion and interstitial infiltrates after tak
after starting mesalamine therapy. The fever ing simvastatin for 6 months. Thoracoscopy
and effusions which had persisted for weeks, revealed no pleural abnormalities. He had
disappeared shortly after the mesalamine was concomitant marked elevations of his liver
stopped (58) . A similar case has been reported function test results. The patient improved
by Sesin (59). There was also one case report in after the simvastatin was stopped and predni
which a patient was thought to have developed sone was administered (66) . There is another
pleuropericarditis 7 months after starting this case report ( 67) of a patient who developed an
treatment (60) . In this instance, the patient was eosinophilic pleural effusion after taking sim
felt to have drug-induced lupus (60) . vastatin for 1 3 years. The effusion went away
Metronidazole. Kristenson and Fryden (6 1 ) when the simvastatin was discontinued.
reported an interesting case o f a patient who Tizanidine. There is one case report of a patient
developed fever and pleural effusions on two who developed a large pleural effusion 6 weeks
different occasions within a day of starting after starting tizanidine, a muscle relaxant
a course of oral metronidazole. On the ini (68) . The effusion was an exudate with 1 0%
tial occasion, pulmonary infiltrates were also eosinophils. The effusion gradually disap
present. peared after tizanidine was discontinued (68) .
346 PLE U RAL D I S EASES
34. Smith JW II. Tolerability and side-effect profile of rhIL- 1 1 . 5 1 . Kalogeropoulos AS , Tsiodras S, Loverdos D, e t al. Eosino
Oncology (Williston Park). 2000 ; 1 4 (suppl 8):4 1 -47. philic pneumonia associated with daptomycin: a case report
35. Isaacs C, Robert NJ, Bailey FA, et al. Randomized placebo and a review of the literature. J Med Case Reports. 2 0 1 1 ; 5 : 1 3 .
controlled study of recombinant human interleukin- ! I to 5 2 . Behnia M, Dowdeswell I , Vakili S . Pleural fluid and serum
prevent chemotherapy-induced thrombocytopenia in patients eosinophilia: association with fluoxetine hydrochloride. South
with breast cancer receiving dose-intensive cyclophosphamide Med] 2000;93 : 6 1 1 -6 1 3 .
and doxorubicin. ] Clin Oncol. 1 997; 1 5 : 3368-3377. 5 3 . Tzanakis N, Bouros D, Siafakas N. Eosinophilic pleural effu
36. Jung JI, Choi JE, Hahn ST, et al. Radiologic features of sion due to gliclazide. Respir Med. 2000;94:94.
all-trans-retinoic acid syndrome. A]R Am J Roentgenol. 54. Goldsby R, Pulsipher M, Adams R, et al. Unexpected pleural
2002; 1 7 8 :475-480. effusions in 3 pediatric patients treated with STI-57 1 . ] Pediatr
37. Frankel SR, Eardley A, Lauwers G, et al. The "retinoic acid Hematol Oncol. 2002;24:694-6 9 5 .
syndrome" in acute promyelocytic leukemia. Ann Intern Med. 5 5 . Bolanos-Meade J, Keng YK, Cobos E. Recurrent lymphocytic
1 992; 1 1 7 :292-296. pleural effusion after intravenous imm unoglobulin. Am ]
3 8 . Kalomenidis I . Effusions due to drugs. In: Light RW, Lee Hematol. 1 999;60:248-249 .
YCG, eds. Textbook of Pleural Diseases. London, England: 56. Takeda A , Ikegame K , Kimura Y, e t al. Pleural effusion during
Arnold Publishers, 2003 :382-393. interferon treatment for chronic hepatitis C. Hepatogastroen
39. Jones SE, Moore M , Blank N, et al. Hypersensitivity to pro terology. 2000;47: 1 43 1 - 1 43 5 .
carbazine (Matulane) manifested by fever and pleuropulmo 5 7 . Milleron BJ, Valcke J, Akoun G M , e t al. Isotretinoin-related
nary reaction. Cancer. 1 9 72;29:498-500. eosinophilic pleural effusion. Chest. 1 996; 1 1 0: 1 1 2 8 .
40. Ecker MD, Jay B, Keohane MF. Procarbazine lung. A]R Am 5 8 . Trisolini R , Dore R, Biagi F, et a l . Eosinophilic pleural effusion
] Roentgenol. 1 978; 1 3 1 : 527-52 8 . due to mesalamine. Report of a rare occurrence. Sarcoidosis
4 1 . Walden PAM, Mitchell-Heggs PF, Coppin C, e t a l . Pleurisy Vase Diffase Lung Dis. 2000; 1 7:288-29 1 .
and methotrexate treatment. Br Med] 1 977;2 : 867. 5 9 . Sesin G P, Mucciardi N , Almeida S . Mesalamine-associated
42. Urban C, Nirenberg A, Caparros B, et al. Chemical pleuritis as pleural effusion with pulmonary infiltration. Am J Health Syst
the cause of acute chest pain following high-dose methotrexate Pharm. 1 99 8 ; 5 5 :2304-23 0 5 .
treatment. Cancer. 1 983;5 1 :34-37. 60. Pent MT, Ganapathy S, Holdsworth CD, et a l . Mesalamine
43. Kamenetsky Z, Da'as N, Esayag Y, et al. Valproic acid-induced induced lupus-like syndrome. BM] 1 992;305 : 1 5 9 .
eosinophilic pleural effusion: a case report and review of the 6 1 . Kristenson M, Fryden A. Pneumonitis caused b y metronida
literature. Neurologist. 20 1 2; 1 8 :39-40. zole. JAMA. 1 98 8;260: 1 84.
44. Thompson J, Chengappa KN, Good CB, et al. Hepatitis, 62. Siddiqui A, Ansari M, Shakil J, et al. Minoxidil-associated
hyperglycemia, pleural effusion, eosinophilia, hematuria and exudative pleural effusion. South MedJ 20 I O; I 03:45 8-460.
proteinuria occurring early in clozapine treatment. Int Clin 63. Gunstream SR, Seidenfield JJ, Sobonya RE, et al. Miromycin
Psychopharmacol. 1 998; 1 3 : 9 5-9 8 . associated lung disease. Cancer Treat Rep. 1 9 83;67:3 0 1 -304.
4 5 . Stanislav S W, Gonzalez-Blanco M . Papular rash and bilateral 64. Middleron KL, Santella R Jr, Couser JI. Eosinophilic pleuritis
pleural effusion associated with clozapine. Ann Pharmacother. due to propylthiouracil. Chest. 1 99 3 ; 1 0 3 : 9 5 5-956.
1 999;33: 1 008-1 009. 6 5 . Sen N, Ermis H, Karatasli M, et al. Propylthiouracil-associated
46. Catalano G, Catalano MC, Frankel Wetter RL. Cloza eosinophilic pleural effusion: a case report. Respiration. 2007;
pine induced polyserositis. Clin Neuropharmacol. 1 997; 74:703-705 .
20:352-3 56. 6 6 . De Groot RE, Willems LN, Dijkman J H . Interstitial lung dis
47. Chatterjee A, Saffermaan A. Cellulitis, eosinophilia and uni ease with pleural effusion caused by simvastin. J Intern Med.
lateral pleural effusion associated with clozapine treatment. 1 996;239:36 1 -363.
] Clin Psychopharmacol. 1 997; 1 7 : 323-333. 67. Roncato-Saberan M, Hustache-Mathieu L, Hoen B . Eosino
4 8 . Daly JM, Goldberg RJ, Braman SS. Polyserositis associated philic pleural effusion caused by simvastatin after 13 years of
with clozapine treatment. Am]Psychiatry. 1 992; 149: 1 274- 1 275. exposure. Eur J Intern Med. 2006; 1 7:450.
49. Kupeli E, Ulubay G, Ulasli SS, et al. Cilazapril-induced pleu 68. Moufarrege G, Frank E, Carstens DD. Eosinophilic exudative
ral effusion: a case report and review of the literature. Ann pleural effusion after initiation of tizanidine treatment: a case
Thorac Med. 20 1 0;5 : 1 7 1 - 1 73 . report. Pain Med. 2003;4:8 5-90.
5 0 . Corp C C , Ghishan P K . Th e sulfone syndrome complicated 69. Kuwahara T, Hamada M, Inoue Y, et al. Warfarin-induced
by pancreatitis and pleural effusion in an adolescent receiving eosinophilic pleurisy. Intern Med. 1 995 ;34:794-796.
dapsone for treatment of acne vulgaris. J Pediatr Gastroenterol 70. Jo YM, Park TH, Jeong IH, et al. Warfarin-induced eosino
Nutr. 1 998;26: 1 03- 1 0 5 . philic pleural effusion. Korean Circf 20 1 1 ;4 1 : 1 09-1 1 2 .
P l e u ra l E ff u s i o n D u e to M i sce l l a n e o u s
D i se a ses
Th e exposure to asbestos is definitely associated with The pathogenesis of the pleural effusion that occurs
the occurrence of benign exudative pleural effusions. after asbestos exposure is not known but is prob
ably similar to that of pleural plaques, which are
described in Chapter 27. In the series of Epler et al.
I n c i d e nce
( 1 ) , 20% of the affected individuals had pleural
Epler et al. ( 1) reviewed the medical histories of 1, 13 5 plaques, whereas in the Hillerdal and Ozesmi series
asbestos workers whom they had followed for several (2) , 39 of 60 patients (65%) had bilateral pleural
years and found that 35 of the workers (3%) had plaques. It is likely that the presence of submicrosco
pleural effusions for which there was no other ready pic asbestos particles in the pleural space provides a
explanation ( 1 ) . In contrast, no unexplained effusions constant stimulation to the pleural mesothelial cells
were seen in the control group of 7 1 7 subjects. These (3) . When mesothelial cells are cultured in the pres
authors found a direct relationship berween the level ence of asbestos particles, they synthesize and release
of asbestos exposure and the development of a pleural a protein fraction with chemotactic activity for neu
effusion. In patients with heavy, moderate, and mild trophils, which appears to be interleukin 8 (IL-8) (4) .
asbestos exposure, the incidence of pleural effusion When crocidolite is instilled into the pleural spaces
was 9.2, 3 . 9 , and 0.7 effusions/ 1 ,000 person-years, of rabbits, chemotactic activity rapidly appears in the
respectively ( 1 ) . Pleural effusions occur sooner after pleural Buid and this chemotactic activity is signifi
asbestos exposure than do pleural plaques or pleural cantly inhibited by a neutralizing antibody to human
calcification. In the series mentioned in the preced IL-8 (4) . In addition, when rat mesothelial cells are
ing text, many patients developed pleural effusions incubated in the presence of crocidolite or chrysotile
within 5 years of the initial exposure, and almost all asbestos fibers, they secrete the fibroblast chemoat
did so within 20 years of the initial exposure. This tractant fibronectin (5) .
finding is in direct contrast to the occurrence of pleu The gross pathologic findings in patients with
ral plaques and pleural calcifications, which usually pleural effusions secondary to asbestos are not well
do not occur until at least 20 years after the initial defined. Mattson (6) performed a thoracoscopy on
exposure. Other investigators, however, have reported nine patients with asbestos pleural effusion and found
a much longer period berween the initial exposure that the visceral pleural surface was completely normal
and the development of the effusion. Hillerdal and in all patients but the parietal pleura was inflamed.
Ozesmi (2) reviewed 60 patients with asbestos pleu In contrast, Gaensler and Kaplan (7) reported that
ral effusions and found that the mean latency after both the visceral pleura and the parietal pleura of their
the initial exposure was 30 years and that only 4 of patients were thickened, and an irregular pleural sym
their patients had developed a pleural effusion within physis was seen in all patients. Perhaps the difference
1 0 years of the initial exposure. berween these series is that pleural disease had been
348
C H APTER 23 I P LE U RAL E F F U S I O N D U E TO M IS C E LLAN E O U S D I S EASES 349
present longer in the second group. Microscopic exam evaluated for exposure to asbestos. If such exposure
ination of the pleura reveals chronic fibrosing pleuritis is found and the patient is asymptomatic with a
with varying degrees of inflammation and vascularity, small pleural effusion, the effusion is probably due to
depending on the acuteness of the process (7,8). asbestos exposure.
Patients with pleural effusions secondary to asbestos The natural history of the patient with an asbestos
have surprisingly few symptoms ( 1 ,2) . In Hillerdal pleural effusion is one of chronicity with frequent
and Ozesmi's (2) series of 60 patients, 47% had no recurrences and sometimes the development of fibro
symptoms, 34% had chest pain, 6% had dyspnea, and sis of the parietal pleura ( 1 ,6-8) . The pleural effusion
the remainder had various other symptoms. Mattson on the average lasts several months, but eventually it
(6) reported that his patients often complained of clears and leaves no residual pleural disease in most
feeling heavy in their chest. Most of Gaensler and patients (2) . In the series of 35 patients followed by
Kaplan's (7) patients complained of pleuritic chest Epler et al. ( 1 ) for a mean period of 9.7 years, 29%
pain or progressive dyspnea, but these patients were of the patients developed recurrent benign effu
referred for symptoms rather than having their disor sions, more commonly on the contralateral side. In
der diagnosed on the basis of serial chest radiographs. approximately 20% of patients, massive pleural fibro
The chest radiograph usually reveals a small-to sis follows the asbestos pleural effusion, whereas in
moderate-sized pleural effusion, which is bilateral in an additional 20%, the ipsilateral costophrenic angle
approximately 1 0% of patients ( 1 ) . Many patients remains blunted after the effusion has resolved. At
have pleural plaques, whereas fewer than 5% have times, malignant mesotheliomas follow asbestos pleu
pleural calcifications, and approximately 50% have ral effusions. Three of the 6 1 patients (5%) in the
some evidence of parenchymal asbestosis ( 1 ,7) . series of Epler et al. developed a mesothelioma during
The pleural fluid associated with asbestos pleural the follow-up period. These mesotheliomas occurred
effusion is an exudate that is serous or serosanguineous 6, 9, and 1 6 years after the initial pleural effusion ( 1 ) .
(6) . The pleural fluid white blood cell (WBC) count
can be as high as 28,000/mm3, and the pleural fluid
PO ST-LU N G TRA N S P LA N TATI O N
differential WBC can reveal either predominantly
polymorphonuclear leukocytes or mononuclear cells Pleural effusions are common after lung transplan
(2) . Pleural fluid eosinophilia appears to be a char tation. Normally, 80% of the fluid that enters the
acteristic of asbestos pleural effusions. In one series, interstitial spaces of the lungs is cleared from the
more than 50% eosinophils were found in 5 of 1 1 lung through the lymphatics, whereas 20% is cleared
asbestos pleural effusions, and an additional 2 effu through the pleural space (see Chapter 2) . In the
sions had more than 1 5 % eosinophils ( 6) . In a second patient with a lung transplant, however, the lymphat
series (2) , 26% of 66 asbestos effusions had pleural ics are transected, and, accordingly, almost all the
fluid eosinophilia. Most asbestos pleural effusions fluid that enters the lung exits through the pleural
contain mesothelial cells (2) . space. The continuity of the lymphatics is restored
within 2 to 4 weeks of lung transplantation (9) .
Pleural effusions are usually not evident in the
D i a g nosis
immediate posttransplant period because the patients
The diagnosis of asbestos pleural effusion is one of have chest tubes. The amount of fluid that drains
exclusion. Patients with a strong history of exposure through the chest tube may be very large, particu
to asbestos and a pleural effusion should be closely larly if the patient has the reperfusion syndrome. In
evaluated for mesothelioma or metastatic broncho one patient with a severe reperfusion syndrome, the
genic carcinoma because these diseases occur much chest tube drained more than 600 ml/hour ( 1 0) .
more commonly in individuals exposed to asbestos. Prolonged chest tube drainage is necessary in many
If these diseases as well as tuberculosis and pulmo patients. In one series of 1 00 patients, the mean time
nary embolism are ruled out, the patient probably for chest tube drainage was 1 9 .3 days with a range of
has an asbestos pleural effusion and should be 5 to 52 days ( 1 1 ) . Most patients do not have large
watched. The occupational history of any patient with amounts of chest tube drainage. Judson et al. ( 1 2)
an undiagnosed exudative pleural effusion should be performed serial analyses on the chest tube drainage
350 PLE U RAL D I S EASES
from seven patients who had undergone lung trans lymphoproliferative diseases, and 1 5 of20 (75%) epi
plantation. They reported that the mean output sodes in which the histopathology was nonspecific.
from the chest tube fell from 400 mL/day on day 1 The high prevalence of effusion with the different
to 200 mL/day on day 4 and that these mean outputs entities after transplantation is probably due to the
were similar to those seen in patients undergoing cor fact that a larger percentage of interstitial fluid exits
onary artery bypass or other cardiothoracic surgeries through the pleural space in the patient after lung
( 1 2) . The pleural fluid on day 1 is a bloody neutrophil transplantation. The pleural effusion that accompanies
predominant exudate, and on day 7 is still bloody with acute lung rejection is a lymphocyte-predominant
a mean protein of 2 g/dL and a mean lactate dehy exudative pleural effusion ( 1 6) .
drogenase (LDH) level of more than twice the upper After lung transplantation, patients are at risk of
limit of normal. Over the same time, the mean WBC developing empyema. In one series of 392 patients
decreased from 1 0,600 to 637 cells/mm3 and neutro from the University of Pittsburgh, empyema devel
phils decreased from 90% to 49% ( 1 2) . Teixeira et al. oped in 1 4 patients (3.6%) at a mean time of 46 days
( 1 3) performed serial measurements of IL- 1 {3, IL-6, after transplantation ( 1 7) . Four of the 1 4 patients
IL-8, and VEGF in the pleural fluid of 20 patients (29%) died from complications of pneumonia or
who had undergone lung transplantation. They sepsis, or both ( 1 7) . In a more recent study ( 1 8) from
found that the levels of all four cytokines were much the Duke University, pleural infection occurred in
higher in the pleural fluid than in the serum and were 27% of 4 5 5 lung transplant recipients within 90 days
highest at 6 hours after the transplantation ( 1 3) . of transplantation. Pleural infection in this study was
With time the levels o f these cytokines tended to defined as a positive bacterial, fungal, or viral culture
gradually decrease ( 1 3) . or a pleural fluid WBC greater than 20,000 ( 1 8) .
Pleural abnormalities are common o n imaging Fungal pathogens accounted fo r more than 6 0 % of
studies of the chest posttransplantation. Ferrer et al. the infections and Candida albicans was the predomi
( 1 1 ) reported that on thoracic computed tomogra nant organism found ( 1 8) .
phy (CT) at 3 months posttransplantation, 34 of Th e omental flap used to prevent dehiscence o f the
58 patients (59%) had a pleural effusion whereas bronchial anastomosis may result in a pseudoeffusion
at 1 2 months posttransplantation 4 of 50 patients on the chest radiograph. The omentum with its blood
(8%) had a pleural effusion. Most patients (62%) at supply is introduced into the chest cavity through a
3 months and (96%) at 1 2 months posttransplanta small incision in the diaphragm. It is particularly likely
tion had pleural thickening ( 1 1 ) . to mimic an effusion on a supine radiograph ( 1 9) .
It appears that some patients may develop a benign Th e management o f the patient with a pleural
effusion 2 to 6 weeks posttransplantation. Shitrit et al. effusion post-lung transplantation is dependent upon
( 1 4) reported that 1 0 of 3 5 patients (29%) developed the size of the effusion and the symptoms of the
more than a minimal pleural effusion between 2 and patient. If the effusion occupies more than 2 5 % of
1 2 weeks posttransplantation. Two of these patients the hemithorax, a thoracentesis should be performed
had a parapneumonic effusion and one had rejec immediately in an attempt to ascertain the etiology
tion, but there was no explanation for the effusion of the effusion. A therapeutic thoracentesis is recom
in the other seven patients ( 1 4) . A thoracentesis was mended because if the patient has the typical post
performed in all the patients. In each instance, the lung transplant pleural effusion, this procedure is
pleural fluid was a lymphocyte-predominant exudate likely to be curative ( 1 4) . The other main consider
( 1 3) . Most did not recur after thoracentesis ( 1 4) . It is ations are pleural infection, chylothorax, congestive
likely that these effusions have a pathogenesis similar heart failure, and rejection of the lung. If the effusion
to those that occur after coronary artery bypass graft recurs after the therapeutic thoracentesis, consider
surgery (see Chapter 1 9) . ation should be given to small-bore catheter drain
It appears that patients who develop complica age (20) . If the effusion is very small, it can probably
tions after their lung transplantation are likely to have be ignored as these effusions are very common ( 1 1 ) .
a pleural effusion. In one series in children, radiologic I f the effusion occupies less than 2 5 % o f the hemi
findings were correlated with histopathologic diag thorax but represents more than just blunting of the
noses in 62 instances ( 1 5) : pleural effusions occurred costophrenic angle, a diagnostic thoracentesis should
with 1 4 of 1 9 (74%) episodes of acute rejection, 7 of be performed if the patient is complaining of short
8 (88%) instances of chronic rejection, 6 of 1 1 (55%) ness of breath or if the patient is not feeling up to
episodes of infection, 3 of 4 (75%) instances with par. It is important to remember that these patients
C H APTER 23 I P LE U RAL E F F U S I O N D U E TO M IS C E LLAN E O U S D I S EASES 35 1
are immunosuppressed and accordingly are more developed a right-sided empyema and an additional
likely to have infections. Moreover, patients with five patients developed a pleural effusion requiring a
pleural infections may not be febrile because of the thoracentesis (26) .
immunosuppression.
Y E L LOW NAI L SYN D RO M E
POST-BO N E M A R ROW
The yellow nail syndrome consists of the triad of
TRA N S P LA N TATI O N
deformed yellow nails, lymphedema, and pleural
Pleural effusions can occur o n occasion as a complica effusions. Until 1 986, only 97 patients had been
tion after bone marrow transplantation. Adam et al. reported with this syndrome (27) . In one series,
(2 1 ) reviewed 860 patients who underwent bone 89% of the reported cases had yellow nails, and these
marrow transplantation between 1 998 and 2006 at were the presenting manifestation in 37%. Lymph
Wayne State University and reported that pleural edema of various degrees was encountered in 80% of
effusions occurred in 64 (7.4%) . Malignancy was the reported cases and was the initial manifestation
responsible for nine effusions while four patients had in 34%. Pleural effusions were found in 36% of all
parapneumonic effusions or empyema (2 1 ) . Diag cases (27) . The three separate entities may become
noses were not established in the remaining patients manifest at widely varying times. For example, one
(2 1 ) . Seber et al. (22) reviewed the medical records patient developed lymphedema in childhood, chronic
of 1 ,905 patients who received bone marrow trans nail changes at age 78, and a pleural effusion in her
plants between 1 97 4 and 1 993 at the University of ninth decade (28). An occasional patient with the
Minnesota. They found seven patients who had unex yellow nail syndrome also has a pericardia! effusion
plained multiple effusions involving two or more of (29,30) or chylous ascites (30) . In a series (3 1 ) of 4 1
the pleural, pericardia!, or peritoneal cavities. The patients from the Mayo Clinic, other chronic respira
pleura was involved in all patients. All of these cases tory manifestations were present including bronchi
of polyserositis occurred in recipients of allogeneic ectasis (44%) , chronic sinusitis (4 1 %) , and recurrent
transplants. The pleural fluid was characterized by pneumonia (22%) . In this series, the median age at
a WBC count below 1 ,000 cells/mm3 and a protein diagnosis was 6 1 years (3 1 ) . There appears to be a rela
level below 3 . 0 g/dL. Because all patients had con tionship between rheumatoid arthritis and the yellow
comitant severe graft versus host disease, the effusions nail syndrome. Since 1 979, 1 0 patients have been
were also attributed to the same disease (22) . reported who had both the yellow nail syndrome and
rheumatoid arthritis (32) . Lastly, a recent report (33)
demonstrated that the yellow nail syndrome could be
P L E U RA L E F F U S I O N I N LIVI N G
associated with either common variable immunodefi
D O N O RS
ciency or the specific antibody deficiency syndrome.
The lungs for some lung transplantations are being The basic abnormality in this syndrome appears to
obtained from living donors. In one series of 62 living be hypoplasia of the lymphatic vessels. Lymphangio
donors, 4 patients (6.4%) had a pleural effusion requir grams of the lower extremity demonstrate hypoplasia
ing drainage with a pigtail catheter and 2 additional of at least some lymphatic vessels in most patients
patients (3.2%) had a loculated pleural effusion (23). with the syndrome (28) . Emerson (34) has postu
In a second report (24) , 2 of 2 1 donors (9.5%) had to lated that pleural effusions may develop when a lower
be readmitted to the hospital for the development of a respiratory tract infection or pleural inflammation
pleural effusion. damages previously adequate but impaired lymphatic
The livers for some liver transplantations are also vessels. Subsequently, the lymphatic drainage of
being obtained from living donors. Pleural effusions the pleural space is insufficient to maintain a fluid
may occur in the living donor partial hepatectomy. free pleural space. In one report, biopsy of the parietal
In one large series of 386 donors, pleural effusion pleura revealed abnormally dilated lymphatics, neo
occurred in 9 (2.3%) (25 ) . Five of these patients genesis of lymphatic channels, and edematous tissues
received a drainage catheter that remained in place for in some areas, suggesting some deficit in lymphatic
a median of 5 weeks (25 ) . In a second study (26) of drainage (3 5 ) . However, the albumin turnover in the
1 1 2 living donors, a CT scan was obtained on day pleural fluid is not greatly decreased in patients with
7 which revealed pleural effusion in 75% which were this syndrome (36) . There has been one report (37) of
bilateral in 5 5 % and right sided in 45%. Three patients the familial occurrence of the yellow nail syndrome.
3 52 PLE U RAL D I S EASES
In this report, the mother, three of four of her Subsequently chylous effusions developed in about
children, and her grandmother had the yellow nail half the animals between 4 and 25 days after the liga
syndrome (37) . tion. An increase in the intravascular hydrostatic pres
With this syndrome, the nails are yellow, thickened, sure in the external jugular vein above 1 0 cm H 2 0
and smooth and may show transverse ridging (38) . correlated with the development of both chylous and
They are excessively curved from side to side, and the bloody effusions in these animals (43 ) .
actual color is pale yellow to greenish. Onycholysis I n the clinical situation, pleural effusions are
(separation of nail from bed) is frequently present, very common with superior vena caval obstruction.
and nail growth is slow (38) . Rice et al. (44) reviewed the chest radiographs of
The pleural effusions are bilateral in approximately 67 patients with the diagnosis of the superior vena
50% of patients and vary in size from small to mas caval syndrome at a tertiary referral hospital. They
sive (38 ). Once pleural effusions have occurred with reported that the incidence of pleural effusions was
this syndrome, they persist and recur rapidly after a 70% in the 43 cases due to malignancy and was 5 8 %
thoracentesis (28) . The pleural fluid is usually a clear i n the 24 cases with benign etiologies (44) . Th e loca
yellow exudate with a normal glucose level and pre tion of the effusions was 23% unilateral on the left,
dominantly lymphocytes in the pleural fluid differen 39% unilateral on the right, and 39% bilateral (44) .
tial WBC (28,34,38) . The pleural fluid LDH tends to Most of the effusions occupied less than 2 5 % of the
be low relative to the pleural fluid protein level. In one hemithorax (44) . The pleural fluid was analyzed in 22
series (3 1 ) , 5 of 16 patients (3 1 % ) had a chylothorax. of the patients and was found either to be an exu
The pleural biopsy reveals fibrosis, nonspecific inflam date or a chylothorax in each instance (44) . The etiol
mation, or lymphocytic cellular infiltrates, none of ogy of the effusions in the patient with malignancy
which is diagnostic of the disease (39) . could well have been the malignancy itself, and one
The diagnosis is made when a patient has a chronic would expect these effusions to be exudative. Only
pleural effusion in conjunction with yellow nails or five of the effusions with benign etiologies were
lymphedema. No specific treatment exists for the sampled and of these, two were exudates, two chy
syndrome, but if the effusion is large and produces lous, and one reported as "bloody." It is not clear why
dyspnea, pleurodesis with a tetracycline derivative or the effusions secondary to the benign processes are
thoracoscopy with pleural abrasion should be consid exudative.
ered (3 1 ,3 5 ,38,40) . One patient with pleural effusion In neonates, superior vena caval thrombosis is also
secondary to the yellow nail syndrome has been treated associated with the development of bilateral pleural
successfully with a pleuroperitoneal shunt (4 1 ) . effusions. Dhande et al. (45) reported a series of five
babies who developed superior vena caval obstruc
tion as a complication of the use of central venous
S U P E R I O R V E N A CAVA L SYN D RO M E
catheters. The effusions occurred 7 to 1 9 days after
In sheep, elevation of the pressure in the superior vena the initial placement or change of a central venous
cava leads to the accumulation of pleural fluid. Allen catheter. All the infants required repeated thoracen
et al. (42) demonstrated that once the pressure in the teses to remove pleural fluid that accumulated at a
superior vena cava was elevated above 1 5 mm Hg, rate of up to 200 mL/kg/day. The fluid was a clear
pleural fluid accumulated. The higher the pressure in transudate (protein level 1 .2-2.2 g/dL) but became
the superior vena cava, the greater the rate of fluid chylous when feedings were given. These workers
accumulation. The fluid was transudative in that the attributed the pleural fluid accumulation to obstruc
ratio of the pleural fluid to serum protein was less tion of thoracic lymph flow into the venous system.
than 0 : 5 . These workers attributed the pleural fluid The incidence of superior vena caval thrombosis in
formation to either lymph leakage out of the lym infants who receive central venous catheters for total
phatics that pass through the chest or obstruction of parenteral nutrition is approximately 1 0% (46) .
lung or chest lymphatics with subsequent leakage of Approximately 3% of neonates who undergo neo
interstitial fluid into the pleural space. natal cardiac surgery will develop the superior vena
Ligation of the superior vena cava also leads to the caval syndrome, and approximately half of these will
formation of pleural fluid in dogs. Blalock et al. (43) develop a chylothorax (47) . Fibrinolytic therapy is
ligated the superior vena cava of dogs and reported sometimes useful in treating the venous thrombosis,
that within a few days of the procedure, bloody, non particularly if it is started within a few days of the
chylous pleural effusions developed almost universally. development of the syndrome (47) .
C H APTER 23 I P LE U RAL E F F U S I O N D U E TO M I SCE LLAN E O U S D I S EASES 353
Another patient had opacification of the entire hemi at least 3 months and reported that 21 % had pleu
thorax, with contralateral mediastinal shift. ral effusion. They had a total of 1 00 patients with
The pleural Buid in uremic pleuritis is an exudate pleural effusions while receiving dialysis, and the effu
that is frequently serosanguineous or frankly hemor sions had the following etiologies: heart failure, 46;
rhagic (5 9-6 1 ,64) . The glucose level is normal, and the uremia, 16; parapneumonic, 1 5 ; atelectasis, 1 1 ; and
differential WBC reveals predominantly lymphocytes miscellaneous, 1 2 (65) . Some patients on dialysis will
in most patients (60) . In one series of seven patients, the develop rounded atelectasis with their pleural effu
mean WBC was 1 ,23 1 cells/mm3, the mean neutro sions (64) . These patients seem particularly likely to
phil percentage was 22, the mean protein was 3.9, and develop pleural fibrosis and early pleurodesis has been
the mean LDH was slightly more than 50% the upper recommended (64) .
limit of normal for serum (65) . Pleural biopsy speci Patients receiving peritoneal dialysis have a high
mens invariably reveal chronic fibrinous pleuritis. prevalence of unexplained exudative pleural effusions
The diagnosis of uremic pleuritis is one of exclu (69) . The transudative pleural effusion due to the
sion in the patient with chronic renal failure (66) . dialysate passing through the diaphragm is discussed
Specifically, Buid overload (in such a case the Buid is a in Chapter 9. Kwan et al. (69) reviewed 1 ,038 patients
transudate) , chronic pleural infection, malignant dis who received peritoneal dialysis from 1 99 5 to 2004
ease, and pulmonary embolism need to be excluded. and reported that 22 patients (2%) had unexplained
There is one report (67) that suggests that measure exudative effusions. They felt that the effusions were
ment of the pleural Buid levels of neopterin might be due to uremia and recommended that the dialysis be
useful in diagnosing uremic pleural effusions. In this intensified in such patients (69) .
report, eight of nine patients (89%) with a uremic
pleural effusion had a pleural Buid neopterin level
A M Y LO I D O S I S
above 200 nmol/L, whereas none of the 85 other
patients with pleural effusions of varying etiologies A significant percentage o f patients with primary sys
had levels this high (67) . temic amyloidosis have a persistent pleural effusion.
Dialysis is the treatment of choice for patients with Berk et al. (70) reviewed the chans of 636 patients
uremic pleuritis. With dialysis, the effusion gradually with primary systemic amyloidosis seen at Boston
disappears within 4 to 6 weeks in approximately 75% University between 1 994 and 200 1 . They found a
of patients. In the remaining 25%, the effusion per total of 3 5 patients ( 5 . 5 %) who had persistent pleural
sists, progresses, or recurs. effusions that were defined as pleural effusions which
In an occasional patient, the pleural thickening failed to resolve despite thoracentesis and aggressive
is progressive and leads to severe restrictive ventila diuresis (70) . The patients underwent a median of
tory dysfunction and marked shortness of breath three thoracenteses (70) . The seriousness of the pleu
(6 1 -64) . At least three such patients have undergone ral effusions is underscored by the fact that 1 8 patients
a decortication procedure, and the operation was had chest tubes placed and 1 1 underwent pleurode
not complicated by severe bleeding in any of these sis (7) . An additional 1 0% to 1 5 % of their patients
patients (6 1 -63) . All three patients reported marked had less significant effusions (7 1 ) . Overall, there have
symptomatic improvement, and one patient's vital been relatively few reports of pleural effusions due to
capacity increased from 8 5 0 mL preoperatively to amyloid in the literature. Berk (7 1 ) could only find
1 ,600 mL 9 months postoperatively. On the basis 23 such cases in the English literature between 1 977
of these reports and the progressive nature of ure and 2004. In 21 of these 23 cases, amyloid infiltra
mic pleuritis, decortication should be considered in tion in the pleura was demonstrated (7 1 ) .
uremic patients with pleural thickening and severe Th e pleural Buids i n general were transudates with
respiratory symptoms. a mean pleural Buid-serum protein ratio of 0 : 3 and
There is a high incidence of pleural effusions in a mean pleural Buid-serum LDH ratio of 0 : 3 8 . How
patients who are receiving chronic hemodialysis. ever, 1 0 of the 27 effusions on which pleural Buid
Coskun et al. (68) reviewed the thoracic CT find analysis was available met Light's exudative criteria
ings of 1 1 7 uremic patients on long-term hemodi (70) . The median pleural Buid WBC count was only
alysis and reported that a pleural effusion was present 306 cells/mm3 and no one cell type predominated
in 5 1 %. The effusions were bilateral in 63%. Jarratt (70) . Two of the patients had chylothoraces (7 1 ) .
and Sahn (65) reviewed the medical records of hos Pleural biopsies were performed i n six patients and all
pitalized patients who had received hemodialysis for revealed amyloid (70).
C H APTER 23 / PLE U RAL E F F U S I O N D U E TO M IS C E LLAN E O U S D I S EASES 355
Th e etiology o f the pleural effusions has been on CT scan (75 ) . Five patients underwent diagnostic
attributed to direct disruption of pleural lymphatics thoracentesis, and the mean percentage of eosinophils
by deposits of amyloid in the parietal pleural. When in four patients was 3 8 % . One patient had pleural
patients with primary systemic amyloidosis and per fluid lymphocytosis without pleural fluid eosino
sistent pleural effusions and those with primary sys philia (75 ) . In a second article, Daimon et al. (76)
temic amyloidosis with heart failure and no effusion reported that pleural effusions were present in 23 of
are compared, there is no difference in septa! thick 29 patients (79%) with eosinophilic pneumonia. The
ness, left ventricular systolic function, or diastolic effusions were bilateral in 22 of the patients (76) .
compliance. However, right ventricular hypokinesis
did occur more frequently in the patients with the
EXTRA M E D U L LA RY H E M ATO PO I E S I S
effusions (70) . Because right ventricular hypokine
sis is not usually associated with the development of Extramedullary hematopoiesis occurs as a compensa
pleural effusions and as there are amyloid deposits in tory phenomenon in various diseases in which there
the parietal pleura, the effusions have been attributed is inadequate production or excessive destruction of
to the amyloid deposits. I have some reservations blood cells. Although the liver and the spleen are the
about this explanation because the amyloid deposits most common sites of extramedullary hematopoiesis,
do not explain the very high daily pleural fluid pro foci can occur in many other organs, including the
duction in amyloid patients with chest tubes. paravertebral areas of the thorax and the pleura.
The management of patients with persistent pleu An occasional patient with extramedullary hema
ral effusions due to primary systemic amyloidosis is topoiesis will develop a symptomatic pleural effusion
difficult. Initially, aggressive diuresis with therapeutic (77-79) . The diagnosis of extramedullary hemato
thoracentesis as needed should be attempted. If thora poiesis is suggested in the patient with severe anemia
centesis is required more frequently than weekly, more by the presence of immature blood cells and mega
aggressive measures are indicated (7 1 ) . In the series of karyocytes in the pleural fluid. However, the presence
Berk (7 1 ) , chest tubes were used to manage the pleural of immature blood cells in the pleural fluid is not
effusions in 1 8 patients and in each case daily fluid loss diagnostic of extramedullary hematopoiesis as it can
exceeded 500 mL for 5 to 1 2 days despite aggressive be seen in patients with sepsis or pleural infection or
diuresis and fluid restriction. Berk (7 1 ) reported that in patients who have undergone transplantation and
thoracoscopy with talc insuffiation was successful only are receiving cyclosporine (80) . Many patients also
if the fluid output per chest tube was less than 200 mL/ have multiple paravertebral masses (77) . Patients with
day. He reported two patients who were managed suc pleural effusions secondary to extramedullary hema
cessfully with an indwelling PleurX catheter (7 1 ) . topoiesis have been managed successfully by both
Anti-vascular endothelial growth factor (anti chemical pleurodesis (77, 8 1 ) and radiotherapy to the
VEGF) antibodies may have a role in the treatment masses of hematopoietic tissue (79) .
of refractory pleural effusions (72,73) . There is one
report (72) in which four patients with systemic amy
R U PT U R E O F A M E D IASTI N A L CYST
loidosis and pleural effusions refractory to diuretic
therapy were treated with bevacizumab, an anti-VEGF On rare occasions, pleural effusions result from rup
antibody. Three of the four patients had improvement ture of a benign germ cell tumor or a bronchogenic
in their pleural effusions, peripheral edema, and func cyst into the pleural space (82-84) . In one series of
tional status (72) . In a case report (73) , the adminis 17 cystic mediastinal teratomas, 4 of the patients
tration of bevacizumad resulted in a decrease in the had preoperative rupture with pleural effusion (83) .
pleural fluid production from 600 ml/day to 0 after Chemical analysis of the pleural fluid is confusing
two treatments 4 weeks apart. at times in these cases. Hiraiwa et al. (82) reported
one patient who developed a right pleural effusion
after rupture of a benign mediastinal teratoma in
ACUTE E O S I N O P H I LI C P N E U M O N IA
which the pleural fluid carcinoembryonic antigen
Acute eosinophilic pneumonia is associated with dif (CEA) level was elevated to 1 60 ug/L. I have seen
fuse pulmonary infiltrates on the chest radiograph another case in which the pleural fluid amylase level
and an increased number of eosinophils and an eleva was elevated due to high amylase levels in the germ
tion of IL-5 (74) . The prevalence of pleural effusion is cell tumor. Khalil et al. (84) reported two cases of
high in patients with acute eosinophilic pneumonia. bronchogenic cysts with associated pleural effusion.
In one study, 1 0 of 14 patients had pleural effusions No description of the fluid was provided (84) .
356 PLE U RAL D I S EASES
Th e pleural effusions, therefore, were attributed to seawater or freshwater ( 1 00) . After rats are drowned,
the radiation (94) . All patients developed their pleu the mean pleural fluid sodium and chloride levels
ral effusions within 6 months of completing radiation after seawater drowning were 1 1 7 and 98 mEq/L,
therapy, and every patient had concomitant radia respectively, whereas those after freshwater drown
tion pneumonitis (94) . The pleural fluid with radia ing were 7 1 and 54 mEq/L, respectively ( 1 00) . The
tion pleuritis has not been well characterized, but potassium levels did not differ in the two types of
one report described the fluid as an exudate with drowning and were very high (>60 mEq/L) . Similar
many vacuolated mesothelial cells (9 5 ) . Most pleu results have been found in humans ( 1 0 1 ) . In one
ral effusions were small, but at least one occupied study ( 1 0 1 ) of 24 cases of seawater drownings and
approximately 50% of the hemithorax. In 4 of the 9 cases of freshwater drownings, the mean sodium
1 1 patients, the fluid gradually disappeared spontane levels were 1 75 mEq/L in the seawater drownings
ously in 4 to 23 months. In the remaining patients, and 75 mEq/L in the freshwater drownings while the
the pleural effusions persisted, gradually decreasing in mean chloride levels were 1 79 mEq/L in the seawater
size over the follow-up period of 1 0 to 40 months. drownings and 57 mEq/L in the freshwater drown
Pleural effusions can also occur as a late compli ings. Although there are no reports of pleural effusion
cation of radiotherapy to the chest. Morrone et al. occurring in patients who survive near-drowning, it
(96) reported one case of bilateral pleural effusion is likely that some have pleural effusion because one
that developed in a patient 1 9 years after receiving of the patients who died in Morild's series (99) had
mediastinal radiotherapy for Hodgkin's disease. The 900 mL of pleural fluid and had been in the water
effusions were exudates with predominantly lympho only 6 to 7 minutes.
cytes. Thoracoscopy revealed that there were enlarged
lymphatic vessels in the visceral pleura. In a second
M I LK O F CA LCI U M P LE U RA L E F F U S I O N
report, a patient developed bilateral pleural effusions
8 years after receiving radiotherapy for Hodgkin's Milk o f calcium is a colloidal suspension o f precipi
disease. In this instance, thoracoscopy demonstrated tated calcium salts. It has been seen in various cystic
diffuse thickening of the pleura (97) . spaces such as the gallbladder, renal calyceal diver
ticula, adrenal cysts, and breast cysts ( 1 02) . Milk
of calcium can also collect in the pleural space. The
E L E CTRICAL B U R N S
radiographic picture is characteristic, showing a half
Individuals who have suffered major electrical burns moon or hemispherical calcium-fluid level ( 1 02) .
may develop a pleural effusion secondary to the burn. Im et al. ( 1 02) reported five patients with pleu
If the contact point for the electrical burn is over the ral milk of calcium who showed a loculated pleu
chest, the underlying pleura is damaged. Accordingly, ral collection with double contour on radiography
a pleural effusion develops within the first week of the and homogeneous calcification on CT scan. Four
accident, and an accompanying pneumonitis may be of the five patients gave a history of pleurisy more
seen (98) . The pleural fluid is an exudate that gradu than 10 years previously. Aspirated materials from
ally resolves over a period of several months. the pleural space consisted of thick yellow fluid
containing gritty particles. The concentration of
calcium in the aspirated material was greater than
D ROWN I N G
5 0 0 mg/dL ( 1 02) . The five patients were essentially
Individuals who drown have a substantial amount asymptomatic from the milk of calcium-fluid col
of pleural fluid at autopsy. Morild (99) reviewed lections and received no therapy for the pleural
the autopsies of 1 33 individuals who had drowned effusion ( 1 02) .
between 1 987 and 1 99 1 . He found that pleural effu
sions were present in 7 1 of the patients (53%) . The
ACUTE R E S P I RATO RY D I ST R E S S
mean amount of fluid was 433 mL, with a maximum
SYN D RO M E
of more than 3,000 mL. Effusions were more com
mon if the patient had been in the water for more There appears to be a high prevalence of small pleural
than 8 hours and were more common with saltwater effusions in patients with the acute respiratory distress
drowning (99) . syndrome (ARDS) . The origin of the pleural fluid in
Analysis of the electrolytes in the pleural fluid patients with ARDS is probably the interstitial spaces
of drowning victims shows significant differences of the lung. Tagliabue et al. ( 1 03) reviewed the CT
depending on whether the drowning occurred in findings in 74 patients with ARDS. They reported
358 PLE U RAL D I S EASES
that pleural effusions were present in 37 (50%) . On 1 to 60 days ( 1 06) . The more common clinical symp
the chest radiograph, the effusion was apparent in 25 toms and signs are dyspnea (82%) , chest pain (46%) ,
of the 37 (68%) . The effusions were bilateral in 2 1 , respiratory failure ( 1 8%) , hypotension ( 1 3%), and
unilateral and right sided in 6 , and unilateral and left cardiac arrest (5%). The chest radiograph may reveal
sided in 1 0 . The effusions were small in 22 and mod unilateral or bilateral pleural effusions with or without
erate in 1 5 . In a second study, Talmor et al. ( 1 04) a widened mediastinum. A unilateral pleural effusion
prospectively studied 1 99 patients with ARDS in a may be ipsilateral or contralateral to the catheter's
surgical intensive care unit who required ventilation insertion site ( 1 06) .
with positive end-expiratory pressure (PEEP) . These The pleural fluid may be pure blood resulting from
investigators found that 19 patients ( 1 0%) who had laceration or puncture of one of the vessels ( 1 07) .
unsatisfactory oxygenation status had effusions visi More frequently, the pleural fluid reflects the charac
ble on the supine chest radiograph. When chest tubes teristics of the infusate. If the patient is receiving an
were inserted into these patients, their oxygenation intravenous fat emulsion, the pleural fluid may appear
status and their lung compliance improved ( 1 04) . milky and is easily confused with a chylothorax ( 1 08).
Patients with ARDS with significant pleural fluid If the patient is receiving 5% of dextrose-containing
should undergo a thoracentesis to delineate the char solutions, the pleural fluid glucose level is invariably
acteristics of the pleural fluid. If the patient is having higher than the simultaneous serum glucose level
trouble with oxygenation or with being weaned off ( 1 06) . The only other situations in which the pleural
the ventilator, a therapeutic thoracentesis should be fluid glucose is significantly higher than the simultane
performed. ously obtained serum glucose is with peritoneal dialy
sis with dialysates containing a high glucose level and
with esophageal rupture when food with high glucose
IATRO G E N I C P L E U RA L E F F U S I O N S
levels enters the pleural space. The pleural fluid protein
At times, physicians are responsible fo r the develop and LDH levels are usually very low ( 1 0 5).
ment of pleural effusions in their patients. The iatro The diagnosis should be suspected in any patient
genic effusions secondary to various pharmaceutical with a central line who has a large pleural effusion.
agents, radiation therapy, endoscopic esophageal If aspiration of the central catheter yields blood, a
sclerotherapy, the ovarian hyperstimulation syn perforation may still be present ( 1 09) . Analysis of
drome, and fluid overload are discussed elsewhere in the pleural fluid usually confirms the diagnosis. The
this book, as are those that occur following coronary treatment is to remove the catheter immediately.
artery bypass surgery, various transplantations and If the patient is in respiratory distress, a therapeutic
abdominal surgery. In this section, the iatrogenic thoracentesis should be performed. If fresh blood is
pleural effusions that result from misplacement of present, a chest tube should be inserted immediately,
percutaneously inserted catheters or enteral feeding and if bleeding persists, an exploratory thoracotomy
tubes, those associated with translumbar aortogra may be necessary ( 1 07) .
phy, and those resulting from rupture of silicone bag Many central lines are inserted to infuse chemo
mammary prosthesis are discussed. therapy. When such lines perforate the vein and enter
the mediastinum, the chemotherapeutic agents may
be quite toxic. Bozkurt et al. ( 1 1 0) were able to find
S u perior Ve n a Cava Perforati o n by a
seven such cases in the literature. The agents extrava
Centra l Cath ete r
sated included vincristine, vinblastine, 5-fluorouracil,
An uncommon but potentially fatal iatrogenic cause epirubicin, and daunorubicin. Pleural effusions devel
of pleural effusion is the misplacement of a percuta oped in four of the patients. The pleural effusion was
neously inserted catheter into the mediastinum or the described as a transudate in one of the patients ( 1 1 0) .
pleural space. As of 1 99 5 , there had been 35 reports
describing 69 patients with central venous catheter
Perfo ration o f P l e u ra with
induced hydrothorax ( 1 0 5 ) . The incidence of this
a N asogastric Tu be
complication has been estimated to be as high as
0 . 5 % ( 1 0 5 ) . The two major risk factors for the devel The development of soft, flexible, small-bore poly
opment of this complication are ( 1 ) catheter insertion urethane feeding tubes has made nasogastric and
from the left and (2) large-bore catheters ( 1 06) . nasoenteric feeding more practical and comfortable
The average time interval from catheter placement for patients. The increasing awareness by physicians
to the onset of symptoms is 2 days, with a range of of the importance of malnutrition and metabolic
C H APTER 23 I P LE U RAL E F F U S I O N D U E TO M IS C E LLAN E O U S D I S EASES 359
support has led t o an increase i n the use o f such tubes. inferior part of the pleural space. An exudative pleu
However, their use has been associated with signifi ral effusion may result from irritation of the pleural
cant pleural complications. In an 1 1 -month period space by the extravasated contrast medium. The
in one institution, there were four instances in which pleural effusion following translumbar aortographic
nasogastric tubes were placed in the tracheobronchial examination is sometimes frankly bloody and proba
tree for an incidence of approximately 0.3% of all bly results from blood leaking from the aorta into the
intensive care unit patients who received nasogastric pleural space. In such situations, a therapeutic thora
tubes ( 1 1 1 ) . These tubes are often inserted into patients centesis is usually sufficient treatment because the
who have taken an overdose. Because such patients are leak stops spontaneously ( 1 1 4) .
obtunded, sometimes the tubes enter the pleural space
without producing respiratory symptoms immedi
After M a m m o p l asty
ately. There have been several instances in which char
coal has been instilled into the pleural space ( 1 1 2) . There have been at least three case reports of a pleu
Pneumothorax i s the most common complication ral effusion developing after rupture of a silicone bag
( 1 1 3) , but the infusion of the enteral formula into the mammary prosthesis ( 1 1 5- 1 1 7) . One patient devel
pleural space or the development of an empyema also oped left-sided pleuritic chest pain 24 hours after
occurs relatively frequently ( 1 1 4) . The stylets used for she had sustained a blow of moderate severity on the
ease of insertion provide stiffness and strength to the left anterior chest wall. Five years previously, she had
tubing and allow easier advancement of the device. undergone bilateral augmentation mammoplasties
With the stylet in place, the tubing becomes stiff and with insertion of silicone bag prostheses. On physi
is able to perforate structures relatively easily. There cal examination, the breasts appeared equal in size.
are frequently no clinical clues that the tube has The chest radiograph revealed a large left pleural
entered the bronchial tree instead of the esophagus. effusion. Thoracentesis revealed slightly turbid fluid
The risk of this complication is much greater if the with a protein level of 4 . 6 g/dL and an LOH level
patient has an endotracheal tube in place or if he or of 372 IU/L. A subsequent pleural biopsy revealed
she is obtunded ( 1 1 3) . To prevent this complication, a dense mixed cellular infiltrate with several granu
these tubes should only be inserted by experienced lomas and large multinucleated giant cells sugges
individuals, and the tube should be removed immedi tive of a foreign body reaction. Two liters of pleural
ately if the patient starts coughing. If any resistance is fluid were aspirated, and an oily layer was observed
felt, no further attempts should be made to advance on the top of the fluid, which was consistent with
the tube. Before feeding is initiated, the position of the presence of silicone gel in the aspirate. After the
the tip of the tube should be confirmed radiographi aspiration, the pleural effusion did not recur. Two
cally ( 1 1 4 ) . The standard tests for the placement of other patients developed pleural effusions approxi
nasogastric tubes such as the insuffiation of air with mately 1 year after their implants had ruptured
auscultation over the left upper quadrant or the aspi ( 1 1 6, 1 1 7) . In one patient, there was viscid, yellow
ration of fluid are often misleading with the small ish pasty material that could only be obtained with
nasogastric tubes ( 1 1 4) . a 1 4-gauge needle ( 1 1 6) . The other fluid was yellow,
I f the tube enters the pleural space and the enteral and scanning electron microscopy was necessary to
solutions or charcoal are infused, tube thoracostomy demonstrate material with the electron energy pat
should be performed after the nasogastric tube has tern of silicone ( 1 1 7) . These three reports demon
been removed. In such a situation, the possibility of strate that silicone can reach the pleural space, but
an empyema should be evaluated because the inci once there, it does not elicit much of an inflamma
dence of empyema is high when these small tubes tory reaction.
enter the pleural space ( 1 1 3) .
REFERENCES
Tra n s l u m ba r Ao rtog ra phy
1 . Epler GR, McLoud TC, Gaensler EA. Prevalence and incidence
Pleural effusion may also complicate translum of benign asbestos pleural effusion in a working population.
bar aortographic examination ( 1 1 4) . Small pleural JAMA. 1 9 82;247: 6 1 7-622.
2 . Hillerdal G, Ozesmi M . Benign asbestos pleural effusion: 73
effusions requiring decubitus radiographs for their
exudates in 60 patients. EurJ Respir Dis. 1 987;7 1 : 1 1 3-1 2 1 .
demonstration occur frequently after translumbar 3 . Sargent EN, Jacobson G , Gordonson J S . Pleural plaques: a
aortographic studies ( 1 1 4) and are thought to be sec signpost of asbestos dust inhalation. Semin Roentgenol. 1 977;
ondary to the passage of the needle through the most 1 2 :287-297.
360 PLE U RAL D I S EASES
4. Boylan AM , Ruegg C, Kim KJ, et al. Evidence of a role for 2 6 . Dondero F, Taille C, Mal H, e t al. Respiratory complications:
mesochelial cell-derived interleukin 8 in the pathogenesis a major concern after right hepatectomy in living liver donors.
of asbesros-induced pleurisy in rabbits. J Clin Invest. 1 992; Transplantation. 2006;8 l : 1 8 1 - 1 86.
89: 1 257-1 267. 27. Nordkild P, Kromann-Andersen H, Struve-Christensen E.
5 . Kuwahara M, Kuwahara M, Verma K, et al. Asbestos exposure Yellow nail syndrome-the triad of yellow nails, lymphedema,
stimulates pleural mesochelial cells to secrete the fibroblast and pleural effusions. Acta Med Scand. 1 9 86;2 1 9 :22 1 -227.
chemoattractant, fibronectin. Am J Respir Cell Mo! Biol. 1 994; 28. Beer DJ, Pereira W Jr, Snider GL. Pleural effusion associated
1 0 : 1 67- 1 76. with primary lymphedema: a perspective on the yellow nail
6. Mattson S-B. Monosymptomacic exudative pleurisy in persons syndrome. Am Rev Respir Dis. 1 978; 1 1 7: 5 9 5-599.
exposed to asbestos dust. Scand] Respir Dis. l 975;56:263-272. 29. Morandi U, Golinelli M, Brandi L, et al. "Yellow nail
7. Gaensler EA, Kaplan AI. Asbestos pleural effusion. Ann Intern syndrome" associated with chronic recurrent pericardia! and
Med. 1 97 1 ;74 : 1 78-1 9 1 . pleural effusions. EurJ Cardiothorac Surg. 1 995;9 :42--44.
8 . Hillerdal G . Non-malignant asbestos pleural disease. Thorax. 30. Malek NP, Ocran K, Tietge UJ, et al. A case of the yellow nail
1 9 8 1 ;36: 669-675. syndrome associated with massive chylous ascites, pleural and
9 . Ruggiero R, Fietsam R Jr, Thomas GA, et al. Detection of pericardia! effusions. Gastroenterology. 1 996;34 :763-766.
canine allograft lung rejection by pulmonary lymphoscinti 3 1 . Maldonado F, Tazelaar HD, Wang CW, et al. Yellow nail
graphy. J lhorac Cardiovasc Surg. 1 994; 1 0 8 :253-25 8 . syndrome: analysis of 41 consecutive patients. Chest. 2008;
1 0. Raju S , Heath BJ, Warren ET, e t a l . Single and double-lung 1 34:375-3 8 1 .
transplantation. Problems and possible solutions. Ann Surg. 32. Lehuede G, Toussirot E, Despaux J, et al. Yellow nail syndrome
1 990;2 1 1 : 6 8 1 -69 1 . associated with thiol compound therapy for rheumatoid arthritis.
1 1 . Ferrer J , Roldan J , Roman A, et al. Acute and chronic pleural Two case reports. joint Bone Spine. 2002;69 :406--408 .
complications in lung transplantation. J Heart Lung Trans 3 3 . Gupta S, Samra D, Ye! L , e t al. T- and B-cell deficiency asso
plant. 2003;22 : 1 2 1 7- 1 22 5 . ciated with yellow nail syndrome. Scand J Immunol. 20 1 1 ;
1 2 . Judson MA, Handy J R , Sahn SA. Pleural effusions following 75: 329-33 5 .
lung transplantation. Time course, characteristics, and clinical 3 4 . Emerson PA. Yellow nails, lymphoedema, and pleural effu
implications. Chest. 1 996; 1 09 : 1 1 90- 1 1 94. sions. Thorax. 1 966;2 1 :247-253.
13. Teixeira RH, Antonangelo L, Vargas FS, et al. Cytokine profile 3 5 . Lewis M, Kallenbach J, Zalczman M, et al. Pleurectomy in
in pleural fluid and serum after lung transplantation. Trans the management of massive pleural effusion associated with
plant Proc. 2 0 1 0;42:53 1 -534. primary lymphoedema: demonstration of abnormal pleural
1 4 . Shicric D, Izbicki G, Fink G, et al. Lace postoperative pleural lymphatics. Thorax. 1 983;38:637-639.
effusion following lung transplantation: characteristics and clini 36. Mambretti-Zumwalt J, Seidman JM, Higano N. Yellow nail
cal implications. EurJ Cardiothorac Surg. 2003;23 :494-496. syndrome: complete triad with pleural protein turnover studies.
1 5 . Medina LS, Siegel MJ, Bejarano PA, et al. Pediatric lung trans South MedJ 1 980;73:995-997.
plantation: radiographic-histopathologic correlation. Radiol 37. Razi E. Familial yellow nail syndrome. Dermatol Online J
ogy. 1 993; 1 87:8 07-8 1 0 . 2006; 1 2 : 1 5 .
1 6 . Judson MA, Handy JR, Sahn SA. Pleural effusion from acute 3 8 . Hiller E , Rosenow E C III, Olsen AM . Pulmonary manifesta
lung rejection. Chest. 1 997; 1 1 1 : 1 1 28-1 1 30 . tions of the yellow nail syndrome. Chest. 1 972;6 1 :452--4 5 8 .
1 7. Nunley D R , Grgurich WF, Keenan RJ, et a l . Empyema 39. Cordasco E M Jr, Beder S, Colcro A , e t al. Clinical features o f
complicating successful lung transplantation. Chest. 1 999; the yellow nail syndrome. Cleve Clin]Med. 1 990:57:472-476.
1 1 5 : 1 3 1 2- 1 3 1 5 . 40. Jiva TM, Poe RH, Kallay MC. Pleural effusion in yellow nail
1 8 . Wahidi MM, Willner DA, Snyder LD, e t al. Diagnosis and syndrome: chemical pleurodesis and its outcome. Respiration.
outcome of early pleural space infection following lung trans 1 994; 6 1 :300-302.
plantation. Chest. 2009; 1 3 5 ;484-49 1 . 4 1 . Brofman JD, Hall JB, Scott W, et al. Yellow nails, lymphedema
1 9 . O'Donovan PB. Imaging of complications of lung transplan and pleural effusion. Treatment of chronic pleural effusion
tation. Radiographies. 1 993; 1 3 :787-796. with pleuroperitoneal shunting. Chest. 1 9 90;97:743-745 .
20. Marom EM, Palmer SM, Erasmus JJ, et al. Pleural effusions in 4 2 . Allen SJ, Laine GA, Drake RE, e t al. Superior vena caval pres
lung transplant recipients: image-guided small-bore catheter sure elevation causes pleural effusion formation in sheep. Am J
drainage. Radiology. 2003;228 :24 1 -245 . Physiol. 1 988;255 :H492-H49 5 .
2 1 . Adam AK, Zamluc M, Soubani AO . Th e yield and safety 4 3 . Blalock A , Cunningham RS , Robinson CS . Experimental pro
of choracentesis in hematopoietic stem cell transplantation duction of chylothorax by occlusion of the superior vena cava.
recipients. Lung. 2007; 1 8 5 : 257-262. Ann Surg. 1 936; 1 04: 3 5 9-364.
22. Seber A, Khan SP, Kersey JH. Unexplained effusions: associa 44. Rice TW, Rodriguez RM, Barnette R, et al. Prevalence and
tion with allogeneic bone marrow transplantation and acute characteristics of pleural effusions in patients with the superior
or chronic graft-versus-host disease. Bone Marrow Transplant. vena caval syndrome. Respirology. 2006; 1 1 :299-3 0 5 .
1 996; 1 7:207-2 1 1 . 4 5 . Dhande V, Kactwinkel J, Alford B . Recurrent bilateral pleu
23. Battafarano RJ, Anderson RC, Meyers BF, et al. Perioperative ral effusions secondary to superior vena cava obstruction as
complications after living donor lobectomy. J lhorac Cardio a complication of central venous cachecerization. Pediatrics.
vasc Surg. 2000 ; 1 20:909-9 1 5 . 1 983 ;72: 1 09- 1 1 3 .
24. Toyazaki T, Chen F, Shoji T, e t al. Postoperatie pleural effusion 4 6 . Swaniker F, Fonkalsrud EW: Superior and inferior vena caval
in living lobar lung transplant donors. Gen lhorac Cardiovasc occlusion in infants receiving total parenteral nutrition. Am
Surg. 20 1 1 ;59:440--442. Surg. 1 995;6 1 :877-8 8 1 .
25. Lee SY, Ko GY, Gwon DI, et al. Living donor liver transplan 47. Sharoni E , Erez E , Birk E , et al. Superior vena cava syndrome
tation: complications in donors and interventional manage following neonatal cardiac surgery. Pediatr Grit Care Med.
ment. Radiology. 2004;230 :443--449 . 200 1 ;2:40-43.
C H APTER 23 / PLE U RAL E F F U S I O N D U E TO M IS C E LLAN E O U S D I S EASES 36 1
4 8 . Chusid EL, Siltzbach LE. Sarcoidosis of the pleura. Ann Intern 7 3 . Bae S H , Hwang JY, Kim WJ, e t al. A case o f cardiac amy
Med. 1 974;8 1 : 1 90- 1 94. loidosis with diuretic-refractory pleural effusions treated with
49. Wilen SB, Rabinowitz JG, Ulreich S, et al. Pleural involve bevacizumab. Korean Circj. 20 1 0;40:67 1 -676.
ment in sarcoidosis. Am ] Med. 1 974;57:200-209. 74. Nishigaki Y, Fujiuchi S, Yamazaki Y, et al. Increased vascular
5 0 . Beekman JF, Zimmer SM, Chun BK, et al. Spectrum of endothelial growth factor in acute eosinophilic pneumonia.
pleural involvement in sarcoidosis. Arch Intern Med. 1 97 6; Eur Respir j. 2003;2 1 :774-778.
1 36:323-330. 75. Philit F, Etienne-Mastroianni B, Parrot A, et al. Idiopathic
5 1 . Nicholls AJ, Friend JAR, Legge JS. Sacco id pleural effusion: three acute eosinophilic pneumonia: a study of 22 patients. Am J
cases and review of the literature. Thorax. 1 980;35 :277-28 1 . Respir Crit Care Med. 2002; 1 66: 1 235-1239.
52. Huggins JT, Doelken P, Sahn SA, e t al. Pleural effusions in 76. Daimon T, Johkoh T, Sumikawa H, et al. Acute eosinophilic
a series of 1 8 1 outpatients with sarcoidosis. Chest. 2006; pneumonia: thin-section CT findings in 29 patients. Eur J
1 2 9 : 1 599-1 604. Radio/. 2008;65 :462-467.
53. Soskel NT, Sharma OP. Pleural involvement in sarcoidosis. 77. Peng MJ, Kuo HT, Chang MC. A case of intrathoracic extra
Curr Opin Pulm Med. 2000;6:45 5-468 . medullary hematopoiesis with massive pleural effusion: suc
54. Iyer S, Afshar K , Sharma OP. Peritoneal and pleural sarcoid cessful pleurodesis with intrapleural minocycline. J Formos
osis: an unusual association-review and clinical report. Curr MedAssoc. 1 994;93 :445-447.
Opin Pulm Med. 2008 ; 1 4:48 1 -487. 78. Bartlett RP, Greipp PR, Tefferi A, et al. Extramedullary hema
55. Akcay S, Pinelli V, Marchetti GP, et al. The diagnosis of sar topoiesis manifesting as a symptomatic pleural effusion. Mayo
coidosis pleurisy by medical thoracoscopy: report of three Clin Proc. 1 99 5 ;70: 1 1 6 1 - 1 1 64.
cases. Tuberk Toraks. 2008 ;56:429-433. 79. Garcia-Riego A, Cuinas C, Vilanova JJ, et al. Extracnedullary
56. Chittock DR, Joseph MG, Paterson NA, et al. Necrotizing hematopoietic effusions. Acta Cytol. 1 99 8 ;42: 1 1 1 6- 1 1 20 .
sarcoid granulomatosis with pleural involvement. Clinical and 80. Vaunois B, Breyton M, Seigneurin D, et a l . Intra-serous
radiographic features. Chest. 1 994; 1 06:672-676. haematopoiesis. In vivo. 2005 ; 1 9 :407-4 1 5 .
57. Bright R. Tabular view of the morbid appearance in 1 00 cases 8 1 . Aessopos A , Tassiopoulos S, Farmakis D, e t al. Extramedul
connected with albuminous urine, with observations. Guys lary hematopoiesis-related pleural effusion: the case of beta
Hosp Rep. 1 836; 1 :380-400. thalassemia. Ann Thorac Surg. 2006;8 1 :2037-2043 .
58. Hopps HC, Wissler RW. Uremic pneumonitis. Am J Pathol. 82. Hiraiwa T, Hayashi T, Kaneda M, et al. Rupture of a benign
1 9 55;3 1 :26 1 -273. mediastinal teratoma into the right pleural caviry. Ann Thorac
59. Nidus BD, Matalon R, Cantacuzino D, et al. Uremic pleuritis: Surg. 1 99 1 ; 5 1 : 1 1 0- 1 1 2 .
a clinicopathological entiry. N Engl] Med. 1 969;28 1 :2 5 5-256. 83. Choi SJ, Lee JS, Song KS, e t al. Mediastinal teratoma: CT
60. Berger HW, Rammohan G, Neff MS, et al. Uremic pleural differentiation of ruptured and unruptured tumors. AJR Am J
effusion: a study in 14 patients on chronic dialysis. Ann Intern Roentgenol. 1 998; 1 7 1 : 5 9 1-594.
Med. 1 975;82:362-364. 84. Khalil A, Carette MF, Milleron B, et al. Bronchogenic cyst
6 1 . Galen MA, Steinberg SM, Lowrie EG, et al. Hemorrhagic presenting as mediastinal mass with pleural effusion. Eur
pleural effusion in patients undergoing chronic hemodialysis. Respir j. 1 995;8:2 1 8 5-2 1 87.
Ann Intern Med. 1 975;82:359-36 1 . 85. Riemer H, Hainz R, Stain C, et al. Severe pulmonary hyper
62. Gilbert L , Ribot S , Frankel H , et al. Fibrinous uremic pleuritis: tension reversed by antibiotics in a patient with Whipple's
a surgical entiry. Chest. 1 975;67:53-56. disease. Thorax. 1 997;52: 1 0 1 4- 1 0 1 5 .
63. Rodelas R, Rakowski TA, Argy WP, et al. Fibrosing uremic 86. Muller C , Stain C , Burghuber 0 . Tropheryma whippelii in
pleuritis during hemodialysis. JAMA. l 980;243 :2424-242 5 . peripheral blood mononuclear cells and cells of pleural effu
6 4 . Horita Y, Noguchi M, Miyazaki M, e t al. Prognosis o f patients sion. Lancet. 1 993;34 1 :70 1 .
with rounded atelectasis undergoing long-term hemodialysis. 87. Im pens N , Warson F, Roels P, et al. A rare cause o f pleurisy.
Nephron. 200 1 ; 8 8 : 87-92. Eur j Respir Dis. l 986;68:388-38 9 .
65. Jarratt MJ, Sahn SA. Pleural effusions in hospitalized patients 8 8 . Moore PJ, Thomas PA. The trapped lung with chronic pleural
receiving long-term hemodialysis. Chest. 1 99 5 ; 1 0 8 :470-474. space, a cause of recurring pleural effusion. Mil Med. 1 967;
66. Maher JF. Uremic pleuritis. Am] Kidney Dis. 1 987; 1 0 : 1 9-22. 1 32:998-1 002.
67. Chiang CS, Chiang CD, Lin JW, et al. Neopterin, soluble 89. Light RW, Jenkinson SG, Minh V, et al. Observations on
interleukin-2 receptor and adenosine deaminase levels in pleu pleural pressures as fluid is withdrawn during thoracentesis.
ral effusions. Respiration. 1 994;6 l : 1 5 0- 1 54. Am Rev Respir Dis. 1 9 80; 1 2 1 :799-804.
68. Coskun M , Boyvat F, Bozkurt B, et al. Thoracic CT findings 90. Lee YC, Vaz MAC, Ely KA, et al. Symptomatic persistent
in long-term hemodialysis patients. Acta Radio/. 1 998;40: postcoronary artery bypass graft pleural effusions requiring
1 8 1 - 1 86 . operative treatment-clinical and hiscologic features. Chest.
69. Kwan B C , Chow KM, Pang WF, e t a l . Unexplained exudative 2001 ; 1 1 9:795-800.
pleural effusion in chronic peritoneal dialysis patients. Perit 9 1 . Pien GW, Gant M, Washam C, et al. Use of an implantable
Dial Int. 20 1 0;30:524-527. pleural catheter for "trapped lung" syndrome in patients with
70. Berk JL, Keane J, Seldin DC, et al. Persistent pleural effu malignant pleural effusion. Chest. 200 1 ; 1 1 9 : 1 64 1 - 1 646.
sions in primary systemic acnyloidosis: etiology and prognosis. 92. Huggins JT, Sahn SA, Heidecker J, et al. Characteristics
Chest. 2003 ; 1 24:969-977. of trapped lung: pleural fluid analysis, manometry, and air
7 1 . Berk JL. Pleural effusions in systemic acnyloidosis. Curr Opin contrast chest CT. Chest. 2007; 1 3 1 :206-2 1 3 .
Pulm Med. 2005; 1 1 :324-328. 9 3 . Doelken P, Huggins JT, Pastis NJ, e t al. Pleural manom
72. Hoyer RJ, Leung N, Witzig TE, et al. Treatment of diuretic etry: technique and clinical implications. Chest. 2004; 1 26:
refractory pleural effusions with bevacizumab in four patients 1 764- 1 769.
with primary systemic amyloidosis. Am J Hematol. 2007; 94. Bachman AL, Macken K. Pleural effusions following supervolt
82:409-4 1 3 . age radiation for breast carcinoma. Radiology. 1 9 59;72:699-709.
362 PLE U RAL D I S EASES
9 5 . Fentanes de Torres E, Guevara E. Pleuritis by radiation: 1 07. Holt S, Kirkman N , Myerscough E . Haemothorax after
report of two cases. Acta Cytol. 1 98 1 ;25 :427-429. subclavian vein cannulation. Thorax. 1 977;32: 1 0 1 - 1 03 .
96. Morrone N, Silva Volpa VLG, Dourado AM, et al. Bilateral 1 0 8 . Wolthuis A , Landewe R B , Theunissen P H , e t al. Chylo
pleural effusion due to mediastinal fibrosis induced by radio thorax or leakage of total parenteral nutrition? Eur Respir J
therapy. Chest. 1 993; 1 04: 1 276- 1 278. 1 998; 1 2 : 1 233-1 235.
97. Rodriguez-Garcia JL, Fraile G, Moreno MA, et al. Recurrent 1 09 . Kollef M H . Fallibility of persistent blood return for confir
massive pleural effusion as a late complication of radiother mation of intravascular catheter placement in patients with
apy in Hodgkin's disease. Chest. 1 99 1 ; 1 00: 1 1 65-1 1 66. hemorrhagic thoracic effusions. Chest. 1 994; 1 06: 1 906-1 908.
9 8 . Baxter CR. Present concepts in the management of maj or 1 1 0. Bozkurt AK, Uzel B, Akman C, et al. Intrathoracic exrrava
electrical injury. Surg Clin North Am. 1 970; 5 0 : 1 40 1 - 1 4 1 8 . sation of antineoplastic agents: case report and systematic
9 9 . Morild I . Pleural effusion i n drowning. A m J Forensic Med review. Am ] Clin Oncol. 2003;26: 1 2 1 - 1 23 .
Pathol. 1 99 5 ; 1 6 :253-256. 1 1 1 . Bankier AA , Wiesmayr M N , Henk C, e t al. Radiographic
1 00. Inoue H, Ishida T, Tsuji A, et al. Electrolyte analysis in detection of intrabronchial malpositions of nasogastric tubes
pleural effusion as an indicator of the drowning medium. and subsequent complications in intensive care unit patients.
Leg Med (Tokyo). 2005;7:96- 1 02. Intensive Care Med. 1 997;23 :406-4 1 0 .
1 0 1 . Usumoto Y, Sameshima N, Hikiji W, et al. Electrolyte analy 1 1 2. Sabga E, Dick A, Lertzman M, et al. Direct administration
sis of pleural effusion as an indicator of drowning in seawater of charcoal into the lung and pleural cavity. Ann Emerg Med.
and freshwater. J Forensic Leg Med 2009; 1 6:32 1 -324. 1 997;30:695-697.
1 02. Im JG, Chung JW, Han MC. Milk of calcium pleural 1 1 3 . Roubenoff R, Ravich WJ . Pneumothorax due to nasogastric
collections: CT findings. ] Comput Assist Tomogr. 1 993; feeding tubes. Report of four cases, review of the literature,
1 7:6 1 3-6 1 6 . and recommendations for prevention. Arch Intern Med.
1 03 . Tagliabue M , Casella TC, Zincone G E , e t al. C T and chest 1 9 89; 149: 1 84- 1 8 8 .
radiography in the evaluation of adult respiratory distress 1 1 4. Bilbrey GL, Hedberg CL. Hemorrhagic pleural effusion
syndrome. Acta Radio!. 1 994;35 :230-234. secondary to aortography: a case report. J 1horac Cardiovasc
1 04. Talmor M, Hydo L, Gershenwald JG, et al. Beneficial effects Surg. 1 967; 54:85-89.
of chest tube drainage of pleural effusion in acute respiratory 1 1 5 . Stevens UM, Burdon JG, Niall JF. Pleural effusion after
failure refractory to positive end-expiratory pressure ventila rupture of silicone bag mammary prosthesis. Thorax. 1 987;
tion. Surgery. 1 99 8 ; 1 23 : 1 37- 143. 42:825-826.
1 05 . Duntley P, Siever J, Forwess ML, et al. Vascular erosion by 1 1 6. Taupmann RE, Adler S. Silicone pleural effusion due to
central venous catheters. Clinical features and outcome. iatrogenic breast implant rupture. South Med J 1 993;
Chest. 1 992; 1 0 1 : 1 633- 1 63 8 . 86:570-57 1 .
1 06. Thurnheer R , Speich R . Impending asphyxia in a 27-year-old 1 1 7. Hirmand H , Hoffman LA, Smith J P. Silicone migration to
woman 14 days afrer a gynecologic operation. Chest. 1 99 5 ; the pleural space associated with silicone-gel augmentation
1 07: 1 1 69-1 1 7 1 . mammaplasty. Ann Plast Surg. l 994;32:645-647.
P n e u m ot h o ra x
A pneumothorax is air in the pleural space, that is, air pneumothorax in males and females aged 1 5 to 34
between the lung and the chest wall. Pneumothoraces was 37.0 and 1 5 .4/ 1 00,000/year, respectively. Since
can be divided into spontaneous pneumothoraces, which most patients in this age range have primary sponta
occur without antecedent trauma or other obvious neous pneumothorax, it appears that the incidence in
cause, and traumatic pneumothoraces, which occur the United Kingdom is greater than that previously
from direct or indirect trauma to the chest. A subcat reported in the United States (2) .
egory of traumatic pneumothorax is iatrogenic pneu
mothorax, which occurs as an intended or inadvertent
consequence of a diagnostic or therapeutic maneuver. Etiolog i c Factors
Spontaneous pneumothoraces are further divided
The general consensus is that primary spontaneous
into primary and secondary spontaneous pneumotho
pneumothorax results from rupture of subpleural
races. Primary spontaneous pneumothoraces occur
emphysematous blebs that are usually located in the
in otherwise healthy individuals, whereas secondary
apices of the lung (3,4) . In one older study, Gobbel
spontaneous pneumothoraces occur as a complication
et al. operated on 3 1 patients with primary sponta
of underlying lung disease, most commonly chronic
neous pneumothorax and found subpleural blebs or
obstructive pulmonary disease (COPD) .
bullae in each patient (3) . In a more recent study,
Lesur et al. (4) obtained computed tomography
P R I M ARY S P O N TA N E O U S (CT) scans on 20 young (mean age 27) patients with
P N E U M OT H O RAX spontaneous pneumothorax and could demonstrate
apical subpleural emphysematous lesions in 1 6 of
I n c i d e nce
the 20 patients (80%) . In another study, Bense et al.
The most complete figures on the incidence of pri obtained CT scans on 27 nonsmoking patients with
mary spontaneous pneumothorax probably come primary spontaneous pneumothorax and reported
from a study of the residents of Olmsted County, that 22 (8 1 %) had emphysema-like changes, mostly
Minnesota, where complete medical records are kept in the upper lobes (5). It appears that the apical blebs
on all residents. Between 1 95 9 and 1 978, 77 cases of present on direct visualization and the emphysema
primary pneumothorax occurred among the county's like changes seen on CT scan represent the same
population that averaged 60,000 over this period. abnormality. It should be noted, however, that there is
The age-adj usted incidence of primary spontaneous some controversy concerning the significance of the
pneumothorax was 7.4/ 1 00,000/year for men and subpleural blebs. Noppen et al. have identified abnor
1 .2 / 1 00,000/year for women ( 1 ) . If these figures are mal regions of the visceral pleura by fluorescein
extrapolated to the entire population of 2 5 0 million enhanced autofluorescence thoracoscopy and suggest
in the United States, one can anticipate approximately that leakage of air through these areas, rather than
1 0,000 new cases of primary spontaneous pneu rupture of blebs, may be responsible for primary
mothorax per annum. In a more recent study from spontaneous pneumothorax (6) . I believe that most
the United Kingdom, the incidence of spontaneous primary spontaneous pneumothoraces are due to
3 63
364 PLE U RAL D I S EASES
rupture of a bleb. The reason that I believe this is that Withers et al. (2 1 ) found that those with pneumo
the symptoms of primary spontaneous pneumotho thoraces were 2 in. taller and 25 lb lighter than the
rax start suddenly. If the pneumothorax were due to average military recruit (2 1 ) . Because the gradient
slow leakage through the visceral pleura, the symp in pleural pressure is greater from the lung base to
toms should not start abruptly. the lung apex in taller individuals (see Chapter 2) ,
The pathogenesis of these subpleural blebs is proba the alveoli at the lung apex are subjected to a greater
bly related to airway inflammation. Respiratory bron mean distending pressure in taller individuals. Over a
chiolitis was found in 70 of 79 patients (89%) who long period, this higher distending pressure could
underwent a surgical procedure for recurrence or per lead to the formation of subpleural blebs in taller
sistence of primary spontaneous pneumothorax (7) . individuals who are genetically predisposed to bleb
All the patients in this study were smokers, and ciga formation.
rette smoking can certainly produce airway inflam The tendency to develop a spontaneous pneumo
mation. Cigarette smoking is known to be strongly thorax may be genetically determined (22) . There is
associated with the development of primary sponta a high incidence of pneumothorax in patients with
neous pneumothorax. When the smoking habits of the Birt-Hogg-Dube syndrome. This syndrome is
5 0 5 patients from four separate studies were analyzed autosomal dominant (23) and is characterized by an
(8- 1 1 ) , 46 1 of the patients (9 1 %) were smokers. Fur increased incidence of spontaneous pneumothorax,
thermore, the occurrence of a spontaneous pneumo benign skin tumors, and renal tumors (23,24) . The
thorax appears to be related to the level of cigarette gene has been mapped to chromosome l 7p l 1 .2 (23) .
smoking. Compared with nonsmokers, the relative Mutations in the folliculin (FLCN) gene are respon
risk of a pneumothorax in men is seven times higher sible for the Birt-Hogg-Dube syndrome (25 ) . At least
in light smokers ( 1 - 1 2 cigarettes/day) , 2 1 times 53 different germline mutations and 3 1 SNPs have
higher in moderate smokers ( 1 3-22 cigarettes/ been identified in patients with the Burt-Hogg-Dube
day) , and 1 02 times higher in heavy smokers syndrome (26) . Pneumothorax occurred in 25 of 1 1 1
(>22 cigarettes/day) . For women, the relative risk patients (22 . 5 %) in one study (27) . Radiographi
is 4, 1 4, and 68 times higher in light, moderate, and cally, 1 5 % to 83% have pulmonary cysts and/or bul
heavy smokers than in nonsmokers, respectively ( 1 1 ) . lae (28). When these patients are explored surgically,
Disease o f the small airways related t o smoking prob they are found to have apical blebs (27) . Microscopic
ably contributes to the development of the subpleural examination of the resected lung tissue reveals cysts
blebs ( 1 2) . Interestingly, the prevalence of smoking comprising intraparenchymal collections of air sur
in Chinese patients with primary spontaneous pneu rounded by normal parenchyma or a thin fibrous
mothorax is only about 50% ( 1 3 , 1 4) . wall or blebs consisting of collections of air within
Two studies concluded that spontaneous pneu the pleura (28) .
mothoraces were more likely to develop following There have been other reports of a familial ten
days when there are broad swings in the atmospheric dency for the development of primary spontaneous
pressure ( 1 5 , 1 6) . It was postulated that the air in the pneumothorax. In one study of primary spontaneous
apical blebs was not in free communication with pneumothorax in the Israeli Defense Forces, 1 1 . 5 %
the airways. Therefore, when the atmospheric pres of 286 patients with spontaneous pneumothorax
sure falls, the distending pressure of the bleb may had a positive family history for primary spontane
increase and could result in its rupture ( 1 5) . It should ous pneumothorax (29) . A more in-depth analysis of
be noted that three other studies found no relation 1 5 families suggested that the mode of inheritance for
ship between change in the atmospheric pressure the tendency for pneumothorax was either autosomal
and the occurrence of a spontaneous pneumothorax dominant with incomplete penetrance or X-linked
( 1 7- 1 9) . In one study, however, there was a signifi recessive (30) . These reports were written before
cant relationship between thunderstorms and the the Birt-Hogg-Dube syndrome was described and
occurrence of pneumothoraces ( 1 7) . Noppen et al. the patients may have had this syndrome. In another
described the development of five episodes of primary report of patients with familial pneumothorax,
spontaneous pneumothorax in four patients upon individuals with human leukocyte antigen (HLA)
exposure to loud music (20) . haplotype A2 , B 40 were found to be much more
Patients with primary spontaneous pneumothorax likely to have a pneumothorax (3 1 ) . Other studies
are usually taller and thinner than control patients. of familial pneumothorax have been unable to docu
In a study of military recruits with pneumothorax, ment any association with the HLA haplotypes (32) .
C H A PT E R 24 I P N E U M OT H O RAX 3 65
Well-known inherited diseases associated with pneu series ofSeremetis ( 1 0) , five patients (3%) complained
mothorax include Marfan's syndrome, homocystin only of generalized malaise. On rare occasions, the
uria, Ehlers-Danlos syndrome, and a 1 -antitrypsin pneumothorax is discovered on a routine chest radio
deficiency (24) . graph (37) . Homer's syndrome has been reported as a
There is a very high prevalence of bronchial abnor rare complication of spontaneous pneumothorax and
malities in nonsmoking patients with spontaneous is thought to be due to traction on the sympathetic
pneumothorax. Bense et al. (33) performed fiberoptic ganglion produced by shift of the mediastinum (3 8).
bronchoscopy on 26 people who had never smoked Primary spontaneous pneumothorax usually
but had a history of spontaneous pneumothorax. develops while the patient is at rest. In the series of
They reported that 25 of 26 (96%) of the patients 2 1 9 patients of Bense et al. (39) , 87% were at rest at
had bronchial abnormalities bilaterally. In compari the onset of symptoms and none were exerting them
son, only 1 of 4 1 control patients had such abnor selves heavily when symptoms began. Other series
malities (33) . The bronchial abnormalities included have reported comparable findings (8, 1 0) .
disproportionate bronchial anatomy (smaller than Many patients with spontaneous pneumothorax
normal dimensions and deviating anatomic arrange do not seek medical attention immediately after the
ments of the airways at various locations) , an acces development of the symptoms. Eighteen percent of
sory bronchus, or a missing bronchus. The most the patients in one series had symptoms for more
common abnormality was the disproportionate than a week before seeking medical attention ( 1 0) ,
bronchial anatomy (33) . whereas 46% i n a second series waited more than
2 days before seeing a physician (8) . Patients with
symptoms for more than 3 days should not have nega
Path o p h ysiolog i c Featu res
tive pressure applied to their chest tubes in view of the
When resected specimens from the lungs of patients higher incidence of reexpansion pulmonary edema
with spontaneous pneumothorax are examined, there with prolonged pneumothorax (see Chapter 28).
is frequently an eosinophilic pleuritis (34) . In addi
tion, some patients have mild pulmonary vascular and
Changes o n Physica l Exa m i nation
perivascular eosinophilia (34) . Many patients also have
pulmonary artery intimal fibrosis and pulmonary vein Physical examination of patients with primary spon
intimal fibrosis (3 5). The eosinophilic pleuritis is prob taneous pneumothorax reveals vital signs that are
ably directly related to the air in the pleural space, and usually normal, with the exception of a moderate
the presence of abnormalities in the pulmonary ves tachycardia. If the pulse rate exceeds 1 40 or if hypo
sels should not serve as an indication to investigate the tension, cyanosis, or electromechanical dissociation is
patient for pulmonary vascular disease. present, a tension pneumothorax should be suspected
The physiological consequences of a pneumotho (see the section later in this chapter on tension pneu
rax are discussed in Chapter 3 . mothorax) . Examination of the chest reveals that
the side with pneumothorax is larger than the con
tralateral side and moves less during the respiratory
C l i n ica l M a n ifestations
cycle. Tactile fremitus is absent, the percussion note
The peak age for the occurrence of a primary spon is hyperresonant, and the breath sounds are absent
taneous pneumothorax is the early 20s, and primary or reduced on the affected side. The trachea may be
spontaneous pneumothorax rarely occurs after age 40. shifted toward the contralateral side. With right-sided
The main symptoms associated with the development pneumothoraces, the lower edge of the liver may be
of primary spontaneous pneumothorax are chest pain shifted inferiorly.
and dyspnea. In a series of 39 patients reported by
Vail et al. (36) , every patient had chest pain or dys
Electroca rd io g ra p h i c C h a n g e s
pnea, and both symptoms were present in 25 of the
39 patients (64%) . Seremetis ( 1 0) reported chest pain Patients with spontaneous pneumothorax may show
in 1 40 of 1 5 5 patients (90%) . The chest pain usually electrocardiographic changes due to the pneumotho
has an acute onset and is localized to the side of the rax (40) . In a study of seven patients with spontaneous
pneumothorax. On rare occasions, the patient has left pneumothorax, Walston et al. (4 1 ) found that a
neither chest pain nor dyspnea. The pneumothorax rightward shift of the frontal QRS axis, a diminution
is bilateral in less than 2% of patients ( 1 3) . In the of precordial R voltage, a decrease in QRS amplitude,
366 PLE U RAL D I S EASES
and precordial T-wave inversion could all occur. When pneumothoraces are induced in mice, the
A different study (40) reports that the V2--6 amplitude number of eosinophils in the pleural space is reduced
was decreased with left-sided pneumothorax. In a in tumour necrosis factor (TNF)-alpha knockout
study (42) of patients with right-sided pneumotho mice and in wild type mice who are given dexametha
rax, prominent R-wave voltage in lead V2 with loss sone (53). In a small percentage of patients, the pleu
of S-wave voltage, mimicking posterior myocardial ral effusion turns out to be a hemothorax that can be
infarction, and reversible reduced QRS voltage were associated with life-threatening hemorrhage (54) (see
reported. In another study (40) , the QRS amplitude Chapter 25) .
was increased in v5-6 with right-sided pneumothorax
(40) . In addition, marked PR-segment elevation in
Q u a ntitat i o n
the inferior leads with reciprocal PR-segment depres
sion in aVR has been reported (43 ) . These changes One should estimate the amount of lung collapse
should not be mistaken for an acute non-Q wave when treating a patient with a pneumothorax. The
myocardial infarction. There has also been a report volume of the lung and the hemithorax are roughly
of a patient with a tension pneumothorax who devel proportional to the cube of their diameters. Therefore,
oped pronounced ST-segment elevation in II, III, a one can estimate the degree of collapse by measuring
VF, and V4--G (44) . When a chest tube was inserted, an average diameter of the lung and the hemithorax,
the ST changes resolved and studies of myocardial cubing these diameters, and finding the ratios.
enzymes were negative (44) . Mathematically,
was very high, improvements can be made in the pre There was no significant relationship between the size of
ceding formula because it does not take into account the original pneumothorax or the treatment of the origi
the patient's size. Obviously, a very large person will nal pneumothorax and the recurrence rates. Twenty-four
have a smaller pneumothorax in relation to the overall of their patients (1 6%) had a pneumothorax on the con
size of the lung than a small person with identical dis tralateral side; in only one patient did the pneumothora
tances between the lung and the chest wall. ces occur simultaneously (63) . Gobbel et al. (3) followed
A third method for estimating the size of a pneu a group of 1 1 9 patients with spontaneous pneumothorax
mothorax, the Rhea method, uses a nomogram that for a mean of 6 years. These investigators found that,
relates the average intrapleural distance to the pneu of the 1 1 0 patients who did not have a thoracotomy at
mothorax size (57). On this nomogram, there is a the time of their initial pneumothorax, 57 (52%) had
1 0% pneumothorax for every centimeter of intra an ipsilateral recurrence. Once a patient had second
pleural distance. A recent article found that there and third pneumothoraces without thoracotomy, the
was a close correlation when the Collins method incidence of subsequent recurrence was 62% and 83%,
and the Rhea method were used to calculate pneu respectively.
mothorax percentage ( 5 8 ) . However, there is no Older studies suggested that there is substantial
close agreement between the Collins method and the risk of recurrence over many years. In the series of
Light index (59) . In general, the Rhea method or the Gobbel et al. (3) , the average interval between the
Collins method is recommended. first and the second pneumothorax was 2.3 years,
The size of a pneumothorax can also be calculated although the average interval for recurrence in the
from a CT scan of the thorax (60) . Cai et al. (60) series of Seremetis was 1 7 months ( 1 0) . However,
demonstrated that was very close agreement for the more recent studies have suggested that most recur
volume of a pneumothorax calculated via a computer rences occur within the first year (49,64,65).
from the chest CT scan and the volume of air aspi Attempts have been made to predict which patients
rated from the hemithorax. with a primary spontaneous pneumothorax are more
Position papers have used much simpler schemes likely to have recurrence. If one could predict which
for semiquantitating the size of pneumothoraces. In patients are more likely to have a recurrence, then
the British Thoracic Society's (BTS) guidelines for the those patients could be treated more aggressively to
management of spontaneous pneumothorax, small prevent a recurrent pneumothorax at the time of their
pneumothoraces were defined as those in which the first pneumothorax. The presence of blebs or bullae on
rim of air between the pleura and the chest wall at chest CT scan does not predict whether the patient will
the level of the hilum was less than 2 cm and large as develop a recurrent pneumothorax (49, 5 1 ) . Abolnik
greater than 2 cm (6 1 ) . In their consensus statement of et al. (29) did report that taller, thinner individuals
the management of spontaneous pneumothorax, the were more likely to have a recurrence. Guo et al. (66) ,
American College of Chest Physicians defined a small using a multivariate analysis of 1 3 8 patients, demon
pneumothorax as one in which the apex-to-cupola dis strated that recurrences were more frequent in taller
tance was less than 3 cm whereas a large pneumothorax patients and patients with lower weights (66) .
was one in which this distance was greater than 3 cm In a recent study, Ganesalingam et al. (67) carefully
(62) . Collins method and Rhea method are preferred studied the chest radiographs taken on the initial presen
to measuring just the apex-to-cupola distance because tation of 1 00 patients for spontaneous pneumothorax
they give a more precise estimate of the size of the for pleural thickening, blebs/bullae, pleural irregularities,
pneumothorax. and pleural adhesions. Over a mean follow-up period
of 57 months, 52% of the patients had a recurrence.
Patients having one, two, and three or more abnormali
Recu rre n ce Rates
ties were 3.0, 5.3, and 1 2.6 times more likely to develop
A patient who has had a primary spontaneous pneumo a recurrence, respectively (67) . They recommended that
thorax is at risk of having a recurrence. Sadikot et al. surgical treatment be offered to patients in whom two
(63) followed up 1 53 patients with primary spontane or more radiological abnormalities were identified (67) .
ous pneumothorax for a mean of 54 months and found
that the recurrence rate was 54.2%. In this study, the
Treatment
recurrence rates were less in men (46%) than in women
(7 1 %) and were less in individuals who stopped smok Therapy for the patient with primary spontaneous
ing ( 40%) than in those who continued smoking (70%) . pneumothorax has two goals: (a) to rid the pleural
368 PLE U RAL D I S EASES
space of its air and (b) to decrease the likelihood of a oxygen through a face mask. It is recommended that
recurrence. hospitalized patients with any type of pneumothorax
Several different treatments can be used for the who are not subjected to aspiration or tube thoracos
management of a patient with a primary spontaneous tomy be treated with supplemental oxygen at high
pneumothorax. These include observation; supplemen concentrations.
tal oxygen; simple aspiration; tube thoracostomy with
Aspiration
or without the instillation of a sclerosing agent; medical
thoracoscopy with the insuffiation of talc; video-assisted The initial treatment for most patients with primary
thoracoscopy with stapling of blebs, instillation of a spontaneous pneumothoraces greater than 1 5 % of
sclerosing agent, or pleural abrasion; and open thora the volume of the hemithorax should probably be
cotomy. In selecting the appropriate treatment for any simple aspiration (6 1 ,72-74) . With this procedure, a
given patient, it should be remembered that a primary 1 6-gauge needle with an internal polyethylene cath
spontaneous pneumothorax is mainly a nuisance and eter is inserted into the second anterior intercostal
is rarely life threatening to the patient. In the follow space at the midclavicular line after local anesthesia.
ing sections, discussions of the various treatments are An alternate site is selected if the pneumothorax is
provided. At the end of the section, recommendations loculated or if adhesions are present. After the needle
for the management of primary spontaneous pneumo is inserted, it is extracted from the cannula. Alterna
thorax are given. tively, one of the commercially available thoracentesis
trays such as the Arrow-Clark Thoracentesis Kit man
Observation
ufactured by Arrow International or the Argyle Turkel
If the communication between the alveoli and the Safety Thoracentesis Set distributed by Sherwood can
pleural space is eliminated, the air in the pleural be used. These kits have an outer cannula with an
space will be reabsorbed for the reasons discussed in inner needle. If they are used, it is important to make
Chapter 2. The rate of spontaneous absorption is slow, a large-enough skin incision so that the catheter does
however. Kircher and Swartzel (68) estimated that not become crumpled during its insertion.
1 .25% of the volume of the hemithorax was absorbed A three-way stopcock and a 60-mL syringe are
every 24 hours. Therefore, a pneumothorax occupying then attached to the catheter. Air is manually with
1 5% of the hemithorax would take 1 2 days to be com drawn until no more can be aspirated. If no resis
pletely reabsorbed. In another study, Kelly et al. (69) tance has been felt after aspirating a total of 4 L, it is
used the Collins method to assess the rate of absorp assumed that no expansion has occurred, and a tube
tion of pneumothoraces. They reported that the mean thoracostomy is performed. After no more air can be
rate of absorption was 2.2% per day, but there was aspirated, the stopcock is closed and the catheter is
much variation from patient to patient (69) . It is not secured to the chest wall. After 4 hours of observa
clear whether some of their patients were on oxygen tion, a chest radiograph should be obtained. If ade
which would increase the rate of absorption. quate expansion persists, the catheter can be removed
It is recommended that only patients with pneu and the patient discharged. Patients should return in
mothoraces occupying less than 1 5 % of the hemitho 24 to 72 hours for a follow-up chest radiograph.
rax be considered for this type of treatment. If the One multicenter, prospective, randomized study
patient is hospitalized, supplemental oxygen should compared manual aspiration versus chest tube drain
be administered to increase the rate of pleural air age for the first episode of primary spontaneous
absorption. pneumothorax. Sixty patients were randomized and
immediate success was obtained in 1 6 of 27 patients
Supplemental Oxygen
(59.3%) in the manual aspiration group and 28 of
The administration of supplemental oxygen acceler 33 (85%) in the chest tube drainage group (72) .
ates the rate of pleural air absorption in experimen Importantly, 1 3 of the 27 manual aspiration patients
tal and clinical situations. Chernick and Avery (70) did not require hospitalization (72) . Devanand et al.
administered humidified 1 00% oxygen to rabbits (75) performed a meta-analysis of three randomized
with experimentally induced pneumothoraces and controlled studies comparing manual aspiration and
found that the oxygen increased the rate of air absorp chest tube drainage and concluded that simple aspi
tion by a factor of 6. Northfield (7 1 ) reported that ration is advantageous in the initial management of
the rate of absorption was increased fourfold when primary spontaneous pneumothorax because of a
patients were treated with high-flow supplemental shorter hospitalization. A recent study randomized
C H A PT E R 24 I P N E U M OT H O RAX 3 69
1 37 patients with their first episode of primary spon might think that the placement of chest tubes would
taneous pneumothorax to aspiration versus tube irritate the pleura and produce at least a partial
thoracostomy (74) . The aspiration was initially suc pleurodesis, diminishing the likelihood of a recurrent
cessful in 40 of the 65 patients (62%) . Only 17 of the pneumothorax, the incidence of recurrent pneumo
65 patients in the aspiration group were admitted thorax is similar whether the initial episode is treated
to the hospital, and their mean hospital stay was by bed rest alone or by tube thoracostomy ( 1 0) .
1 . 8 days. All patients in the tube thoracostomy group When patients with spontaneous pneumothorax
were admitted with a mean stay of 4.04 days (74) . are managed with tube thoracostomy, several ques
One might worry that the use of manual aspira tions need to be addressed, such as what size of tube
tion would be associated with a higher recurrence rate needs to be used, whether the patient can be managed
in the patients who were successfully aspirated. This as an outpatient, whether to apply suction, when to
does not appear to be the case. For the patients in the remove the tube, and when to resort to more aggres
study of Noppen (72) , the recurrence rate in patients sive therapy. Although one earlier study (79) con
successfully treated with aspiration was 3 of 1 6 ( 1 9%), cluded that the success rates were poor when patients
whereas it was 4 of 1 1 (36%) in patients in whom with spontaneous pneumothorax were treated with
the aspiration was unsuccessful (73) . In a recent study, small chest tubes ( 1 3 F), subsequent studies have
Ayed et al. (74) reported that the recurrence rate over reported that most pneumothoraces are effectively
2 years was 29% in 40 patients in whom aspiration managed with small chest tubes (80-82) . Minami
was successful compared with 25% in 72 patients et al. (80) treated 71 episodes of spontaneous pneu
who were treated with tube thoracostomy. One might mothorax using a small caliber catheter (No. 5 . 5 or
hypothesize that the patients in whom aspiration is 7.0 F) connected to a Heimlich valve. They reported
successful have smaller blebs than those in whom it that the treatment was successful in 60 patients
fails (73). One study (76) reported that the admin (84 . 5 %) and ineffective in the remaining 1 1 patients.
istration of 300 mg minocycline after a successful air Only 6 of these latter 1 1 patients were successfully
aspiration reduced the incidence recurrent pneumo managed when a large chest tube was placed (80) . Liu
thorax from 1 1 of 33 (33%) to 4 of 3 1 ( 1 2.9%) . et al. (82) retrospectively reviewed the results of treat
One complication that has been reported with ing 1 02 patients with spontaneous pneumothoraces
anterior needle aspiration of pneumothoraces is life with 8- to 1 0-Fr pigtail catheters or conventional
threatening hemorrhage (77) . In one report, three chest tubes. They demonstrated that the results with
cases were described that developed life-threatening both treatments were comparable (82) .
hemorrhage after the procedure (77) . I know of no When patients with pneumothorax are treated
other similar reports, but the authors of this article with tube thoracostomy, small tubes (7- 1 4 F) should
suggested that it would be preferable to do the aspira be tried initially because their insertion is much less
tion in the fifth intercostal space in the anterior axil traumatic than that of larger tubes (6 1 ,62,83). They
lary line (77) . are best inserted using a guidewire technique, as
If aspiration fails, the two primary alternatives are described in Chapter 2 8 . If the lung does not reex
tube thoracostomy and VATS. Chen et al. (78) reported pand with the small tube, then a larger tube can be
that the recurrence rates were less after VATS and the inserted; however, it appears that most patients can
duration of hospitalization was less. be successfully managed with the small tube.
Patients with primary spontaneous pneumo
Tube Thoracostomy
thorax can be managed with tube thoracostomy on
With tube thoracostomy, the air in the pleural space an outpatient basis (84,8 5 ) . Ponn et al. (84) inserted
can be rapidly evacuated. The chest tube should 1 2-F or 1 6-F short catheters in 96 patients with
be positioned in the uppermost part of the pleural spontaneous pneumothorax. To prevent kinking,
space, where residual air accumulates. The management the tube was placed intracorporeally for most of its
of chest tubes in general is discussed in Chapter 29. length, with only 1 or 2 cm plus the Rared end left
Tube thoracostomy effectively evacuates the pleural outside the body. A Heimlich valve was connected
air if the tube is properly inserted. In one series of and secured with tape. An occlusive dressing cov
8 1 patients, only 3 patients (4%) had persistent air ered the entry site, and a gauze sponge was secured
leaks after several days of chest tube drainage ( 1 O) . The over the open end of the valve with a rubber band.
average duration of hospitalization in this series was Patients returned every 2 to 5 days for physical exami
only 4 days, with a range of 3 to 6 days. Although one nation and a chest radiograph. Using this procedure,
370 PLE U RAL D I S EASES
92 of 96 patients (96%) were treated successfully. Not all primary spontaneous pneumothoraces are
Dernevik et al. (85) reported that they managed 3 1 treated successfully with a small chest tube. If the
of 3 5 patients (8 8 . 5 %) on an outpatient basis with patient is initially treated with a small caliber chest
a device called a Tru-Close Thoracic Vent (Uresil, tube and the lung does not expand within 48 hours, a
Sweden) . This device is self-contained and consists larger chest tube should be placed. If the lung has not
of a chest tube, a Heimlich valve, and a thoracic vent expanded or a bronchopleural fistula persists after 3 or
(8 5). The cost reduction associated with outpatient 4 days, consideration should be given to more invasive
management is obvious; most patients with primary therapy such as thoracoscopy or thoracotomy or to
spontaneous pneumothorax who are subjected to performing an autologous blood patch (see later sec
tube thoracostomy should be treated as outpatients. tions in this chapter) . The insertion of additional chest
It is recommended that no suction be applied to tubes is not recommended (90) .
chest tubes inserted for spontaneous pneumothorax
Tube Thoracostomy with Instilla tion of a
(6 1 ) . The chest tubes can either be connected to a
Sc/erasing Agen t
Heimlich valve or an underwater seal. Two studies
(79 , 86) have concluded that the rate at which the Approximately 50% of patients with an initial pri
lung reexpands is similar whether suction is applied mary spontaneous pneumothorax have a recurrence
or not. Because the risk of reexpansion pulmonary whether they are treated with observation, aspira
edema is greater when suction is applied to the chest tion, or tube thoracostomy. Efforts have been made
tube (87) , and because the suction appears to offer no to diminish the recurrence rates by injecting various
benefit, suction is not recommended. In one series agents into the pleural space in an attempt to cre
(88), reexpansion pulmonary edema occurred in 1 6 ate an intense inflammatory reaction that would
o f 8 4 patients ( 1 9%) with spontaneous pneumotho obliterate the pleural space. Many different materi
rax who had - 20 cm H 2 0 suction applied to the als including quinacrine (9 1 ) , talc slurry (92) , olive
chest tube. Although this is a much higher incidence oil (93) , and tetracycline (65 ,94) have been instilled
than in most other studies, it does emphasize the rec through the chest tube at the time of the initial pneu
ommendation to not use suction. If the lung does mothorax in an effort to create a pleurodesis and pre
not expand after 24 hours of water-seal drainage or vent a recurrence.
Heimlich valve drainage, suction should be applied to The two agents that appear to be the best scleros
the chest tube. A Heimlich valve comes with some of ing agents are talc slurry and the tetracycline deriva
the commercially available kits (see Chapter 2 8 ) . It is tives. Most commonly, when talc is used as an agent
important to hook up the Heimlich valve in the right to affect a pleurodesis, it is insuffiated at the time
direction, or a tension pneumothorax can result (89) . of thoracoscopy or thoracotomy (see the discussion
The chest tube should remain in place for 24 hours in the following text) . There have been two reports
after the lung reexpands and the air leak ceases. If the (92,9 5 ) , however, with a total of 32 patients in whom
chest tubes are removed too soon after the lung reex 5 to 1 0 g of talc suspended in 2 5 0 mL of saline was
pands and the air leak ceases, there is a high likelihood administered intrapleurally. The recurrence rate in
of an early recurrence. Sharma et al. (86) reported a these 32 patients was less than 1 0%.
recollapse rate of25% in 20 patients in whom the chest The primary drawback to using talc is that a small
tubes were removed within 6 hours of lung expansion, percentage of patients develop the acute respira
but a recollapse rate of 0% in 20 patients in whom the tory distress syndrome (ARDS) from its instillation
chest tubes were removed 48 hours after lung expan intrapleurally (96) . The mortality rate is not incon
sion. There is controversy as to whether the chest tube sequential. In one recent study in which talc slurry
should be clamped if the lung has reexpanded and if was used to treat 240 patients with malignant pleural
there is no air leak. The thought behind this procedure effusion, the incidence of respiratory failure was 4%,
is that if there is a small air leak that is not obvious, and 5 (2%) of the patients died (97) . Although most
then a small pneumothorax will develop if the tube is deaths following talc slurry have occurred in patients
clamped and then chest tube drainage can be reiniti who were being treated for malignant pleural effu
ated. There are no studies to my knowledge evaluating sions rather than pneumothorax, the fact that ARDS
how many pneumothoraces will be detected with this and death do occur after administering talc slurry
approach. Chest tubes should certainly not be clamped intrapleurally should make one hesitant to use it for a
if there is an air leak because clamping in this situation benign condition in young healthy individuals (96) .
could lead to a tension pneumothorax. Talc insuffiation for prevention of pneumothorax
C H A PT E R 24 I P N E U M OT H O RAX 37 1
recurrence has also been associated with the develop requirements for parenteral antibiotics. Tetracycline
ment of chronic debilitating pain requiring thora derivatives appear comparable in effectiveness to tet
cotomy (98) . A more in-depth discussion of the side racycline. In the rabbit model, minocycline ( 1 03)
effects of talc can be found in Chapter 1 0 . or doxycycline ( 1 04) is as effective as tetracycline in
An alternative agent fo r pleurodesis i n patients producing a pleurodesis at approximately one fourth
with pneumothorax is a tetracycline derivative. In the the dose of tetracycline. Intrapleural doxycycline is
Veterans Administration (VA) cooperative study on also an effective treatment for malignant pleural effu
pneumothorax, 229 patients who were being treated sion (see Chapter 1 0) . Accordingly, 500 mg of doxy
with tube thoracostomy for spontaneous pneumo cycline intrapleurally is recommended for patients
thorax were randomized to receive 1 , 500 mg of tetra with spontaneous pneumothorax who are treated
cycline or nothing through their chest tube. During with chest tubes. An alternative agent is minocy
the 5-year study period, the 25% recurrence rate in cline (300 mg intrapleurally) (78) . Bleomycin should
the tetracycline group was significantly less than the not be used because it is ineffective in producing a
41 % recurrence rate in the control group (65) . In a pleurodesis in rabbits with a normal pleura ( 1 05) and
second study, Alfageme et al. (99) reported that the expensive.
recurrence rate was 9% in 66 patients treated with The intrapleural injection of tetracycline deriva
tetracycline intrapleurally, whereas the recurrence rate tives is an intensely painful experience for many
was 36% for the 79 patients treated with observation patients. In the VA cooperative study (65), more than
or chest tubes only (99) . In a more recent study of 50% of the patients reported severe pain at the time
1 3 8 patients that was not randomized, Guo et al. of the tetracycline injection, and 70% of the indi
used proportional hazards analysis and demonstrated viduals stated that the pain was greater at the time
that the administration of a sclerosing agent (tetracy of the tetracycline injection than at either the onset
cline 45 patients, gentamicin 23 patients) was associ of pneumothorax or at the time the chest tube was
ated with a significantly lower recurrence rate (66) . placed. The intrapleural administration of 1 00 mg of
In summary, the evidence presented earlier strongly lidocaine (Xylocaine) was not effective in amelio
suggests that the intrapleural injection of a sclerosing rating the intense chest pain. However, some have
agent in patients with spontaneous pneumothorax sig recommended that when a tetracycline derivative is
nificantly reduces the subsequent recurrence rates. It administered intrapleurally for pneumothorax, the
appears that at least in rabbits, the effectiveness of the injection be preceded by 4 mg/kg of Xylocaine up
pleural sclerosant is decreased if the animal is receiv to a maximal dose of 250 mg ( 1 06 ) . The patient
ing corticosteroids ( 1 00, 1 0 1 ) . The administration of should also be premedicated with an agent such as a
the nonsteroidal anti-inflammatory agent diclofenac short-acting benzodiazepine (e.g. , midazolam) .
(2 mg/kg body weight) also decreased the effectiveness It is recommended that the tetracycline derivative be
of pleurodesis from mechanical abrasion in pigs ( 1 02) . injected as soon as the lung has reexpanded. The patient
Therefore, one should attempt to minimize cortico should be positioned so that the tetracycline comes into
steroids and anti-inflammatory agents in patients in contact with the apical pleura. In experimental animals,
whom pleurodesis is attempted. Which patients with the presence of a small pneumothorax at the time of the
spontaneous pneumothorax should receive the intra injection does not decrease the efficacy of the pleurode
pleural injection of an agent in an attempt to produce sis ( 1 07) . A persistent air leak is not a contraindication
a pleurodesis and decrease recurrence rates? It is rec to tetracycline injection. There is, however, no evidence
ommended that all patients with primary or second that the intrapleural injection of tetracycline leads to an
ary spontaneous pneumothorax who are treated with earlier closure of the bronchopleural fistula (65 , 1 08).
tube thoracostomy receive an agent unless they are
Autologous Blood Patch for
subjected to thoracoscopy or thoracotomy.
Persisten t A ir Leak
What agent should be used? Currently, the recom
mended agent for an attempted pleurodesis through a In the past decade, there have been approximately
chest tube is a tetracycline derivative. Ifit were not for 50 papers reporting on the use of the autologous
the reported cases of ARDS after the administration blood patch for the treatment of a persistent air leak in
of talc slurry to patients with malignant pleural effu patients with spontaneous pneumothorax ( 1 09, 1 1 0) .
sions, talc slurry would be the recommended agent. With this technique, 5 0 to 1 00 m L o f blood is drawn
Parenteral tetracycline is no longer available in most from a vein and then promptly injected, without anti
countries due to increasingly stringent manufacturing coagulation, through the chest tube into the pleural
372 PLE U RAL D I S EASES
are treated is with an endoscopic stapling device. The postoperative hospital stay was only 3 days and the
primary disadvantage of the endostapler is its expense. recurrence rate was 1 .74%, with a mean follow-up of
The Endo-GIA model costs approximately US$500, 20 months ( 1 3 3 ). Cardillo et al. ( 1 34) used VATS to
and additional cartridges (of which an average of two treat 432 patients with primary spontaneous pneu
per procedure are used) cost US$500 each ( 1 24) . Previ mothorax between 1 992 and 1 99 8 . They used sub
ously, the bullae were treated with electrocoagulation, total pleurectomy to induce a pleurodesis in some
which was associated with a higher recurrence rate patients and talc insuffiation in others. The conver
( 1 25) . An alternative method of dealing with the api sion rate to open procedures was 2.3%, most often
cal bullae is to ligate the bullae with a Roeder loop due to extensive pleural adhesions. The mean time to
( 1 26) . However, Inderbitzi et al. ( 1 26), who have chest tube removal was 5 . 4 days, and the mean hos
reported one of the largest series using VATS for the pital stay was 6. 1 days. The recurrence rate was 4.4%,
treatment of pneumothorax, have reported a rela with a mean follow-up of 3 8 months ( 1 34) . In a more
tively high recurrence rate after use of the loop and recent study, Margolis et al. ( 1 38) performed VATS
recommend that it be abandoned in favor of wedge with stapling of blebs and pleural abrasion in 1 5 6
resection with the endostapler. One series concluded young (median age 1 9) patients with their first spon
that endostapling of the lung apex was associated taneous pneumothorax and reported no recurrences
with a decreased recurrence rate even if no blebs with a mean follow-up of 62 months. The mean total
are visible ( 1 27) . The recurrence rate during a mean hospital stay was 2.4 days ( 1 3 8 ) .
follow-up of 3 8 . 7 months was 7% in the 57 patients Cardillo e t al. ( 1 39) reported the largest series ever
who did not receive the stapling, whereas there were reported from a single institution. They treated 86 1
no recurrences in the 69 patients who received the patients with VATS and talc poudrage. If the patients
stapling ( 1 27) . had no visible blebs, they were treated with only talc
Once the lesion in the lung is treated, some attempt poudrage. Ifblebs or bullae were visible, they were sta
should be made to create a pleurodesis. Recurrence pled ( 1 39) . After a mean follow-up of 5 2 . 5 months,
rates are higher when only the blebs are treated the recurrence rate was 2.4% in the group without
( 1 28, 1 29) . The primary alternatives are mechanical blebs and 1 .7% in the group with blebs ( 1 39). Since
abrasion of the pleura, partial parietal pleurectomy, I would expect the group with no blebs to have less
talc insuffiation, and argon beam coagulation ( 1 30) . recurrences, this study suggests to me that treatment
Of these four, mechanical abrasion of the pleura is the of the blebs is important.
simplest. In one study ( 1 3 1 ) of 569 patients treated One paper randomly assigned 202 patients to
with VATS for spontaneous pneumothorax, the resection of blebs plus pleural abrasion or resection of
recurrence rates over a 5-year follow-up period were blebs plus pleural abrasion plus the instillation of300
3.6% for abrasion, 1 . 1 % for poudrage, and 2 . 5 % for to 400 mg minocycline at the end of the procedure
pleurectomy, which did not differ significantly. Since ( 1 40 ) . The group of patients that received minocy
there is no other evidence that partial pleurectomy, cline had more pain postoperatively, but there was a
talc poudrage, or argon beam coagulation is associated significant decrease in the recurrence rate for pneumo
with less recurrences than mechanical abrasion, it is thorax ( 1 .9% vs. 8 . 1 %) ( 1 40) . However, it should be
the method of choice. In another approach, Marcheix noted that the recurrence rates in the patients treated
et al. ( 1 32) treated 603 patients with stapling of the with resection of blebs plus pleural abrasion is much
blebs and spraying 1 % silver nitrate on the parietal higher than in most series. Accordingly, minocycline
pleural and reported that the long-term recurrence in addition to pleural abrasion is not recommended
rate was only 1 . 1 % . currently.
There have been several series, each with more Patients with hemopneumothoraces in which there is
than 1 00 patients, in which patients with spontane significant pleural hemorrhage are probably best man
ous pneumothorax were treated with VATS. In gen aged with VATS. Hwong et al. ( 1 4 1 ) performed VATS
eral, VATS with stapling of bullae is very effective at on 25 patients with spontaneous hemopneumothorax
managing spontaneous pneumothorax, with an over and reported that the bleeding was controlled in all the
all recurrence rate of approximately 3% ( 1 33- 1 37) . patients. VATS also appears to be effective in patients
Yim and Liu ( 1 33) treated 4 8 3 patients with VATS who have had a recurrence after talc insuffiation with
using mechanical pleurodesis plus some other proce medical thoracoscopy ( 1 42) . Doddoli et al. ( 1 42) suc
dures such as endostapling or endoloop for manage cessfully managed 27 of 39 such patients (69%) with
ment of the bullae. They reported that their median VATS . VATS is also an effective management strategy
374 P L E U RAL D I S EASES
for recurrent pneumothorax after a previous VATS conservative, and I will allow my patients to fly 7 days
procedure ( 1 43-1 45). after resolution of the pneumothorax.
Open Thoracotomy
Reco m m e n d ations
The indications for open thoracotomy are the same
as those for thoracoscopy. IfVATS is available, thora My recommendations for the management of a
cotomy is recommended only after VATS has failed. patient with a primary spontaneous pneumothorax
The reason for this recommendation is that the hospi are as follows. If the pneumothorax is small ( < 1 5 %
talization is shorter and the postoperative pain is less of the hemithorax) and the patient is asymptomatic,
severe after thoracoscopy ( 1 24, 1 46, 1 47) . However, observation is recommended. If the patient is at a site
in a meta-analysis of 29 studies ( 1 48) the recurrence where oxygen is available, high-flow supplemental
rate after open thoracotomy ( 1 . 1 %) was significantly oxygen should be administered. If the pneumotho
lower than that after VATS (5.4%), but many of the rax is > 1 5 % of the hemithorax, an attempt should
studies were done when thoracoscopy was first being be made to aspirate the pneumothorax. If this is suc
used for pneumothorax and recurrence rates fell as cessful and there is no recurrence over several hours,
the surgeons became more experienced. It should be the patient can be sent home. If the aspiration is
mentioned, however, that some surgeons still prefer unsuccessful, then the patient should be admitted
axillary mini-thoracotomy to VATS for the treatment to the hospital. If possible, thoracoscopy should be
of spontaneous pneumothorax ( 1 48, 1 49). The reason performed as soon as possible. Thoracoscopy is pre
for this preference is that time is saved because dou ferred to tube thoracostomy because it is associated
ble-lumen intubation is not required, the operating with less recurrences, a shorter mean hospital stay,
time is short, there is a good cosmetic result, and it is and a shorter duration of chest tube drainage ( 1 53) .
less expensive ( 1 4 9) . In general, if both medical thoracoscopy and VATS
At thoracotomy, the apical pleural blebs are over are available, VATS is preferred because with it
sewn and the pleura is scarified. This procedure is the apical blebs can be treated. If thoracoscopy is not
effective in controlling the pneumothorax and dimin available, tube thoracostomy should be performed
ishing the rate of recurrence. In one large series in with the injection of a sclerosing agent. The sclerosing
which 362 patients underwent parietal pleurectomy, agent of choice is doxycycline 500 mg. Talc is not rec
only two documented ipsilateral recurrences were ommended because its administration intrapleurally
reported, with an average follow-up of 4 . 5 years in can lead to ARDS that can be fatal. If a patient has a
3 1 0 patients ( 1 5 0) . The low rates of morbidity and recurrent pneumothorax, VATS or medical thoracos
mortality of the procedure are attested to in the same copy should be performed. VATS or medical thora
article ( 1 50). Only 1 operative death was reported in coscopy should also be performed if the patient has
the 362 operative procedures, and the average post an occupation, for example, airplane piloting, or an
operative period of hospitalization was only 6 days. avocation, for example, diving, where the occurrence
Various methods proposed for scarification of the of a pneumothorax might be life threatening. VATS
pleura range from visceral and parietal pleurectomy or medical thoracoscopy is not indicated if the aspi
to mere abrasion of the pleura with dry sponges. All ration is successful because most such patients will
of these procedures appear to be effective (8), but never have a recurrence.
because pleural abrasion with dry gauze is less trau
matic than pleurectomy and does not affect a poten
S E C O N DARY S P O N TA N E O U S
tial later thoracotomy, it is the procedure of choice.
P N E U M OT H O RAX
Air Tra vel
Secondary spontaneous pneumothoraces are more
If a patient has a spontaneous pneumothorax, when serious than primary spontaneous pneumothoraces
should they be allowed to travel by air? Commercial because they decrease the pulmonary function of a
airlines have adopted a 6-week "no fly" rule between patient with already compromised pulmonary func
pneumothorax occurrence and air travel ( 1 5 1 ) . This tion. The secondary spontaneous pneumothoraces
rule seems very arbitrary and there are no data to sup that occur in patients with the acquired immunodefi
port it. Currie et al. ( 1 52) reported two patients with ciency syndrome (AIDS), cystic fibrosis, tuberculosis,
loculated pneumothoraces who flew without inci lymphangioleiomyomatosis (LAM), and Langerhans
dent. I believe that the 6-week "no fly" rule is very cell histiocytosis are discussed in separate sections.
C H A PT E R 24 I P N E U M OT H O RAX 375
diagnosis of pneumothorax in patients with primary for primary spontaneous pneumothorax ( 64,65 , 1 63) .
spontaneous pneumothorax and traumatic pneumo Videm et al. ( 1 63) followed a total of 303 patients
thorax, it is less reliable in patients with COPD ( 1 5 9) . for a median period of 5 . 5 years and reported that 24
I f pleural gliding is present, one can b e confident that of the 54 patients (44%) with COPD had a recur
there is no pneumothorax. However, pleural gliding rence. In patients without COPD, 96 of 249 (39%)
is sometimes absent in patients with COPD who do had a recurrence ( 1 63 ) . In the VA cooperative study,
not have a pneumothorax ( 1 5 9 ) . 92 patients with secondary spontaneous pneumotho
Th e diagnosis o f pneumothorax i s established rax were treated with chest tubes without pleural scle
by the demonstration of a visceral pleural line. It is rosis and the recurrence rate was 47% with a median
sometimes difficult to see this line because the lung is follow-up of 3 years (65) . In this study, the recur
hyperlucent and little difference exists in radiodensity rence rate with primary spontaneous pneumothorax
between the pneumothorax and the emphysematous was 32% (65 ) . Guo et al. in a multivariate analysis of
lung. Frequently, the presence of the pneumothorax is the factors related to recurrent pneumothorax found
overlooked on the initial chest radiograph. One must that patients with secondary spontaneous pneumotho
distinguish a spontaneous pneumothorax from a large, races were significantly (p < 0.007) more likely than
thin-walled, air-containing bulla. The pleural line with patients with primary spontaneous pneumothorax to
a pneumothorax is usually oriented in convex fashion have a recurrence (66) .
toward the lateral chest wall, whereas the apparent
pleural line with a large bulla is usually concave toward
Treatm e n t
the lateral chest wall. If there is any doubt as to whether
the patient has a pneumothorax or a giant bulla, CT The goals of treatment of the patient with secondary
scan should be obtained because the two conditions spontaneous pneumothorax, as with primary spon
are easily differentiated with this procedure (45 , 1 60) . taneous pneumothorax, are to rid the pleural space
It is important to make the distinction between a large of air and to decrease the likelihood of a recurrence.
bulla and a pneumothorax because only the pneumo Achievement of these goals, particularly the second, is
thorax should be treated with tube thoracostomy. more important in the patient with secondary sponta
In patients with cystic lung disease, the presence of neous pneumothorax, however. A primary spontane
cysts and pleural adhesions sometimes makes it diffi ous pneumothorax or its recurrence is mostly just a
cult to determine whether a pneumothorax is present nuisance. In contrast, the occurrence of a pneumotho
on the routine chest radiographs. If patients with cystic rax in a patient with lung disease may be life threat
lung disease present with increased shortness of breath, ening, even though the mortality rate from recurrent
the possibility of a pneumothorax should be consid secondary spontaneous pneumothorax is low (65).
ered, particularly if the hemithoraces are asymmetric in The treatment options for the patient with a
size. In such cases, the CT scan will delineate whether a secondary spontaneous pneumothorax are the same
pneumothorax is present and will also assist in selecting as those for a patient with a primary spontaneous
the appropriate site for chest tube placement ( 1 6 1 ) . pneumothorax, as discussed earlier in this chapter.
Occasionally, secondary spontaneous pneumo There are far fewer articles written on the treatment
thoraces result from primary carcinoma of the lung of secondary spontaneous pneumothorax than there
with bronchial obstruction. One must recognize the are on primary spontaneous pneumothorax. The rec
radiologic signs of bronchial obstruction in these ommendations for the treatment of the patient with
patients because the insertion of chest tubes is con a secondary spontaneous pneumothorax differ from
traindicated. When a patient has a totally collapsed those of the patient with a primary spontaneous
lung, one should search for air bronchograms in the pneumothorax in the following ways.
lung. Air bronchograms are absent when there is an Nearly every patient with a secondary spontane
obstructing endobronchial lesion, but otherwise they ous pneumothorax should initially be hospitalized
are present ( 1 62) . If no air bronchograms are present, and managed by tube thoracostomy (90) . Aspiration
a bronchoscopic examination should be performed of the pneumothorax is not recommended because it
before a chest tube is inserted. is less likely to be successful ( 1 64, 1 65) and does noth
ing to diminish the likelihood of a recurrence. Even if
the pneumothorax is small, its evacuation can lead to
Recu rre n ce Rates
a rapid improvement in symptoms. Pigtail catheters
The recurrence rates for secondary spontaneous pneu appear efficacious in treating secondary spontaneous
mothorax appear to be somewhat higher than those pneumothorax ( 1 66) . Arterial blood gases usually
C H A PT E R 24 I P N E U M OT H O RAX 377
improve within 24 hours of instituting tube thora might be managed with lung transplantation. In the
costomy ( 1 5 8 ) . If the patient has respiratory failure past, patients were excluded from lung transplanta
necessitating mechanical ventilation, a chest tube tion if attempts at pleurodesis had been made on
should definitely be placed because the pneumotho the side of the proposed transplant owing to the
rax is likely to enlarge during mechanical ventilation. increased difficulty of the procedure and the risk of
Tube thoracostomy is less efficacious in secondary excessive bleeding. However, a consensus conference
spontaneous pneumothorax than in primary pneu statement in 1 998 on lung transplantation concluded
mothorax, however. In primary spontaneous pneu that pleurodesis was not a contraindication to lung
mothorax, the lung usually expands, and the air leak transplantation in patients with cystic fibrosis ( 1 72) .
ceases within 3 days. In secondary spontaneous pneu In one study, the outcome of 1 8 patients with a
mothorax due to COPD, the mean time for the lung previous intrapleural procedure was compared with
to expand is 5 days. In approximately 20% of patients 1 8 paired controls without previous intrapleural pro
with secondary spontaneous pneumothorax, the lung cedures and there was no difference in the outcomes
remains unexpanded or an air leak persists after of the two groups ( 1 73) .
7 days ( 1 5 8 , 1 62, 1 63, 1 67, 1 68). If the lung does not expand within 72 hours or
Once the lung has expanded, it is recommended if there is a persistent air leak for more than 3 days,
that attempts be made to prevent the recurrence of strong consideration should be given to performing
a pneumothorax (90) . This can be done with VATS, VATS or medical thoracoscopy ( 1 74, 1 75 ) . At thora
medical thoracoscopy, mini-thoracotomy, or the instil coscopy, the blebs are excised with a stapling instru
lation of a sclerosant through the chest tube. IfVATS ment and some other procedure is done to create a
is available, it is the procedure of choice because the pleurodesis. In one study, 22 patients with secondary
stapling of blebs and pleural abrasion reduces the spontaneous pneumothorax due to COPD, with a
likelihood of recurrence to less than 5 % . Onuki et al. mean age of 70 and a mean preoperative FEV1 that
( 1 69) subjected 53 patients with a mean age of 6 5 . 1 , was 48% of predicted, were subjected to VATS for
most of whom had persistent air leaks, to VATS . either persistent air leak ( 1 8 patients) or recurrent
They reported that during the mean follow-up period pneumothorax (4 patients) ( 1 74) . The mean duration
of 988 days, there was only one recurrence ( 1 69) . of the procedure was only 57 minutes, and only one
Shaikhrezal et al. ( 1 3 1 ) performed VATS plus a pro patient required mechanical ventilation during the
cedure to create a pleurodesis in 94 patients with immediate postoperative period. The mean duration
secondary spontaneous pneumothorax and reported of postoperative hospitalization was 9 days. The fact
that only 3.9% needed an additional operation that the mean hospitalization was 1 8 days before
over the next 5 years. Medical thoracoscopy with surgery suggests that the hospitalization would have
the insuffiation of talc also appears to be an effec been shortened if the procedure had been performed
tive treatment. Lee et al. ( 1 70) insuffiated 3 g talc in sooner. VATS failed in 4 of the 22 patients ( 1 8%) ,
4 1 patients with secondary spontaneous pneumo in that there was a large air leak postoperatively that
thorax whose mean age was 70.7 and whose mean necessitated thoracotomy. There were two deaths in
FEV1 was 4 1 % of predicted. Although the mortal this series; one patient developed a contralateral
ity was 1 0%, there were only two recurrences in the tension pneumothorax and a subsequent fatal myo
37 surviving patients ( 1 70) . If these procedures are cardial infarction, and the other developed broncho
not available, or if the patient refuses or is too sick to pneumonia after revisional thoracotomy and died
undergo surgery, then doxycycline can be injected of respiratory failure. None of the surviving patients
through the chest tube, as described earlier. However, had a recurrent pneumothorax ( 1 74) .
the intrapleural injection of a tetracycline derivative The three studies detailed earlier ( 1 69, 1 70, 1 74)
will only decrease the recurrence rate from approxi document that elderly patients with severe COPD
mately 50% to 30% (65) . It should be emphasized and pneumothorax can be managed successfully with
that there is a dearth of randomized controlled stud VATS or medical thoracoscopy with acceptable rates
ies comparing the different treatment modalities with of morbidity and mortality. Since the results with
secondary spontaneous pneumothorax ( 1 7 1 ) . these procedures are so good, it is recommended that
When one contemplates an attempt to prevent a they be considered if the lung remains unexpanded or
recurrent pneumothorax, the effect that the agent will if there is a persistent air leak after 3 days. If facilities
have on a future lung transplant should be considered are available for VATS, it is advisable to attempt it
if the patient has a disease such as LAM, cystic fibro relatively early rather than attempting to reexpand the
sis, interstitial pulmonary fibrosis, or COPD that lung with several chest tubes. Indeed, some authors
378 PLE U RAL D I S EASES
have recommended that all patients with secondary prophylactic pentamidine has varied widely. In one
spontaneous pneumothorax undergo VATS (90, 1 75 ) . series of 408 patients who were receiving prophylac
I f the patient i s a good operative candidate and i f tho tic pentamidine therapy in San Francisco, there were
racoscopy is readily available, I would agree with this 1 7 cases (4%) of spontaneous pneumothorax but only
recommendation. approximately one fourth of the 408 patients had
One procedure that should be considered in had P jirovecii pneumonia previously ( 1 83). Renzi
patients with persistent airleaks due to secondary et al. ( 1 84) reported that 5 of48 patients ( 1 0%) with
spontaneous pneumothorax is a blood patch as a history of P jirovecii infection who were receiving
discussed in the section on primary spontaneous prophylactic pentamidine therapy developed a spon
pneumothorax. Aihara et al. ( 1 76) reported that the taneous pneumothorax. In another small series of 1 3
blood-patch technique was successful in treating the hemophiliacs who were infected with human immu
airleak in 1 6 of 22 patients (73%) of patients with nodeficiency virus (HIV) and who were receiving
secondary spontaneous pneumothorax due to inter prophylactic pentamidine, 4 patients (3 1 %) developed
stitial lung disease. a spontaneous pneumothorax ( 1 8 5 ) .
In patients who have a persistent airleak, pro Th e etiology o f pneumothorax i n the patient with
cedures can be performed endobronchially in an AIDS may be changing. In a review ( 1 86) of 9,83 1
attempt to stop the air leak. Zeng et al. ( 1 77) studied non treated HIV-infected patients, 1 0 5 ( 1 .06%) of the
40 patients with persistent air leaks. They first patients had a spontaneous pneumothorax. The most
attempted to identify the bronchus from which common etiology was bacterial pneumonia (34.3%)
the air originated by occluding the various bronchi followed by P jirovecii infection (29 . 5 %) and tubercu
with a balloon. Then they attempted to occlude the losis ( 1 5 .2%) . Patients with CD4+ counts above 200
bronchus with a mixture of fibrin and autologous were more likely to have bacterial pneumonia while
blood. They reported that they were able to identify those with CD4+ counts below 200 were more likely
the bronchus leading to the air leak in 36 patients to have P jirovecii infection.
and were successful in stopping the airleak in 28 of The explanation for the high incidence of sponta
the 40 (70%) patients ( 1 77) . An alternative to neous pneumothorax in these patients with P jirovecii
occluding the bronchus with a mixture of fibrin and infection appears to be the presence of multiple sub
aurologous blood is to occlude it with an endobron pleural lung cavities, which are associated with sub
chial valve ( 1 78) . pleural necrosis ( 1 87- 1 90) . These bullous changes
and pulmonary cysts develop because of repeated epi
sodes of inflammation and cytotoxic effects of HIV
P N E U M OT H O RAX S E C O N DARY TO
on pulmonary macrophages ( 1 8 8 ) . Most patients
ACQ U I R E D I M M U N O D E F I C I E N CY
have radiologic evidence of fibrocystic parenchymal
SYN D RO M E
disease (Fig. 24. 1 ) ( 1 87) . If these patients are sub
In the 1 990s, a substantial proportion of secondary jected to surgery, there is diffuse involvement of the
spontaneous pneumothoraces occurred in patients lung parenchyma with greater involvement in the
with AIDS . Between 1 98 3 and 1 99 1 , 1 20 patients with upper lobe than in the lower lobe. Areas of necrosis
a spontaneous pneumothorax were seen at Parkland are usually present in consolidated areas of the lung,
Memorial Hospital in Dallas, Texas, and 32 (27%) and these areas are exceedingly friable and prone to
occurred in patients with AIDS ( 1 67) . Most patients laceration with the slightest manipulation. Emphyse
with AIDS who have a spontaneous pneumothorax matous blebs are located on the surface of the lungs,
have a history of Pneumocystis jirovecii infection are and the apex of the lungs contains multiple cysts based
on prophylactic pentamidine and have a recurrence on consolidated parenchyma ( 1 9 1 ) . Microscopically,
of the P jirovecii infection ( 1 79- 1 8 1 ) . Pulmonary the tissue specimens invariably reveal extensive necro
tuberculosis and pulmonary cryptococcosis are sis with a complete loss of the inherent architecture
also associated with spontaneous pneumothorax in ( 1 9 1 ) . When patients are studied prospectively, those
patients with AIDS. In one series, 1 3 of 35 patients with a lower diffusion capacity are more likely to
(37%) with spontaneous pneumothorax and AIDS develop a spontaneous pneumothorax ( 1 84) . The rea
had tuberculosis ( 1 82) . Most patients have a CD4+ son for the relationship between the aerosolized pent
count less than 1 00 cells/ mm3 ( 1 8 1 ) . The reported amidine and the occurrence of the pneumothorax is
prevalence of pneumothorax in patients who have not clear, but it is probably related to the fact that the
a history of P jirovecii infection and who receive aerosolized pentamidine does not reach the periphery
C H APTE R 24 I P N E U M OT H O RAX 379
or the inability of the immunocompromised host to Thoracoscopy should also be performed if the air leak
mount a brisk inflammatory response. If pleurodesis persists or the lung remains unexpanded for 3 days
is attempted through a chest tube, doxycycline is rec after tube thoracostomy is performed. Indeed, one
ommended. There has been one report in which six of can make a good case for thoracoscopy for all cases of
seven spontaneous pneumothoraces in patients with pneumothorax secondary to cystic fibrosis as soon as
AIDS were managed successfully with intrapleural the patient is stabilized with tube thoracostomy.
doxycycline (200) .
P N E U M OT H O RAX S E C O N DARY TO
P N E U M OT H O RAX S E C O N DARY TO TU B E RC U LO S I S
CYST I C F I B RO S I S
Th e prevalence o f secondary spontaneous pneu
Secondary spontaneous pneumothorax is also fre mothorax in patients hospitalized with pulmonary
quent with cystic fibrosis, a disease with a high preva tuberculosis is between 1 % and 3% (206) . In one
lence of severe CO PD. Flume et al. (20 1 ) reviewed the series from Spain, tuberculosis was the second lead
Cystic Fibrosis Foundation Patient Registry of more ing cause of secondary spontaneous pneumothorax
than 28,000 cystic fibrosis patients and reported that after COPD (207) . All patients with pneumothorax
3.4% of the patients had had a spontaneous pneumo secondary to tuberculosis should be treated with tube
thorax and that the annual incidence was 0 .64%. The thoracostomy. In one older series of 28 patients, 1 1
mean age of the patients at the time of their initial were treated by observation or repeated pleural aspi
pneumothorax was 2 1 .9 years (20 1 ) . Pneumothorax ration, and 7 of the 1 1 (64%) died. In contrast, of
occurs more frequently in patients with severe respira the 17 patients treated with chest tubes, only 1 (6%)
tory impairment. Seventy-five percent of cystic fibrosis died. Once chest tubes are placed in such patients,
patients with pneumothorax will have an FEV1 <40% a long period of chest tube drainage can be antici
of that predicted (20 1 ) . Cystic fibrosis patients with pated. The duration of tube thoracostomy ranged
a pneumothorax have a higher mortality than those from 5 days to 6 months, with a mean duration of
without even when the level of pulmonary dysfunc 50 days in one series (206) . Patients with pneumotho
tion is taken into consideration (20 1 ) . Approximately rax secondary to TB should have surgery if the airleak
1 6% to 20% of patients with cystic fibrosis who are persists more than a few days or if they have a relapse.
older than 1 8 years will experience a pneumothorax Recently Freixinet et al. (208) reviewed the cases of
some time in their lives (202) . pneumothorax secondary to tuberculosis at a hospital
The treatment of the secondary spontaneous in the Canary Islands and reported that there were
pneumothorax associated with cystic fibrosis is simi 47 cases between 1 9 89 and 20 1 0 . They inserted chest
lar to the treatment of that associated with COPD. tubes in all but one patient (208). They performed
Because the recurrence rate approaches 50% (203), thoracotomy or VATS in patients in whom the air
consideration should be given to preventing a recur leak lasted more than 1 0 days or who had a recurrent
rence. Almost all patients should initially be treated pneumothorax. An atypical pulmonary segmentec
with tube thoracostomy unless the pneumothorax tomy was performed in 1 0 patients for persistent air
is small. If the air leak ceases and the lung remains leak and 3 patients for recurrence. Five patients had a
expanded, consideration should be given to the pre lateral thoracotomy and 7 patients had VATS (208) .
vention of a recurrence with either thoracoscopy or There were no postoperative deaths and the duration
the intrapleural injection of a sclerosant. However, of hospitalization ranged from 4 to 73 days (208) .
the Cystic Fibrosis Pulmonary Guidelines do not
recommend either procedure after the initial pneu
PN E U M OTH O RAX D U E TO
mothorax (204) . Because many patients with cystic
LYM PHAN G I O LE I O M YO MATO S I S (LAM}
fibrosis are candidates for lung transplantation, one
has to consider the effect of the preventive measures LAM i s a rare condition characterized b y peribron
on the subsequent lung transplantation. As discussed chial, perivascular, and perilymphatic proliferation of
earlier, it appears that efforts to create a pleurodesis abnormal smooth muscle cells. LAM almost exclu
do not preclude a lung transplantation and do not sively affects women of childbearing age and presents
add appreciably to its complications ( 1 72, 1 73,20 5 ) . with slowly progressive breathlessness, chylothorax,
Accordingly, the procedure o f choice for prevention is recurrent spontaneous pneumothorax, or hemopty
VATS with the stapling of blebs and pleural abrasion. sis (209) . There is a high incidence of spontaneous
C H A PT E R 24 I P N E U M OT H O RAX 38 1
during the first 48 hours of her menstrual flow intermittent positive-pressure ventilation with posi
should be considered to have a probable catamenial tive end-expiratory pressures (23 1 ) . Overall, pneu
pneumothorax. The medical treatment of catame mothorax develops in 5% to 8% of babies with birth
nial pneumothorax is aimed at treating the endo weights 500 to 1 , 500 g (232) .
metriosis by suppressing the ectopic endometrium
(2 1 5 , 2 1 9,224) . This can be accomplished by admin
Pathog e n e s i s
istering gonadotropin-releasing hormone antagonists
such as Luprin (22 5 ) . Hormonal therapies that allow The pathogenesis of neonatal pneumothorax in
for menses do not always prevent catamenial pneu infants without RDS is related to the mechanical
mothorax and the recurrence rate for patients on problems of first expanding the lung. Karlberg (233)
hormonal therapy exceeds 50% (225 , 22 5 ) . The sur has demonstrated transpulmonary pressures averag
gical treatment for catamenial pneumothorax is tho ing 40 cm H 2 0 during the first few breaths of life,
racoscopy with closure of the diaphragmatic defects, with occasional transpulmonary pressures as high
stapling of any blebs in the lung, and pleural abra as 1 00 cm H2 0 . At birth, the alveoli usually open
sion. However, the recurrence rates in patients with in rapid sequence, but if bronchial obstruction
catamenial pneumothorax after this procedure are occurs from the aspiration of blood, meconium, or
approximately 30% (226) and are higher than with mucus, high transpulmonary pressures may lead to
any other condition. Bagan et al. (227) recommend rupture of the lung (230) . A transpulmonary pres
the coverage of the diaphragmatic surface by a poly sure of 60 cm H 20 ruptures adult lungs (230) ,
glactin mesh even when the diaphragm appears nor whereas a transpulmonary pressure of only 45 cm
mal. If facilities for thoracoscopy are not available, H 2 0 ruptures neonatal rabbit lungs (234) . There
then the same procedures can be performed with has been one family reported in which the maternal
a thoracotomy (2 1 5 ,228) . There has been one case grandfather, a maternal aunt, and an older sister,
report in which recurrent pneumothoraces developed along with the patient had spontaneous neonatal
after thoracoscopy with pleural abrasion and thora pneumothorax (23 5 ) .
cotomy with partial pleurectomy and plication of the I n infants with RDS, the pneumothoraces also
diaphragm; however, after bilateral tubal ligation, occur because of high transpulmonary pressures.
there were no recurrences (229 ) . With the infant breathing spontaneously, abnormally
negative transpulmonary pressures can be generated
because of the reduced lung volumes and the noncom
N E O NATA L P N E U M OT H O RAX
pliant lung. Intermittent positive-pressure ventilation
Spontaneous pneumothorax occurs more commonly is even more likely to produce high transpulmonary
in the newborn period than at any other age. In pressures and pneumothorax.
radiologic surveys, a pneumothorax is present shortly
after birth in 1 % to 2% of all infants (230), and a
C l i n i ca l M a n ifestati o n s
symptomatic pneumothorax is present in approxi
mately 0 . 5 % (230) . Spontaneous neonatal pneumo Depending on the size of the pneumothorax, the signs
thorax is twice as common in boys as in girls, and vary from none to severe acute respiratory distress.
the infants are usually full term or postterm (230) . In the infant with a small pneumothorax, no clini
In most instances, the baby has a history of fetal cal signs or mild apneic spells with some irritability
distress requiring resuscitation or a difficult delivery or restlessness may be present. Large pneumothoraces
with evidence of aspiration of meconium, blood, or incur varying degrees of respiratory distress, and, in
mucus (230) . severe cases, marked tachypnea (up to 1 20/minute) ,
The incidence of pneumothorax in infants grunting, retractions, and cyanosis are present (230) .
with respiratory distress syndrome (RDS) is high The detection of pneumothorax by physical examina
(230,23 1 ) . The more severe the RDS, the more likely tion is often difficult because abnormal physical signs
the infant will develop a pneumothorax. In one series are often not found. The most reliable sign is a shift
of 295 infants with RDS , 1 9% developed a pneumo of the apical heart impulse away from the side of the
thorax (23 1 ) . Pneumothorax developed in only 3 . 5 % pneumothorax. Because breath sounds are widely
o f those n o t requiring respiratory assistance, but it transmitted in the small neonatal thorax, apprecia
occurred in 1 1 % of those requiring continuous posi tion of diminished breath sounds on the affected side
tive airway pressure and in 29% of those requiring is difficult (230) .
C H A PT E R 24 I P N E U M OT H O RAX 383
In infants who develop pneumothorax as a com because o f the possibility that the pneumothorax
plication of RDS , the onset of the pneumothorax will enlarge or that a tension pneumothorax (see
is frequently heralded by a change in the vital signs the section later in this chapter) will develop (230) .
(23 1 ) . In the series of Ogata et al. (23 1 ) of 49 infants Supplemental oxygen can increase the speed at which
with pneumothorax complicating RDS , cardiac arrest the pneumothorax is absorbed, but it should be
marked the development of the pneumothorax in administered with care, particularly in the preterm
1 2 (24%). Most of the other babies had a decrease infant because of the dangers of retrolental fibroplasia
in the pulse of 1 0 to 90 beats/minute, a decrease in (230) . A thoracentesis should be performed or a chest
the blood pressure of 8 to 22 mm Hg, or a decrease tube should be inserted in the neonate who is more
in the respiratory rate of 8 to 20 breaths/minute than mildly symptomatic (23 8) . Smith et al. (238)
(23 1 ) . Although the Pao2 decreased with the devel reported that 22 of 54 neonates at the University
opment of pneumothorax, no consistent changes of Michigan with neonatal pneumothorax required
were seen in the pH or Paco2• The infant with RDS tube thoracostomy. In this study (23 8), 1 0 of these
who develops hypotension as a result of a pneumo 22 patients required mechanical ventilation for pro
thorax is at high risk of having an intraventricular gressive respiration failure. In addition (238), 1 0 of
hemorrhage. In one series, 32 of 36 infants (89%) these 22 neonates had pulmonary hypertension neces
with pneumothorax associated with hypotension sitating treatment with inhaled nitric oxide in 7 and
had a grade 3 or 4 intraventricular hemorrhage. In extracorporeal membrane oxygenation in 4. All these
contrast, only 3 of 3 1 ( 1 0%) infants with pneumo neonates survived (23 8).
thorax and normal blood pressure developed an It is important to be careful about the location o n
intraventricular bleed (236) . It is hypothesized that the chest wall where the chest tube is placed. Rainer
the hypotension results in a cerebral infarction, with et al. (239) reported two teenagers who had signifi
the intraventricular hemorrhage occurring after the cant breast deformities due to chest tube placement
systemic blood pressure has been raised to normal when they were infants. The chest tube should be
values (236) . The infants who developed hypoten placed in the anterior axillary line at a distance 4 to
sion had a higher mortality rate and more residual 5 cm inferior to the nipple in the fifth or sixth inter
brain damage than did those who maintained their costal space (239) .
blood pressure (236) . Tube thoracostomy should almost always be per
formed in infants with RDS and pneumothorax
because the pneumothorax compromises the patient's
D i a g nosis
already poor ventilatory status and often increases in
The diagnosis of pneumothorax should be enter size. Usually, the air leak is small, and intermittent
tained in any neonate with respiratory distress or positive-pressure ventilation can maintain adequate
in any infant with RDS who deteriorates clinically. gas exchange. In certain patients, however, air leaks
A radiograph of the chest is essential to differentiate are so large that most of the ventilation delivered
pneumothorax from pneumomediastinum, hyaline by the respirator exits the lung through the bron
membrane disease, aspiration pneumonia, congenital chopleural fistula. In such patients, high-frequency
cyst of the lung, lobar emphysema, and diaphrag ventilation may be the only method by which ade
matic hernia. A clinically significant pneumothorax quate gas exchange can be maintained (240) (see the
should be evident on a high-quality anteroposterior discussion on bronchopleural fistulas at the end of
or posteroanterior chest radiograph (230) . Transil this chapter) .
lumination of the chest with a high-intensity trans
illuminating light is also a rapid, accurate, and easy
IATRO G E N I C P N E U M OT H O RAX
way to make the diagnosis of pneumothorax in the
neonate (237) . The incidence of iatrogenic pneumothorax is high
and is likely to increase as the use of invasive pro
cedures continues to increase. In Olmsted County,
Treatm ent
Minnesota, between 1 9 5 0 and 1 974, 1 02 instances
The neonate without RDS who is asymptomatic or is of iatrogenic pneumothorax were reported, as com
mildly symptomatic can be treated by close observa pared with 77 cases of primary and 64 cases of sec
tion, and the pneumothorax resolves in most patients ondary spontaneous pneumothorax ( 1 ) . In the VA
over a few days. Close observation is necessary cooperative study on spontaneous pneumothoraces
3 84 PLE U RAL D I S EASES
in the 1 9 80s, data were collected on the incidence of 90 cases of spontaneous pneumothorax. The most
iatrogenic pneumothoraces at the same time (24 1 ) . common cause of iatrogenic pneumothorax was
These investigators reported that during the 4-year transthoracic needle aspiration (3 5), followed by tho
study period, there were 5 3 8 instances of iatrogenic racentesis (30), subclavian venipuncture (23), and
pneumothorax and 520 instances of spontaneous positive-pressure ventilation (7) . There was substantial
pneumothorax. This study probably underestimates morbidity from the iatrogenic pneumothoraces in this
the relative incidence of iatrogenic pneumothorax series. Most patients (65 of98) were treated with large
because some of the medical centers did not appear chest tubes that were in place 4.7 ± 3 . 9 days. Nine of
to be diligent in searching for iatrogenic pneumotho the patients required a second chest tube. Two patients
races (24 1 ) . The major causes of iatrogenic pneumo died from the iatrogenic pneumothorax (244) .
thorax in this study are shown in Table 24. 1 . There At present, the leading cause of iatrogenic pneu
is a significant cost for iatrogenic pneumothoraces. mothorax is transthoracic needle aspiration of lung
Zhan et al. (242) reported that iatrogenic pneumo masses. The incidence of iatrogenic pneumothorax
thoraces occur in 0 .67/ 1 ,000 hospitalized patients with this procedure in five series, each with more than
and that these patients stayed in hospitals for an 300 patients, ranged from 1 9% to 40% (245-249) .
average of 4.4 extra days, incurred approximately The percentage of patients undergoing needle aspi
US$ 1 8,000 in excess charges, and had a 6% higher ration of the lung who are treated with chest tubes
risk of death in the hospital. ranges from 2 to 8 (245-249) . The two primary fac
The incidence of iatrogenic pneumothoraces is tors related to the development of the pneumotho
particularly high in patients in the intensive care unit. rax are the depth of the lesion and the severity of the
In one prospective study of 3,430 patients admitted underlying lung disease (245,247) . In one study, the
for more than 24 hours to an intensive care unit, incidence of pneumothorax was 1 5 % if no aerated
the incidence of pneumothorax was 3.0% (243) . The lung was traversed and approximately 50% if aerated
etiologies of the iatrogenic pneumothoraces in these lung was penetrated (247) . In this same study, the
patients were mechanical ventilation in 42, central incidence of pneumothorax was 49% if emphysema
venous catheters in 28, thoracentesis in 2 1 , and mis was present on the CT scan and 3 5 % if emphysema
cellaneous in 3 (243) . was absent. Patients with emphysema were three times
There is a substantial rate of morbidity and even more likely to undergo chest tube drainage than were
some deaths associated with iatrogenic pneumotho patients without emphysema (247) . In a more recent
rax. Despars et al. (244) reviewed the cases of iatro study (250) of 1 ,098 CT fluoroscopy-guided lung
genic pneumothoraces at the VA Medical Center in biopsies with a #20 coaxial cutting needle, the overall
Long Beach, California, and reported that between incidence of pneumothorax was 42.3% and 1 1 .9%
October 1 983 and December 1 98 8 , there were 1 05 of the patients required a chest tube. Multivariate
cases of iatrogenic pneumothorax in comparison to analysis in this study (25 0) revealed that lesions in
the lower lobe, greater lesion depth, and a needle
trajectory <45° were associated with an increased
TABLE 24. 1 • Lea d i n g Ca uses of
incidence of pneumothorax.
Iatrog e n i c P n e u m ot h o rax in the Veterans
Ad m i n istration Coope rative Study
In one study, the use of a smaller coaxial needle
(# 1 9) in comparison to a larger needle (# 1 8) reduced
Proced u re N u m be r Pe rcenta g e the incidence of pneumothorax from 38% to 23%
(249) . Positioning the patient with the biopsied lung
Tra n s t h o ra c i c need l e 1 28 24
a s p i rati o n
down is not effective (246,25 1 ) , even though it is
S u bc l a vi a n need l e stick 1 19 22 in animals. Similarly, the use of a blood-patch tech
T h o racentesis 1 01 19 nique was ineffective in decreasing the incidence of
Tra n s b ro n c h i a l b i o psy 53 10 pneumothorax in two studies ( 1 1 2, 1 1 3) , but it did
P l e u r a l b i o psy 45 8 decrease the incidence of large pneumothoraces in
Positive- p ressu re 38 7 a third study ( 1 1 4) . The pneumothorax following
ve n t i l a t i o n transthoracic needle biopsy may be delayed. Choi
S u p ra c l avi c u l a r n e e d l e 24 5
et al. (252) reported that a delayed pneumothorax,
sti c k
that is, one not visible on radiograph at 3 hours but
N e rve b l o c k 16 3
M i sce l l a n e o u s 5
which was subsequently visible, developed in 3 .3 % of
458 patients.
C H A PT E R 24 I P N E U M OT H O RAX 385
A preliminary study (253) suggested that the pneumothorax fell dramatically from 1 2% once
incidence of pneumothorax could be reduced with a training programs were initiated (257) . Because more
lung biopsy tract plug. The plug consists of a desic than 1 million subclavian catheters are inserted annu
cated polyethylene glycol hydrogel which is extruded ally in the United States, this procedure is responsible
as a solid cylinder (25 3) . Upon contact with moist for a substantial number of pneumothoraces. In one
tissue, the hydrogel absorbs fluids and expands to recent study from Denmark, only 2 of 473 patients
fill the void created by the needle puncture. It then (0.4%) developed a pneumothorax following cen
absorbs over time. In one randomized multicenter tral venous catheter insertion (262) . The authors of
study with 339 patients, the incidence of pneumo this study recommended that routine postprocedure
thorax was significantly lower ( 1 9%) in the treat radiographs not be obtained unless a complication
ment group than in the control group (3 1 %) (25 3) . was suspected (262) . Maury et al. (263) recommend
There were also significantly fewer chest tubes placed ultrasound rather that chest radiographs to assess the
in the treatment group (4%) than in the control possibility of pneumothorax and the placement of
group ( 1 1 %) . the catheter. Pneumothoraces appear to be more com
Another preliminary study (2 1 6) suggested that mon with internal j ugular cannulations (2.6%) as
the use of fibrin glue as a sealant might decrease the opposed to subclavian cannulations ( 1 .3%) and with
incidence of pneumothorax after lung aspiration. Swan-Ganz catheters as opposed to central venous
In a prospective randomized study in patients with catheters. One recent paper (264) suggests strongly
COPD, 26 patients received 1 mL of fibrin glue that the incidence of pneumothorax can be markedly
as the needle was withdrawn, whereas 32 control reduced if the catheter is inserted using ultrasound
patients received nothing. The incidence of pneu guidance. In this paper (264) , 1 ,948 catheters were
mothorax was 1 9 .2% in the group that received inserted in the internal j ugular for cancer treatment
the fibrin glue compared with 40.6% in the control and there were no pneumothoraces. It is important to
group. In the group that received the fibrin glue, note that the pneumothorax following the insertion
one patient (3.8%) received a chest tube, whereas of a central line may not be apparent on the immedi
six patients ( 1 8%) in the control group received a ate postprocedure radiograph (265 ) .
chest tube (254) . In a third preliminary study (25 5 ) , Thoracentesis i s probably the third leading cause of
1 40 patients were randomized t o receive 2 t o 4 m l of iatrogenic pneumothorax. At present, the incidence
saline into the whole puncture access during extrac of pneumothorax after thoracentesis is approximately
tion of the trocar needle. The incidence of pneumo 6%, with approximately 34% of those with a pneu
thorax with the saline (8%) was significantly less than mothorax receiving a chest tube (266) . The incidence
in the control group (34%) (25 5 ) . It remains to be of pneumothorax is higher if the patient has COPD .
seen which of the above three techniques survives Pneumothorax is more common following therapeu
the test of time. However, the big advantage of the tic than diagnostic thoracentesis (266) . The incidence
saline technique is that its cost is minimal. of pneumothorax can be decreased if the thoracente
It appears that it is safe to travel by air within sis is performed by experienced interventional radi
24 hours of having a transthoracic needle aspiration. ologists or pulmonologist trained in thoracentesis
Tam et al. (25 6) contacted 1 79 patients who under methods with ultrasound guidance. At my previous
went air travel within 14 days of having a transtho institution, the incidence of pneumothorax was 2 . 5 %
racic needle aspiration including 65 who developed a after the performance of 94 1 thoracenteses with
pneumothorax. No patient required in-flight medi ultrasound guidance and only 0 . 8 % received tube
cal attention or flight diversion (256) . Most patients thoracostomy (267) . At the Mayo Clinic, the inci
traveled less than 48 hours after the biopsy (25 6) . dence of pneumothorax was 8.6% in the pulmonary
Th e second leading cause o f iatrogenic pneu outpatient clinic in 200 1 -2002 (268) . This incidence
mothorax is probably the insertion of a central line fell to 1 . 1 % after a program was instituted where all
(24 1 ) . The reported incidence of iatrogenic pneu thoracenteses were performed with ultrasound guid
mothorax following subclavian vein catheterization ance and only by physicians who had undergone a
has varied from 0% to 1 2% , with the average being training program (268) . If tactile fremitus is present
approximately 2% (257-26 1 ) . The importance of over the upper lung field after thoracentesis, if the
monitoring the incidence of complications in indi patient is not symptomatic, and if the physician does
vidual institutions is demonstrated by the report of not suspect a pneumothorax, a chest radiograph after
Lockwood (257) . In this report, the incidence of thoracentesis is not indicated (269) .
3 86 PLE U RAL D I S EASES
Although mechanical ventilation was the leading neoplasms, 1 1 . 3% to 42% (278 ,279) ; transbron
cause of iatrogenic pneumothorax in the 1 970s (270) , chial lung biopsy, 1 % to 2% (280,28 1 ) ; laparoscopy,
it is probably now only the third or fourth lead 0.2% (282) ; cardiac surgery 1 .4% (283) , intercostal
ing cause of iatrogenic pneumothorax. The relative nerve block for fractured rib 5 .6% (284) , and liver
decrease in the incidence of iatrogenic pneumothorax biopsy 0 . 3 5 % (28 5 ) . Bronchoscopy with an ultra
caused by mechanical ventilation is probably due to fine bronchoscopy without transbronchial biopsy
a combination of two factors. First, procedures such has been associated with pneumothorax (286) . In
as transthoracic needle aspiration and subclavian this instance, the ultrathin bronchoscope perforates
vein catheterization were used much less commonly the visceral pleura (286) . The reported incidences are
40 years ago. Second, newer ventilatory modes have probably minimum percentages because the authors
made it possible to ventilate patients with lower peak of articles are usually more experienced in the various
inspiratory pressures and lower mean airway pres procedures they describe than is the average physi
sures. In a series of 5 5 3 patients requiring ventilatory cian. Iatrogenic pneumothorax may occur following
support from nearly 3 5 years ago, the incidence of iat tracheostomy, when air passes into the mediasti
rogenic pneumothorax was 4% (27 1 ) . In this series, num and pleural space via the cervical fascia! planes.
the frequency of pneumothorax was increased if the Iatrogenic pneumothorax also frequently compli
patient had aspiration pneumonia (37%) , COPD cates cardiopulmonary resuscitation. In an autopsy
(8%), intubation of the right main stem bronchus series, 1 2 patients had tension pneumothoraces that
( 1 3%), or treatment with positive end-expiratory were undiagnosed during life, and 9 of these patients
pressure ( 1 5%) (27 1 ) . had undergone cardiopulmonary resuscitation (287) .
Th e incidence o f pneumothorax is relatively high Resuscitation-related rib fractures were found in only
in patients with ARD S . Weg et al. (272) reported that three of the nine patients.
the incidence of pneumothorax was 9.2% in a series Physicians treating heart-lung transplant recipients
of 644 patients with ARD S. Although the occurrence or other patients who have undergone mediastinal
of pneumothoraces in this situation in the past had surgery should be aware of the fact that these patients
been attributed to high inspiratory pressures or mean do not have an intact mediastinum. Because they are
airway pressures, these pressures were very similar in likely to undergo procedures that are associated with
patients with and without pneumothorax in Weg's iatrogenic pneumothorax such as transthoracic needle
series (272) . In this series, the mortality rate was aspiration, bronchoscopy, thoracentesis, and central
not significantly different in those with and without line insertion, they may develop life-threatening bilat
pneumothorax. Boussarsar et al. (273) reviewed the eral pneumothoraces. Paranjpe et al. (288) reported
literature on the relationship between pneumothorax that 1 5 of 72 heart-lung transplant recipients devel
and ARDS and concluded that there was a higher oped iatrogenic pneumothoraces, and the pneumo
incidence of pneumothorax with plateau pressures thoraces were bilateral in six of the patients. Lee et al.
above 35 cm H20 and lung compliances less than (289) reported the development of a contralateral ten
30 mL/cm H20 . The presence of mediastinal emphy sion pneumothorax following the unilateral chest tube
sema may precede the development of the pneumo drainage of bilateral pneumothoraces.
thorax. In one series of 20 patients who developed
a pneumothorax while on mechanical ventilation,
C l i n ical M a n ifestations
previous chest radiographs had shown the presence of
mediastinal emphysema in 10 (50%) (274) . The clinical manifestations of iatrogenic pneumotho
Overall, the incidence of pneumothorax with rax depend both on the patient's condition and on the
severe acute respiratory syndrome (SARS) is relatively initiating procedure. If the pneumothorax occurs as a
low. Sihoe et al. reported that only 1 .7% of 3 5 6 SARS complication of mechanical ventilation, the patient
patients had a pneumothorax (275), but most of their is likely to demonstrate a sudden clinical deteriora
patients did not receive mechanical ventilation. Kao tion. A sensitive indicator of the development of a
et al. (276) reported that 5 of 4 1 patients ( 1 2%) who pneumothorax in such patients is an increasing peak
were treated with mechanical ventilation developed a and plateau pressure on the respirator if the patient
pneumothorax. is on volume-controlled ventilation, or a decreasing
Other procedures associated with iatrogenic pneu tidal volume if the patient is on pressure support. The
mothorax and their approximate incidence are pleu development of a pneumothorax during cardiopul
ral biopsy, 1 0% (277) ; radiofrequncy ablation of lung monary resuscitation is heralded by more difficulty
C H A PT E R 24 I P N E U M OT H O RAX 3 87
in ventilating the patient. In contrast, many patients With mechanical ventilation, positive pressure in the
who develop pneumothorax after thoracentesis, pleu alveoli leads to increased entry of air into the pleural
ral biopsy, transbronchial biopsy, or percutaneous space and the likelihood that a tension pneumothorax
lung aspiration have no symptoms referable to the will develop. The chest tube should be left in place for
pneumothorax. at least 48 hours after the air leak stops if the patient
continues to receive mechanical ventilation. At times,
an ipsilateral recurrent pneumothorax develops in a
D i a g nosis
patient on mechanical ventilation while the chest tube
The diagnosis of iatrogenic pneumothorax should be is still in place. This development is usually due to
suspected in any patient treated by mechanical ventila placement of the chest tube in a fissure (293) . Malpo
tion. The presence of mediastinal emphysema should sitioning of the chest tube is suggested if the chest tube
serve as an indicator to look closely for a pneumotho is perpendicular to the lateral chest wall; the chest tube
rax. Recognition of the pneumothorax in the patient should be relatively parallel to the chest wall (293) .
on mechanical ventilation is more difficult because Bronchopleural fistulas and mechanical ventilation are
the chest radiographs are obtained with the patient discussed later in this chapter.
supine or semisupine. When the patient is in this When an iatrogenic pneumothorax develops after
position, the most superior part of the chest (where a procedure, symptoms vary from none to severe
the air accumulates) is the anterior costophrenic respiratory distress. In general, if the patient has no
sulcus. In one series of 1 1 2 pneumothoraces seen on symptoms or just mild symptoms and the pneumo
supine radiographs, the most common location of air thorax occupies less than 40% of the hemithorax, the
was anteromedial in 38%, followed by subpulmonic patient can be managed with observation. The admin
in 26%, apicolateral in 22%, and posteromedial in istration of supplemental oxygen will increase the rate
1 1 % (290) . Air in the anterior costophrenic sulcus is at which air is absorbed from the pleural space (see
manifested as hyperlucency over the upper abdomi Chapter 2) (7 1 ) . If the patient is more than mildly
nal quadrants (290) . Pneumothoraces are frequently symptomatic, if the pneumothorax occupies more
not recognized on the supine radiographs. Kollef than 40% of the hemithorax, or if the pneumothorax
(29 1 ) prospectively reviewed all 464 medical inten continues to enlarge, however, one should consider
sive care unit admissions at Fitzsimons Army Medical removing the intrapleural air.
Center over a I -year period and reported that 9 of In general, most iatrogenic pneumothoraces should
28 pneumothoraces (32%) were not originally recog first be treated with aspiration. If the initial aspira
nized. Three of these nine patients subsequently went tion is unsuccessful, then a Heimlich valve should be
on to develop a tension pneumothorax. attached to the catheter. Only when the lung does not
The occurrence of an iatrogenic pneumothorax expand and remains expanded with the Heimlich valve
should also be suspected in patients who become more is a larger chest tube inserted (294) . Delius et al. (295)
short of breath after a medical or surgical procedure treated 79 needle-induced iatrogenic pneumothoraces
known to be associated with the development of an by aspiration through an 8-F radiopaque Teflon cath
iatrogenic pneumothorax. The signs and symptoms eter. The initial aspiration was successful in 59 patients
of the pneumothorax are similar to those of primary (75%), and an additional 9 patients ( 1 5%) were suc
and secondary pneumothorax, and the diagnosis is cessfully managed with a Heimlich valve attached to
confirmed by chest radiographs. Ultrasound can also this small catheter (29 5). In another study (296) in
be used to diagnose the pneumothorax (292) . which 1 02 patients were treated with aspiration after
developing a pneumothorax post needle aspiration of
a nodule, only 1 5 required a chest tube. A chest tube
Treatment
was more likely to be necessary if more than 670 ml
The treatment of iatrogenic pneumothorax differs from of pleural air was aspirated (296) . Patients can be
that of spontaneous pneumothorax in that recurrence managed as outpatients with small intrapleural cath
is not likely, and, therefore, one need not try to create eters and Heimlich valves (297) . In one study, 1 7 of
a pleurodesis, as is done frequently with spontaneous 20 patients (85%) who developed pneumothorax
pneumothorax. When a pneumothorax occurs dur after transthoracic needle aspiration and were treated
ing positive-pressure ventilation, tube thoracostomy with manual aspiration did not require a chest tube
should be performed immediately in most cases to (298) . The injection of 1 5 ml blood after aspiration is
prevent the development of a tension pneumothorax. complete may decrease the subsequent need for tube
388 PLE U RAL D I S EASES
thoracostomy (299) . In general, patients are more 5% of multiple trauma patients have a pneumothorax
likely to require a chest tube if they have COPD (300) . and at least 40% of the pneumothoraces are occult
(304) . For example, in one series of 2,048 multiple
trauma patients, there were 90 patients (4.4%) who
TRA U M AT I C { N O N IATRO G E N I C}
had a pneumothorax (305) . Thirty-five of these pneu
P N E U M OT H O RAX
mothoraces (3 8 . 8 %) were occult (305) . In a second
Traumatic pneumothorax can result from either study of 3, 7 1 2 trauma patients, a pneumothorax was
penetrating or nonpenetrating chest trauma. present in 230 (6.2%) and the pneumothorax was
occult in 1 26 of them (54.8%) (306) .
An alternative imaging procedure to diagnose
Mech a n i s m
traumatic pneumothorax is ultrasound. The presence
The mechanism of the pneumothorax is easily under of pneumothorax is characterized by two features:
stood with penetrating chest trauma because the (a) absence of pleural lung sliding and (b) absence
wound allows air to enter the pleural space directly of comet-tail artifacts (307,308). Ultrasound appears
through the chest wall. In addition, the visceral to be better at diagnosing pneumothorax than the
pleura is frequently penetrated, allowing air to enter supine radiograph (309,3 1 0,3 1 1 ) . Soldati et al. (3 1 O)
the pleural space from the alveoli. With nonpen performed supine chest radiographs, ultrasound, and
etrating trauma, the ribs may become fractured or CT scan on 1 86 patients with blunt chest trauma.
dislocated, and the visceral pleura may thereby be lac Using the CT scan as the gold standard, they reported
erated, leading to a pneumothorax. In most patients that ultrasound identified 55 of 56 (98%) pneumo
with pneumothorax secondary to nonpenetrating thoraces while the chest radiograph only identified
trauma, however, no associated rib fractures occur 30 of 56 (54%) (3 1 0) . In a second study, Blaivas
(30 1 ,302) . The mechanism of the pneumothorax in et al. (3 1 1 ) examined supine chest radio graphs and
such patients is thought to be as follows (30 1 ,302) . bedside ultrasound performed by the emergency room
With sudden chest compression, the alveolar pressure physicians for the delineation of pneumothorax in
increases and this may cause alveolar rupture. Air then 1 76 trauma victims of whom 53 had a pneumothorax
enters the interstitial spaces and dissects either toward on CT scan which they used as the gold standard
the visceral pleura or toward the mediastinum to (3 1 1 ) . They reported that the sensitivity and speci
produce mediastinal emphysema. A pneumothorax ficity of ultrasound were 9 8 . 1 % and 99.2%, respec
results when either the visceral or the mediastinal tively, whereas the sensitive and specificity of the
pleura ruptures. supine chest radiograph was 75 . 5 % and 1 00%,
respectively (3 1 1 ) .
I n c i d e nce a n d D i a g nosis
Treatm e n t
The diagnosis of traumatic pneumothorax should
be considered in any patient who suffers significant Most traumatic pneumothoraces should be treated
trauma. In most instances, the initial chest radiograph with tube thoracostomy. Small tubes 1 0 to 14 Fr are
on trauma patients is obtained in the supine position adequate in almost all instances (3 1 2) . If a hemopneu
and small pneumothoraces may not be apparent. These mothorax is present, one chest tube should be placed
supine chest radiographs are insensitive in diagnosing in the superior part of the hemithorax to evacuate the
both pneumothorax and hemothorax. In one series air and another should be placed in the inferior part of
of 1 03 patients with blunt chest trauma, thoracic CT the hemithorax to remove the blood (see Chapter 25) .
scans revealed pneumothorax in 44 patients whereas With traumatic pneumothorax, the lung expands and
supine chest radiographs revealed pneumothorax in the air leak usually ceases within 72 hours. If the lung
only 17 patients (303) . In the same series, the CT does not expand or an air leak persists, VATS should be
scan revealed hemothorax in 44 patients whereas the performed within the first few days to evaluate the rea
supine chest radiograph revealed hemothorax in only son for the air leak (3 1 3) . Carrillo et al. (3 1 4) reported
23 patients (303) . In view of this series, a case can be their results in 1 3 patients who had persistent air
made for obtaining a thoracic CT scan in all severely leaks or unexpanded lung 72 hours post trauma. They
injured patients with blunt chest trauma. obtained chest CT scans and performed bronchoscopy
Pneumothoraces seen only on the CT scan are labe on all patients to ascertain that there were no significant
led as occult pneumothoraces. Overall, approximately problems causing the pneumothorax. At thoracoscopy,
C H A PT E R 24 I P N E U M OT H O RAX 389
the source of the air leak was sought by inspecting the Indeed, the practice guidelines for management of
lung surface from apex to base. If there was no obvi occult pneumothorax states that occult pneumo
ous air leak, 250 mL saline solution was instilled with thoraces subjected to mechanical ventilation may
slight lung ventilation to identify the air leak. Once the be observed (323) . However, if the patient does not
air leak was identified, a topical surgical sealant CoSeal receive a chest tube, they should be observed closely
(Baxter, Freemont, CA) was applied. They did not because of the possibility of the development of a
attempt to surgically close the leak. This procedure was tension pneumothorax, which could be fatal. Supple
successful in all patients in that 1 1 of the 1 3 patients mental oxygen should be administered to facilitate
had their chest tubes removed within 24 hours and the the reabsorption of the pleural air (324) .
remaining 2 patients had their chest tubes removed Whenever a patient with a traumatic pneumotho
without 48 hours (3 14). rax is seen, two uncommon diagnostic possibilities,
Tube thoracostomy may not be necessary for both indications for immediate thoracic operation,
patients with small pneumothoraces or those with should be considered. One is fracture of the trachea
occult pneumothoraces. Knottenbelt and van der Spuy or a major bronchus; the second is traumatic rupture
(3 1 5) observed 333 patients with small ( < 1 . 5 cm of the esophagus. Bronchial rupture should be sus
from lung to chest wall) pneumothoraces due to chest pected in patients with persistent air leak following a
trauma and reported that only 33 ( 1 0%) required traumatic pneumothorax, particularly if there is sub
subsequent drainage for an enlarging pneumothorax. cutaneous emphysema, pneumomediastinum, deep
Ordog et al. (3 1 6) observed 47 patients with small cervical emphysema, hemoptysis, or rib, or clavicu
pneumothoraces (<20%) secondary to stab wounds lar fractures (325 ,326) . The possibiliry of bronchial
of the chest and reported that only 32% required a rupture should be assessed in such patients with
chest tube or showed progression of the pneumothorax fiberoptic bronchoscopy. Thoracic CT scan does not
within 24 hours. They recommend that such patients definitively establish the diagnosis in most patients
be admitted to the hospital and have repeat radio (326) . The treatment of choice is surgical repair.
graphs at 6 hours, 24 hours, and again at 48 hours. Traumatic rupture of the esophagus usually pro
The patients are then discharged if the pneumothorax duces a hydropneumothorax. Therefore, if a patient
is unchanged or shows evidence of resolving (3 1 6) . with a traumatic pneumothorax also has a pleural
Most patients with occult pneumothoraces need effusion, the possibility of esophageal rupture should
not be treated with tube thoracostomy (3 1 7-3 1 9) . be entertained. A reliable screening test for esopha
Wolfman et al. (3 1 9) classified occult pneumothora geal rupture is measurement of the pleural fluid amy
ces as minuscule ( < 1 cm in greatest anteroposterior lase level (327) . If the patient's pleural fluid amylase
thickness and seen on no more than four contigu level is elevated, contrast radiographic studies of the
ous CT images), anterior (> 1 cm but not extend esophagus should be performed.
ing beyond the midcoronal line) , and anterolateral If a patient has suffered a traumatic pneumotho
(extending posteriorly beyond the midcoronal line) . rax, how long should they wait before they travel by
Of the 28 occult pneumothoraces in their series, air? The Aerospace Medicine Association has suggested
6 were minuscule, 14 were anterior, and 8 were that patients should be able to fly 2 to 3 weeks after
anterolateral. The patients with the minuscule and radiologic resolution of the pneumothorax (328) . The
the anterior occult pneumothoraces were less likely following study appears to have validated these recom
to receive tube thoracostomy (3 1 9) . De Moya et al. mendations. Cheatham and Safcsak (329) studied 1 2
(320) developed another system for semiquantitating consecutive patients with recent traumatic pneumo
the size of an occult pneumothorax. They took the thorax who desired to travel by commercial airline. Ten
maximum distance from the chest wall in millimeters patients waited at least 1 4 days and all were asymp
and added 1 0 if the pneumothorax did not go below tomatic in-flight. One of two patients who flew earlier
the hilum and 20 if the pneumothorax went below than 14 days developed respiratory distress in-flight
the hilum (320) . They reported in a series of 1 ,295 with symptoms suggesting a recurrent pneumothorax.
patients with pneumothorax that the average score of
those with chest tubes was 34 while the average score
TRA U M AT I C P N E U M OT H O RAX
for those without chest tubes was 2 1 (320) .
S E C O N DARY TO D R U G A B U S E
Some patients who have an occult pneumotho
rax and receive mechanical ventilation have been Intravenous drug abuse has become endemic i n many
managed without tube thoracostomy (32 1 ,322) . urban areas. It appears that there is a high incidence
390 PLE U RAL D I S EASES
of traumatic pneumothorax in intravenous drug throughout expiration and often during inspiration as
users. Douglass and Levison (330) reviewed 525 well. Most tension pneumothoraces occur in patients
diagnoses of pneumothorax between January 1, 1 9 82 who are receiving positive-pressure ventilation either
and December 3 1 , 1 984 at the Detroit Receiving from mechanical ventilation or during resuscitation. If
Hospital. They reported that 1 1 3 (2 1 . 5%) occurred the patient is not receiving positive-pressure ventilation,
because of drug abuse. The user or a companion had then the mechanism by which a tension pneumothorax
attempted to inject the drug into the subclavian or develops is probably related to some type of one-way
internal j ugular vein. It has been recommended that valve process in which the valve is open during inspi
intravenous drug users with traumatic pneumotho ration and closed during expiration. During inspira
rax be managed with tube thoracostomy (330) . In tion, owing to the action of the respiratory muscles, the
the series of Douglass and Levison (330) , the average pleural pressure becomes negative and air moves from
number of days for chest tube management was 4.4. the alveoli into the pleural space. Then, during expira
It is probable, however, that many such cases could be tion, with the respiratory muscles relaxed, the pleural
managed with simple aspiration. pressure becomes positive. A one-way valve mechanism
must be implicated; otherwise, on expiration, when
the pleural pressure is positive with respect to the
P N E U M OT H O RAX EX VA CUO alveolar pressure, gas would flow from the pleural
Pneumothorax ex vacuo is said to occur when patients space into the alveoli and no positive pressure would
develop a pneumothorax secondary to acute bronchial develop in the pleural space.
obstruction. The theory is that the acute collapse of
the lung results in negative intrapleural pressure, which
leads to the accumulation of gas that originated in the Path ophysiologic Featu res
ambient tissues and blood in the pleural space (33 1 ) . The development of a tension pneumothorax is
Recently, the definition has been altered t o indi usually heralded by a sudden deterioration in the
cate the development of a pneumothorax after a cardiopulmonary status of the patient. The precise
thoracentesis because the lung is unable to reexpand explanation for sudden deterioration is not known,
and fill the pleural space (332) . In one series of 282 but it is probably related to the combination of a
patients undergoing 437 thoracenteses, 10 patients decreased cardiac output due to impaired venous
(4%) developed what the authors called p neumotho return and marked hypoxemia (334) . Older studies
rax ex vacuo (332) . These pneumothoraces tend to in unventilated animals suggested that the primary
persist and the authors recommend that they not be pathophysiologic abnormality was a precipitous fall
treated with chest tubes (332) . in the Pao 2 to below 30 mm Hg (33 5,336) . However,
It is not obvious to me that this entity actually more recent studies in ventilated animals have sug
exists. In order for air to come out of the tissues and gested that the primary problem is decreased cardiac
into the pleural space, the pleural pressure would have output. Carvalho et al. (337) induced right-sided
to be more negative than - 60 cm H 2 0 (see Chap tension pneumothoraces with mean pleural pressures
ter 2) . Heidecker et al. (333) performed manometry in of + 1 0 and then + 2 5 cm H 2 0 in 1 0 mechanically
eight patients who developed an unintentional pneu ventilated adult sheep. The mean cardiac output in
mothorax after a thoracentesis. In none of the cases these animals fell from 3 . 5 L/minute to approxi
were the pleural pressures below - 20 cm H 2 0 (333). mately 1 .2 L/minute, and the mean blood pressure
All theses pneumothoraces were associated with radio fell from 80 mm Hg to less than 50 mm Hg as the
graphic signs of unexpandable lung. These investigators pleural pressure increased from -5 to + 2 5 cm H 2 0
speculated that the pneumothoraces developed from (337) . The decrease in the Pao2 was much less life
transient, parenchymal-pleural fistulae caused by non threatening; the Pao 2 was 1 5 0 mm Hg at baseline
uniform stress distribution over the visceral pleura that and fell to 59 mm Hg when the pleural pressure was
develop during large-volume drainage if the lung can 25 cm H 2 0 (337) . The inspiratory airway pressure
not conform to the shape of the thoracic cavity (333) . nearly doubled from 1 9 to 35 cm H p (337) . How
ever, studies by Barton et al. (338) in ventilated swine
suggested that the fall in the Sao 2 was at least as
TE N S I O N P N E U M OT H O RAX
important as the fall in cardiac output. They mea
A tension pneumothorax is said to be present when sured the cardiac output and the Sao 2 as 1 00 mL
the intrapleural pressure exceeds atmospheric pressure aliquots of air were introduced into the pleural space.
C H A PT E R 24 I P N E U M OT H O RAX 39 1
They found that when the mean intrapleural pressure D i a g nosis a n d Treatm ent
had increased to 1 1 mm Hg with the introduction
The diagnosis of tension pneumothorax should be
of 700 mL of air, the mean cardiac output had fallen
suspected in patients whose condition suddenly dete
from 2 . 8 to 1 . 9 L/minute, the mean arterial pressure
riorates, who are receiving mechanical ventilation and
had fallen only from 90 to 73 mm Hg, but the Sao 2
who have undergone a procedure known to cause a
had fallen from 97% to 5 5 % (33 8) .
pneumothorax. If difficulty is encountered in the ven
In humans, for obvious reasons, there are no
tilation of a patient during cardiopulmonary resusci
systematic studies of the blood gases or the hemo
tation or a patient has electromechanical dissociation,
dynamics associated with tension pneumothorax.
a tension pneumothorax should also be suspected.
Beards and Lipman (339) did report on the hemo
In a series of 3 , 5 00 autopsies, unsuspected tension
dynamics of three patients receiving mechanical ven
pneumothorax was found in 12 patients; 10 of these
tilation who developed tension pneumothorax. In
had been supported by mechanical ventilators, and 9
their three patients, the cardiac indices, which were
had undergone cardiopulmonary resuscitation (287) .
7.3, 4 . 8 , and 3.6, respectively, at baseline, fell to 3.0,
2 There is one report of three cases of tension pneu
3 . 1 , and l .4 l/min/m , respectively, with the develop
mothorax that occurred during hyperbaric oxygen
ment of a tension pneumothorax, whereas the base
therapy for acute carbon monoxide poisoning (340) .
line mean arterial pressures, which were 97, 96, and
The presence of a chest tube in a patient with a pneu
68, respectively, fell to 33, 68, and 57, respectively.
mothorax does not preclude the possibility of a ten
The oxygenation status did not deteriorate nearly as
sion pneumothorax because the chest tube might be
dramatically (339) . The patients in this study did not
malpositioned (342) . There is a report of two cases
have consistent changes in their heart rates (339) . In
of tension pneumothorax that occurred when the
another report, a 67-year-old man with COPD devel
Heimlich valve used for treating pneumothorax was
oped a tension pneumothorax while on mechanical
attached backward (343) .
ventilation. This patient's cardiac output fell from
It is important to assess carefully the chest radio
7. 1 1 to 3 . 8 0 L/minute and the stroke volume fell
graph for pneumothorax in patients who are receiving
from 56 to 27 mL, whereas the pulse increased from
mechanical ventilation. Patients with unrecognized
1 27 to 1 42 (340) . In the same patient, the Pao 2 fell
pneumothoraces who are receiving mechanical ven
from 76 mm Hg to 47 mm Hg with the development
tilation are those most likely to develop a tension
of tension pneumothorax.
pneumothorax. Kollef (29 1 ) reviewed 464 medical
In summary, the disastrous effect of a tension
intensive care unit admissions at Fitzsimons Army
pneumothorax in patients appears to be the result of
Medical Center over a I -year period and reported
the combination of a marked decrease in the cardiac
that 28 patients acquired a pneumothorax during
output and in the Pao2 • In patients, the decrease in
their stay in the intensive care unit. Pneumothorax
the cardiac output is most life threatening, but the
was not originally recognized in nine of the patients,
marked decrease in the Pao 2 should not be ignored.
and three of them (33%) subsequently developed
a tension pneumothorax. In a second series, Tocino
C l i n ical M a n ifestations
et al. (290) reported that a pneumothorax was orig
Although tension pneumothorax occasionally evolves inally missed in 34 of 1 1 2 patients in an intensive
from a spontaneous pneumothorax, it is much more care unit and 1 6 of these 34 patients developed
frequent in patients who develop pneumothorax tension pneumothorax. The diagnosis of pneumo
while receiving mechanical ventilation or during car thorax on the supine chest radiograph is discussed
diopulmonary resuscitation (34 1 ) . The clinical status earlier in this chapter in the section on iatrogenic
of patients with tension pneumothorax is striking. pneumothorax.
The patient appears distressed with rapid labored Tension pneumothorax is a medical emergency.
respirations, cyanosis, and usually profuse diapho Although the diagnosis of tension pneumothorax
resis, hypotension, and marked tachycardia. Arterial can be established radiographically by demonstrating
blood gases reveal marked hypoxemia and, some severe contralateral mediastinal shift and ipsilateral
times, respiratory acidosis. The physical findings diaphragmatic depression, valuable time should not
are those of any large pneumothorax, but in addition, be wasted on radiologic studies because the clinical
the involved hemithorax is larger than the contralat situation and the physical findings are usually suf
eral hemithorax with the interspaces widened. The ficient to establish the diagnosis. When the diagno
trachea is usually shifted toward the contralateral side. sis is suspected, the patient should immediately be
3 92 PLE U RAL D I S EASES
given a high concentration of supplemental oxygen the pulmonary parenchyma distal to a segmental
to combat the hypoxia. Then, the elevated pressure bronchus and the pleural space. Most air leaks are due
in the pleural space must be eliminated. Optimally, to alveolar-pleural fistulas. Bronchopleural fistulas
this is done with a silicon catheter such as that advo occur almost exclusively after pneumonectomy, lobec
cated for thoracentesis (see Chapter 2 8 ) . Ideally, tomy, or segmentectomy and almost always require
the catheter should be attached to a three-way stop reoperation or some type of surgical intervention.
cock and a 5 0-mL syringe partially filled with sterile Alveolar-pleural fistulas rarely require reoperation.
saline solution. After the catheter is inserted into Air leaks occurring concomitantly with spontane
the pleural space, it is connected through the three ous pneumothorax have been discussed earlier in this
way stopcock to the syringe. Then the stopcock is chapter. In this section, the problem of air leaks in
opened to the syringe, and the plunger is withdrawn. patients on mechanical ventilation and in patients
A rush of air bubbling outward through the fluid after pulmonary surgery is discussed.
in the syringe establishes the diagnosis of tension
pneumothorax.
If a tension pneumothorax is confirmed, the cath Class ification of A i r Lea ks
eter should be left in place and in communication
When a patient is seen with an air leak, it is important
with the atmosphere until air ceases to exit through
to semiquantitate the amount of leak. Cerfolio (346)
the syringe. Additional air can be withdrawn from
has developed the following classification that is quite
the pleural space with the syringe and the three-way
useful. In his classification, there are four types of
stopcock. There have been instances reported where
air leaks. The largest and the most uncommon is the
1 6-gauge cannulae were insufficient in the drainage
continuous (C) leak that is present throughout the
of air from tension pneumothorax (344) . If a ten
entire respiratory cycle. There is continuous bubbling
sion pneumothorax is present, preparations should
in the air leak chamber as the patient breathes in and
be made for the immediate insertion of a large
out. These types of leaks are rare and are usually seen
chest tube. If no bubbles escape from the syringe,
only in the patient who is using a ventilator or who
the patient does not have a tension pneumotho
has a bronchopleural fistula (346) . The second larg
rax and the catheter should be withdrawn from the
est type of air leak, which is also uncommon, is an
pleural space.
inspiratory (I) air leak. These leaks are present only
In recent years, some emergency medical services
during inspiration. They are seen almost exclusively
have begun performing needle thoracostomies pre
in the patient who is using a ventilator and has a siz
hospital in patients suspected of having a tension
able alveolar-pleural fistula or a small bronchopleural
pneumothorax (345) . The outpatient use of needle
fistula. The third largest type of air leak, which is very
thoracostomy is controversial, and its utilization
common, is an expiratory (E) air leak. An E leak is
varies markedly from region to region (345) . Warner
present only during expiration. This type of leak is
et al. reviewed 20,330 advanced life support calls
commonly seen after pulmonary surgery and its pres
in the Seattle area and found that 39 had a needle
ence suggests an alveolar-pleural fistula. The smallest
thoracostomy placed for suspected tension pneu
type of air leak is the forced expiratory (FE) leak that
mothorax. Four of these patient had unexpected
is present only when the patient performs a forced
survivals. It appears that if the emergency medical
expiration or coughs. As leaks begin to resolve or heal,
technicians are well trained in needle thoracostomy,
they usually go from an E leak to an FE leak.
its benefits are greater than its risks.
A new method of quantitating air leaks was
reported by Anegg et al. (347) who used a device
B RO N C H O P L E U RA L A N D
(AIRFIX) that would provide digital readouts of the
A LV E O LAR-P LE U RA L F I STU LAS
leak per breath and per minute. They reported that
When a chest tube i s i n place, there will be a n air leak if the leakage volume was less than 1 mL/breath or
through the tube if there is communication between 20 mL/minute, the chest tube could be removed with
a bronchus or the pulmonary parenchyma and the no recurrence of the pneumothorax (347) . If these
pleural space. A bronchopleuralfistula is present when results can be confirmed, use of this device could cer
there is a communication between a mainstem, lobar, tainly diminish the time that chest tubes remain in
or segmental bronchus and the pleural space. An place postoperatively. This paper suggests that the air
alveolar-pleural fistula is a communication between leak need not be zero before the tube is removed.
C H A PT E R 24 I P N E U M OT H O RAX 393
and the incidence of prolonged air leak was reduced bronchial stump. After a pneumonectomy, the early
(2% vs. 1 6%) (363) . Wain et al. (364) demonstrated fistula is massive and persistent. The patient fre
that the application of a synthetic sealant at the end of quently develops massive subcutaneous emphysema
the surgical procedure in 1 89 patients decreased the and may exhibit varying degrees of respiratory insuf
incidence of air leaks from 49% to 1 1 % and decreased ficiency (360) .
the mean time to the last observable air leak from When the bronchial leak occurs later in the post
52.3 to 30.9 hours. operative course (e.g., 7-1 0 days) , it may be caused by
The management of the chest tubes postop failure of healing because of inadequate viable tissue
eratively also influences the duration of air leaks. coverage of the stump, or as the result of infection of
Cerfolio et al. (365) demonstrated that when chest the fluid within the space and rupture of the empyema
tubes were managed with water seal the air leak through the suture line of the bronchial stump. At this
stopped sooner than when they were managed with stage, the patient coughs up variable quantities of sero
suction. However, in a larger study Brunelli et al. sanguineous, frothy fluid from the respiratory tract. The
(366) were unable to confirm these findings. Patients patient should be placed with the affected side down
who had large air leaks on the first postoperative day to decrease the danger of flooding the remaining lung.
were more likely to have prolonged air leaks (365 ) . In When a bronchopleural fistula occurs more than
a meta-analysis of six studies, Deng et al. (367) found 2 weeks after pneumonectomy, it is usually due to
that there was no significant difference in the dura rupture of a frank empyema through the bronchial
tion of air leaks or the incidence of prolonged pneu stump, although, at times, it may be due to failure
mothorax for patients who received suction or no of healing of the bronchial stump (360) . The patient
suction although the trends favored no suction. Inter appears chronically ill with a cough and fever. Thora
estingly Waller et al. (368) demonstrated that flutter centesis reveals that the pleural fluid is infected.
valves were more effective than water-seal systems for The management of a postoperative broncho
the management of postoperative air leaks (368) . Pro pleural fistula depends on the time of its develop
longed air leaks are a very frequent problem after lung ment and its underlying cause. If the bronchopleural
volume reduction surgery. In one study, prolonged air fistula occurs early in the postoperative period, it can
leaks occurred in 35 of 1 97 patients ( 1 7.8%) under sometimes be managed with reoperation and repair
going lung volume reduction surgery, but only three of the bronchial stump. If primary repair of the bron
leaks persisted after 14 days (366) . Shackcloth et al. chial stump is attempted, it is imperative that the new
(369) randomized 20 patients with prolonged air bronchial suture line be covered. This can be done
leaks after lobectomy to receive 1 20 ml autologous with a transposed muscle flap (372) , the pericardia!
blood intrapleurally or nothing and reported that the fat pad, or an omental pedicle flap (360) . With direct
treatment group had a significantly shorter hospital closure, the bronchial stump should be shortened as
stay and duration of chest tube drainage. In general, much as possible (373) .
this treatment results in cessation of the air leak in If primary repair of the bronchopleural fistula is
>70% of patients within 24 hours (369,370) . Since not attempted or is unsuccessful, the patient should
this is such a simple procedure, it should be tried first be treated with a chest tube. Several studies have
in patients with prolonged air leaks. reported on the implantation of different materials in
A bronchopleural fistula is observed in approxi the bronchus through a bronchoscope in an attempt
mately 4% of patients after a pneumonectomy and to close the air leak. When the patient has under
less frequently after lobectomy, segmentectomy, or gone less than a pneumonectomy, a Fogarty balloon
lesser procedure (360) . In patients with lung cancer, catheter is passed down the working channel of the
significant risk factors for the development of a bron bronchoscope, and systematic occlusion of all lung
chopleural fistula include residual carcinomatous tis segments on the side of the air leak is undertaken.
sue at the bronchial stump, preoperative irradiation, The segment or segments leading to the fistula can
and diabetes mellitus (37 1 ) . A bronchopleural fistula be noted by observing decreases or disappearance
is more common after resections for inflammatory of the air leak (374) . Materials placed in the appro
disease of the lung, especially in patients with active priate bronchus to close the fistula have included
tuberculosis and positive sputum cultures (360) . Gelfoam (374) , doxycycline and blood (375), fibrin
After pulmonary surgery, a bronchopleural fistula glue (376,377) , vascular occlusion coils (378) , self
may develop immediately or weeks to months later. expandable stents (379) , and endobronchial one
The early appearance of a fistula (e.g., 1-6 days) is way valves (380) . Fibrin glue appears to be the most
frequently due to a technically poor closure of the promising material. Hollaus et al. (377) applied
C H A PT E R 24 I P N E U M OT H O RAX 395
fibrin glue to 45 patients with bronchopleural fistula 1 7. Smit HJ, Deville WL, Schramel FM, e t al. Atmospheric
after pneumonectomy (40 patients) or lobectomy pressure changes and outdoor temperature changes in relation
to spontaneous pneumothorax. Chest. 1 999; 1 1 6: 676-68 1 .
(5 patients) . They reported that 9 of 29 patients who 1 8 . Suarez-Varela MM, Martinez-Selva MI, Llopis-Gonzalez A,
were treated only endoscopically were cured. Small et al. Spontaneous pneumothorax related with climatic
fistulas ( <3 mm) were particularly likely to respond. characteristics in the Valencia area. Eur J Epidemiol. 2000;
The ultimate place of this approach in the manage 16: 1 93- 1 9 8 .
1 9 . Chen C H , Kou YR, Chen C S , e t al. Seasonal variation i n the
ment of patients with a postoperative bronchopleural
incidence of spontaneous pneumothorax and its association
fistula remains to be determined. with climate: A nationwide population-based study. Respirol
Bronchopleural fistulas that occur late after surgery ogy. 20 1 0; 1 5 :296-302.
are almost always associated with empyema. Such 20. Noppen M, Verbanck S, Harvey J, et al. Music: a new cause
fistulas are discussed in Chapter 1 2 in the section on of primary spontaneous pneumothorax. Thorax. 2004; 5 9 :
722-724.
postpneumonectomy empyema.
2 1 . Withers JN, Fishback ME, Kiehl PV, et al. Spontaneous pneu
mothorax. Am ] Surg. 1 964; 1 0 8 :772-776.
22. Menko FH, van Steensel MA, Giraud S, et al. European
REFERENCES BHD Consortium. Birt-Hogg-Dube syndrome: diagnosis and
management. Lancet Oncol. 2009; 1 0: 1 1 99-1206.
I . Melton LJ, Hepper NGG, Offord KP. Incidence of spontane 23. Khoo SK, Giraud S, Kahnoski K, et al. Clinical and
ous pneumothorax in Olmsted County, Minnesota: 1 9 5 0 to genetic studies of Birt-Hogg-Dube syndrome. J Med Genet.
1 974. Am Rev Respir Dis. 1 979; 1 2 0 : 1 379- 1 382. 2002;39:906-9 1 2 .
2 . Gupta D, Hansell A, Nichols T, et al. Epidemiology of pneu 2 4 . Chie HT, Garcia C K . Familial spontaneous pneumothorax.
mothorax in England. Thorax. 2000; 5 5 :666-67 1 . Curr Opin Pulm Med. 2006; 1 2 :268-272.
3 . Gobbel WG Jr, Rhea WG Jr, Nelson IA, et al. Spontaneous 25. Painter JN, Tapanainen H, Somer M, et al. A 4-bp deletion in
pneumothorax. ] Thorac Cardiovasc Surg. I 963;46:33 1 -34 5 . the Birt-Hogg-Dube Gene (FLCN) causes dominantly inher
4. Lesur 0, Delorme N, Fromaget J M , e t a l . Computed tomog ited spontaneous pneumothorax. Am J Hum Genet. 2005;
raphy in the etiologic assessment of idiopathic spontaneous 76: 5 22-527.
pneumothorax. Chest. 1 990;98:34 1-347. 26. Wei MH, Blake PW, Shevchenko J, et al. The folliculin
5. Bense L, Lewander R, Eklund G, et a!. Nonsmoking, non-alpha mutation database: an online database of mutations associ
1 -antitrypsin deficiency-induced emphysema in nonsmokers ated with Birt-Hogg-Dube syndrome. Hum Mutat. 2009;
with healed spontaneous pneumothorax, identified by com 30:E8 80-E890.
puted tomography of the lungs. Chest. 1 993; I 03:433-43 8 . 27. Zbar B, Alvord WG, Glenn G, et al. Risk of renal and
6. Noppen M, Dekeukeleire T, Hanon S, e t al. Fluoresceine colonic neoplasms and spontaneous pneumothorax in the
enhanced autofluorescence thoracoscopy in primary spon Birt-Hogg-Dube syndrome. Cancer Epidemiol Biomarkers
taneous pneumothorax. Am J Respir Crit Care Med. 2006; Prev. 2002; 1 1 :393-400.
1 74:26-30. 28. Butnor KJ, Guinee DG Jr. Pleuropulmonary pathology of
7. Corrin V, Streichenberger N, Gamondes JP, et al. Respiratory Birt-Hogg-Dube syndrome. Am J Surg Pathol. 2006;30:
bronchiolitis in smokers with spontaneous pneumothorax. 395-399.
Eur Respir J 1 99 8 ; 1 2:702-704. 29. Abolnik IZ, Lossos IS, Gillis D, et al. Primary spontaneous
8 . O'Hara VS . Spontaneous pneumothorax. Mil Med. 1 978; pneumothorax in men. Am ] Med Sci. 1 993;305 :297-303.
1 43:32-3 5 . 30. Abolnik IZ, Lossos IS, Zlotogora J, et al. On the inheritance
9 . Jansveld CAP, Dijkman J H . Primary spontaneous pneumo of primary spontaneous pneumothorax. Am J Med Genet.
chorax and smoking. Br Med] 1 975;4: 5 5 9-560. 1 9 9 1 ;40: 1 5 5- 1 5 8 .
10. Seremetis MG. The management of spontaneous pneumotho 3 1 . Sharpe IK, Ahmad M, Braun W. Familial spontaneous pneu
rax. Chest. 1 970;57:65-68 . mothorax and HLA antigens. Chest. 1 9 80;78:264-268.
1 1 . Bense L, Eklund G, Wiman L G . Smoking and the increased 32. Lenler-Petersen P, Grunnet N, Jespersen TW, et al. Familial
risk of contracting spontaneous pneumothorax. Chest. 1 9 87; spontaneous pneumothorax. Eur Respir J 1 990;3:342-345.
92: 1 009-1 0 1 2. 33. Bense L, Eklung G, Wiman LG. Bilateral bronchial anomaly.
1 2 . Cheng YL, Huang TW, Lin CK, et al. The impact of smoking A pathogenetic factor in spontaneous pneumothorax. Am Rev
in primary spontaneous pneumothorax. J Thorac Cardiovasc Respir Dis. 1 992; 1 46: 5 1 3-5 1 6 .
Surg. 2009; 1 3 8 : 1 92- 1 9 5 . 3 4 . Luna E , Tomashefski JF Jr, Brown D, e t al. Reactive eosino
1 3 . Lee S C , Cheng YL , Huang CW, e t al. Simultaneous bilateral philic pulmonary vascular infiltration in patients with sponta
primary spontaneous pneumochorax. Respirology. 2008 ; 1 3 : neous pneumothorax. Am ] Surg Pathol. 1 994; 1 8 : 1 9 5-1 99.
1 45-148. 3 5 . Cyr PV, Vincic L, Kay JM. Pulmonary vasculopathy in idio
1 4 . Ho KK, Ong ME, Koh MS, et al. A randomized controlled pathic spontaneous pneumothorax in young subjects. Arch
trial comparing minichest tube and needle aspiration in out Pathol Lab Med. 2000; 1 24:7 1 7-720.
patient management of primary spontaneous pneumochorax. 36. Vail WJ, Alway AE, England NJ. Spontaneous pneumochorax.
Am ] Emerg Med. 20 1 1 ;29: 1 1 5 2-1 1 57. Dis Chest. 1 960;38 : 5 1 2-5 1 5 .
1 5 . Scott GC, Berger R, McKean HE. The role of atmospheric 37. Maeda A, Ishioka S, Yoshihara M , e t al. Primary spontane
pressure variation in the development of spontaneous pneu ous pneumothorax detected during a medical checkup. Chest.
mothoraces. Am Rev Respir Dis. 1 989; 1 39:659-662. 1 999; 1 1 6: 847-848.
16. Bense L. Spontaneous pneumothorax related to falls in atmo 3 8 . Aston SJ, Rosove M . Homer's syndrome occurring with spon
spheric pressure. EurJ Respir Dis. 1 98 5 ; 6 5 :544-546. taneous pneumothorax. N Engl] Med. 1 972;287: 1 09 8 .
396 PLE U RAL D I S EASES
3 9 . Bense L, Wiman LG, Hedenstierna G . Onset of symptoms in 5 9 . H o i K , Turchin B, Kelly AM . How accurate i s the Light
spontaneous pneumothorax: correlations to physical activity. index for estimating pneumothorax size? Australas Radio!.
Eur J Respir Dis. 1 9 87;7 1 : 1 8 1 - 1 86. 2007;5 1 : 1 9 6- 1 9 8 .
40. Krenke R, Nasilowski J, Przybylowski T, et al. Electrocardio 6 0 . Cai W, Tabbara M, Takata N, e t al. MDCT fo r automated
gtaphic changes in patients with spontaneous pneumothorax. detection and measurement of pneumothoraces in trauma
J Physiol Pharmacol. 2008;59 Suppl 6:36 1 -373. patients. AJR Am J Roentgenol. 2009; 1 92:830-836.
4 1 . Walston A, Brewer DL, Kitchens CS, et al. The electrocar 6 1 . MacDuff A, Arnold A, Harvey J; BTS Pleural Disease Guide
diographic manifestations of spontaneous left pneumothorax. line Group. Management of spontaneous pneumothorax:
Ann Intern Med. 1 974;80 :375-379 . British Thoracic Society Pleural Disease Guideline 20 1 0 .
42. Alikhan M, Biddison JH. Electrocardiographic changes with Thorax. 2 0 1 0;65 (suppl 2) :ii l 8-ii3 1 .
right-sided pneumothorax. South Med]. 1 998;9 1 :677-680. 62. Baumann MH, Strange C , Heffner JE, et al. Management
43. Strizik B , Forman R. New ECG changes associated with a of spontaneous pneumothorax: an American College of
tension pneumothorax. Chest. 1 999; 1 1 5 : 1 742-1 744 . Chest Physicians Delphi Consensus Statement. Chest. 200 1 ;
44. Janssens U, Koch KC, Graf J, et al. Severe transmyocardial 1 1 9 : 5 90-602.
ischemia in a patient with tension pneumothorax. Crit Care 63. Sadikot RT, Greene T, Meadows K, et al. Recurrence of pri
Med. 2000;28 : 1 63 8- 1 64 1 . mary spontaneous pneumothorax. Thorax. 1 997;52:8 05-809.
4 5 . O'Connor AR, Morgan WE. Radiological review o f pneumo 64. Lippert HL, Lund 0, Blegvad S, et al. Independent risk fac
thorax. BM]. 2005;330 : 1 493- 1 497. tors for cumulative recurrence rate after first spontaneous
46. Seow A, Kazerooni EA, Pernicano PG, et al. Comparison pneumothorax. Eur Respir f. 1 9 9 1 ;4:324-33 1 .
of upright inspiratory and expiratory chest radiographs for 6 5 . Light RW, O'Hara VS, Moritz TE, et al. lntrapleural tetracy
detecting pneumothoraces. AJR Am J Roentgenol. 1 996; cline for the prevention of recurrent spontaneous pneumotho
1 66:3 1 3-3 1 6. rax. JAMA. l 990;264:2224-2230.
47. Dulchavsky SA, Hamilton DR, Diebel LN, et al. Thoracic ultra 66. Guo Y, Xie C, Rodriguez RM, et al. Factors related to recurrence
sound diagnosis of pneumothorax. J Trauma. l 999;47:970-971 . of spontaneous pneumothorax. Respirology. 2005 ; 1 0 :378-384.
4 8 . Harte S, Casey RG, Mannion D, et al. When is a pneumotho 67. Ganesalingam R, O'Neil RA, Shadbolt B, et al. Radiological
rax not a pneumothorax? J Pediatr Surg. 2005;40 : 5 8 6-587. predictors of recurrent primary spontaneous pneumothorax
49. Mitlehner W, Friedrich M, Dissmann W. Value of computer following non-surgical management. Heart Lung Circ. 20 1 0 ;
tomography in the detection of bullae and blebs in patients 1 9 :606-6 1 0 .
with primary spontaneous pneumothorax. Respiration. 1 992; 6 8 . Kircher LT Jr, Swartzel RL. Spontaneous pneumothorax and
59:22 1 -227. its treatment. JAMA. 1 9 54; 1 5 5 :24-29.
50. Smit HJ, Wienk MA, Schreurs AJ, et al. Do bullae indicate 69. Kelly AM, Loy J, Tsang AY, et al. Estimating the rate of
a predisposition to recurrent pneumothorax? Br J Radio!. re-expansion of spontaneous pneumothorax by a formula
2000;73 :356-3 5 9 . derived from com pured tomography volumetry studies. Emerg
5 1 . Martinez-Ramos D, Angel-Yepes V, Escrig-Sos J, e t al. Use Medj. 2006;23:780-782.
fulness of computed tomography in determining risk of 70. Chernick V, Avery ME. Spontaneous alveolar rupture at birth.
recurrence after a first episode of primary spontaneous pneu Pediatrics. l 963;32: 8 1 6-824.
mothorax: therapeutic implications. Arch Bronconeumol. 7 1 . Northfield TC. Oxygen therapy for spontaneous pneumotho
2007;43 :304-308. rax. Br Med]. 1 97 1 ;4:86-8 8 .
52. Smit HJ, van den Heuvel MM, Barbierato SB, et al. 7 2 . Noppen M , Alexander P, Driesen P, e t al. Manual aspiration
Analysis of pleural fluid in idiopathic spontaneous pneu versus chest tube drainage in first episodes of primary sponta
mothorax; correlation of eosinophil percentage with the neous pneumothotax: a mulricenter, prospective, randomized
duration of air in the pleural space. Respir Med. 1 999; pilot study. Am] Respir Crit Care Med. 2002; 1 6 5 : 1 240- 1 244.
93:262-267. 73. Light RW. Manual aspiration: the preferred method for
53. Kalomenidis I, Moschos C, Kollintza A, et al. Pneumothorax managing primary spontaneous pneumothorax? Am J Respir
associated pleural eosinophilia is rumour necrosis factor Crit Care Med. 2002; 1 6 5 : 1 202- 1 203.
alpha-dependent and attenuated by steroids. Respirology. 74. Ayed AK, Chandrasekaran C, Sukumar M . Aspiration versus
2008; 1 3 :73-78. rube drainage in primary spontaneous pneumothorax: a ran
54. Kakaris S, Athanassiadi K, Vassilikos K, et al. Spontaneous domised study. Eur Respir f. 2006;27:477-482.
hemopneumothorax: a rare bur life-threatening entity. Eur J 7 5 . Devanand A, Koh MS, Ong TH, et al. Simple aspiration
Cardiothorac Surg. 2004;2 5 : 8 5 6-8 5 8 . versus chest-rube insertion in the management of primary
5 5 . Nop pen M, Alexander P, Driesen P, e t al. Quantification o f the spontaneous pneumothorax: a systematic review. Respir Med.
size of primary spontaneous pneumothorax: accuracy of the 2004;98:579-5 90.
Light index. Respiration. 200 1 ;6 8 :396-399. 76. Chen JS, Tsai KT, Hsu HH, et al. lntrapleural minocycline fol
56. Collins CD, Lopez A, Mathie A, et al. Quantification of lowing simple aspiration for initial treatment of primary spon
pneumothorax size on chest radiographs using interpleural taneous pneumothorax. Respir Med. 200 8; 1 02 : 1 004- 1 0 1 0 .
distances: regression analysis based on volume measurements 77. Rawlins R, Brown KM, Carr CS, et al. Life threatening
from helical CT. AJR Am f Roentgenol. 1 99 5 ; 1 65 : 1 1 27- 1 1 30. hemorrhage after anterior needle aspiration of pneumo
57. Rhea JT, Deluca SA, Green RE. Determining the size of thoraces. A role for lateral needle aspiration in emergency
pneumothorax in the upright patient. Radiology. 1 982; decompression of spontaneous pneumothorax. Emerg Med f.
1 44:733-736. 2003;20 :383-3 84.
5 8 . Kelly AM, Weldon D, Tsang AY, et al. Comparison berween 78. Chen JS, Hsu HH, Tsai KT, et al. Salvage for unsuccessful
rwo methods for estimating pneumothorax size from chest aspiration of primary pneumothorax: Thoracoscopic surgery
X-rays. Respir Med. 2006; 1 0 0 : 1 3 5 6- 1 35 9 . or chest tube drainage? Ann 1horac Surg. 2008;85 : 1 908- 1 9 1 3 .
C H A PT E R 24 I P N E U M OT H O RAX 397
79. So SY, Yu DYC. Catheter drainage of spontaneous pneumo 1 0 1 . Xie C, Teixeira LR, McGovern JP, e t al. Systemic cortico
thorax: suction or no suction, early or late removal. Thorax. steroids decrease the effectiveness of talc pleurodesis. Am ]
1 9 82;37:46-48. Respir Crit Care Med. 1 99 8 ; 1 57: 1 44 1 - 1 444.
80. Minami H, Saka H, Senda K, et al. Small caliber catheter 1 02 . Lardinois D, Vogt P, Yang L, et al. Non-steroidal anti
drainage for spontaneous pneumothorax. Am J Med Sci. inflammatory drugs decrease the quality of pleurodesis
1 992;404: 345-347. after mechanical pleural abrasion. Eur J Cardiothorac Surg.
8 1 . Vedam H, Barnes DJ . Comparison of large- and small-bore 2004;25 : 865-87 1 .
intercostal catheters in the management of spontaneous 1 03 . Light RW, Wang N-S, Sassoon CSH, et al. Comparison of
pneumothorax. Intern Med J 2003;33 :49 5-499. the effectiveness of tetracycline and minocycline as pleural
82. Liu CM, Hang LW, Chen WK, et al. Pigtail tube drainage sclerosing agents in rabbits. Chest. 1 9 94; 1 06:577-5 82.
in the treatment of spontaneous pneumothorax. Am J Emerg 1 04 . Wu W, Teixeira LR, Light RW. Doxycycline pleurodesis in
Med. 2003;2 1 :24 1 -244. rabbits. Comparison of results with and without chest tube.
83. Light RW Pleural controversy: optimal chest tube size for Chest. 1 998 ; 1 1 4: 5 63-568.
drainage. Respirology. 2 0 1 1 ; 1 6 :244-248. 105. Vargas FS, Wang N-S, Lee HM, et al. Effectiveness of bleo
84. Ponn RB, Silverman HJ, Federico JA. Outpatient chest tube mycin in comparison to tetracycline as pleural sclerosing
management. Ann Thorac Surg. 1 997;64: 1 437- 1 440. agent in rabbits. Chest. 1 993; 1 04: 1 5 82- 1 5 84.
85. Dernevik L, Roberts D , Hamraz B, et al. Management of 1 06. Sherman S, Ravikrishnan KP, Patel AS, et al. Optimum
pneumothorax with a mini-drain in ambulatory and hospi anesthesia with intrapleural lidocaine during chemical
talized patients. Scand Cardiovasc] 2003;37: 1 72- 1 76. pleurodesis with tetracycline. Chest. 1 9 88;93:533-536.
86. Sharma TN, Agnihotri SP, Jain NK, et al. Intercostal tube 1 07. Xie C, Teixeira LR, McGovern JP, et al. Effect of pneumo
thoracostomy in pneumothorax: facrors influencing re thorax on pleurodesis induced with talc in rabbits. Chest.
expansion of lung. Indian J Chest Dis Allied Sci. 1 9 88; 1 998; 1 1 4: 1 143- 1 146.
30:32-3 5 . 108. Wang YT, Ng KY, Poh SC. Intrapleural tetracycline for
8 7 . Shaw TJ, Caterine J M . Recurrent re-expansion pulmonary spontaneous pneumothorax with persistent air leak.
edema. Chest. 1 984;86:784-786. Singapore Medf 1 9 88;29 :72-73 .
8 8 . Kim YK, Kim H , Lee CC, et al. New classification and clini 1 09 . Chambers A, Routledge T, Bille A, et al. Is blood pleurodesis
cal characteristics of reexpansion pulmonary edema after effective for determining the cessation of persistent air leak?
treatment of spontaneous pneumothorax. Am ] Emerg Med. Interact Cardiovasc Thorac Surg. 20 1 O; 1 1 :468-472.
2009;27: 9 6 1 -967. 1 1 0 . Manley K, Coonar A, Wells F, et al. Blood patch for persis
89. Lizotte PE, Whitlock WL, Prudhomme JC, et al. Tension tent air leak: a review of the current literature. Curr Opin
pneumothorax complicating small-caliber chest tube inser Pulm Med. 20 1 2 ; 1 3 :333-338.
tion. Chest. 1 990;97:759-760 . 1 1 1 . Cagirici U, Sahin B, Cakan A, et al. Autologous blood patch
90. Baumann MH, Strange C, Heffner J E , e t a l . Management pleurodesis in spontaneous pneumothorax with persistent air
of spontaneous pneumothorax: an ACCP Delphi Consensus leak. Scand Cardiovasc] 1 998;32:75-78.
Statement. Chest. 200 1 ; 1 1 9 : 5 90-602. 1 1 2 . Bourgouin PM, Shepard JA, McLoud TC, et al. Transtho
9 1 . Larrieu AJ, Tyers GFO, Williams EH, et al. Intrapleural racic needle aspiration biopsy: evaluation of the blood patch
instillarion of quinacrine for rreatment of recurrent sponta technique. Radiology. 1 98 8 ; 1 66:93-9 5 .
neous pneumothorax. Ann Thorac Surg. 1 979;28: 1 46- 1 5 0 . 1 1 3 . Herman SJ, Weisbrod G L . Usefulness of the blood patch
92. Almind M, Lange P, Viskum K. Spontaneous pneumo technique after transthoracic needle aspiration biopsy.
thorax: comparison of simple drainage, talc pleurodesis, Radiology. 1 990; 1 76:39 5-397.
and tetracycline pleurodesis. Thorax. 1 9 89;44:627-630 . 1 1 4 . Lang EK, Ghavami R, Schreiner VC, et al. Autologous blood
9 3 . Ofoegbu RO . Pleurodesis for spontaneous pneumothorax: clot seal to prevent pneumothorax at CT-guided lung biopsy.
experience with intrapleural olive oil in high risk patients. Radiology. 2000;2 1 6:93-96.
Am ] Surg. 1 980; 1 40:679-68 1 . 1 1 5 . Cobanoglu U, Melek M , Edirne Y. Aurologous blood
94. Goldszer RC, Bennett ] , VanCampen J , et al. Intrapleural pleurodesis: A good choice in patients with persistent air
tetracycline for spontaneous pneumothorax. JAMA. 1 979; leak. Ann Thorac Med. 2009;4: 1 82- 1 86.
24 1 :724-725 . 1 1 6. Kinoshita T, Miyoshi S, Katoh M, et al. Intrapleural admin
9 5 . Spector ML, Stern RC. Pneumothorax i n cystic fibrosis: a istration of a large amount of diluted fibrin glue for intrac
26-year experience. Ann Thorac Surg. 1 9 89;47:204-207. table pneumothorax. Chest. 2000; 1 1 7:790-79 5 .
96. Light RW. Diseases of the pleura: the use of talc for 1 1 7 . Loddenkemper R . Thoracoscopy-state o f the art. Eur Respir
pleurodesis. Curr Opinion. 2000;6:255-2 5 8 . J 1 99 8 ; 1 1 : 2 1 3-22 1 .
9 7 . Dresler C M , Olak J, Herndon JE I I , e t al. Phase III inter 1 1 8 . Pompeo E, Tacconi F, Mineo D , et al. The role of awake
group study of talc poudrage vs talc slurry sclerosis for malig video-assisted thoracoscopic surgery in spontaneous pneu
nant pleural effusion. Chest. 200 5 ; 1 27:909-9 1 5 . mothorax. ] Thorac Cardiovasc Surg. 2007; 1 33:786-790.
9 8 . Milton R , Cale AR. Chronic pain due t o talc pleurodesis 1 1 9 . Tschopp JM, Brutsche M, Frey JG. Treatment of complicated
for spontaneous pneumothorax. Ann Thorac Surg. 2003; spontaneous pneumothorax by simple talc pleurodesis under
76: 1 740- 1 74 1 . thoracoscopy and local anaesthesia. Thorax. 1 997;52:329-332.
9 9 . Alfageme I , Moreno L , Huetas C , e t al. Spontaneous pneu 120. Tschopp JM, Boutin C, Astoul P, et al. Talcage by medical
mothorax. Long-term results with tetracycline pleurodesis. thoracoscopy for primary spontaneous pneumothorax is
Chest. 1 994; 1 06:347-350. more cost-effective than drainage: a randomised study. Eur
1 00. Teixeira LR, Wu W, Chang DS, et al. The effect of cortico Respir J 2002;20 : 1 003- 1 009.
steroids on pleurodesis induced by doxycycline in rabbits. 1 2 1 . Noppen M . Management of primary spontaneous pneumo
Chest. 2002; 1 2 1 :2 1 6-2 1 9 . thorax. Curr Opin Pulm Med. 2003;9 :272-275 .
398 PLE U RAL D I S EASES
1 22. Bridevaux PO, Tschopp JM, Cardillo G, et al. Short term 1 4 1 . Hwong TM, Ng CS, Lee TW; et al. Video-assisted thoracic
safety of thoracoscopic talc pleurodesis for recurrent primary surgery for primary spontaneous hemopneumothorax. Eur]
spontaneous pneumothorax. Eur RespirJ 2 0 1 1 ;38:770-773. Cardiothorac Surg. 2004;26: 8 93-896.
1 23 . Yim AP, Ng CS. Thoracoscopy in the management of pneu 1 42. Doddoli C, Barlesi F, Fraticelli A, et al. Video-assisted tho
mothorax. Curr Opin Pulm Med. 200 1 ;7:2 1 0-2 1 4 . racoscopic management of recurrent primary spontaneous
1 24. Hazelrigg S R , Landreneau RJ, Mack M, e t al. Thoracoscopic pneumothorax after prior talc pleurodesis: a feasible, safe
stapled resection for spontaneous pneumothorax. Thorac and efficient treatment option. Eur J Cardiothorac Surg.
Cardiovasc Surg. 1 993; 1 0 5 :389-393. 2004;26: 8 8 9-892.
125. Takeno Y. Thoracoscopic treatment of spontaneous pneumo 1 43 . Cardillo G, Facciolo F, Regal M, et al. Recurrences following
thorax. Ann Thorac Surg. l 993;56:688-690. videothoracoscopic treatment of primary spontaneous
1 26. Inderbitzi RGC, Leiser A, Furrer M, et al. Three years' experi pneumthorax: the role of redo-videothoracoscopy. Eur J
ence in video-assisted thoracic surgery (VATS) for spontaneous Cardiothorac Surg. 200 1 ; 1 9 : 396-399.
pneumothorax. Thorac Cardiovasc Surg. 1 994; 1 07: 1 4 1 0-14 1 5 . 1 44. Akiba T, Marushima H, Kobayashi S, et al. Video-assisted
1 27. Czerny M, Salat A , Fleck T, e t al. Lung wedge resection thoracic surgery for recurrent primary spontaneous pneumo
improves outcome in stage I primary spontaneous pneumo thorax in reoperated chests. Surg Today. 2009;39: 944-946.
thorax. Ann Thorac Surg. 2004;77 : 1 802- 1 8 0 5 . 1 4 5 . Chen JS, Hsu HH, Kuo SW, et al. Management of recur
1 2 8 . Loubani M, Lynch V. Video assisted thoracoscopic bullec rent primary spontaneous pneumothorax after thoracoscopic
tomy and acromycin pleurodesis: an effective treatment for surgery: should observation, drainage, redo thoracoscopy, or
spontaneous pneumothorax. Respir Med. 2000;94 : 8 8 8-890. thoracotomy be used? Surg Endosc. 2009 ; 1 4 :462-463.
1 29 . Horio H, Nomori H, Kobayashi R, et al. Impact of addi 1 46. Radberg G, Dernevik L, Svanvik J, et al. A comparative
tional pleurodesis in video-assisted thoracoscopic bullectomy retrospective study of thoracoscopy versus thoracotomy for
for primary spontaneous pneumothorax. Surg Endosc. 2002; the treatment of spontaneous pneumothorax. Surg Laparosc
1 6:630-634. Endosc. 1 99 5 ; 5 : 90-93 .
1 30. Bobbio A, Ampollini L, lnternullo E, et al. Thoracoscopic 1 47. Hyland MJ, Ashrafi AS, Crepeau A , e t al. I s video-assisted
parietal pleural argon beam coagulation versus pleural abra thoracoscopic surgery superior to limited axillary thora
sion in the treatment of primary spontaneous pneumotho cotomy in the management of spontaneous pneumothorax?
rax. EurJ Cardiothorac Surg. 2006;29 :6-8. Can Respir J 200 1 ; 8 :339-343.
1 3 1 . Shaikhrezai K, Thompson AI, Parkin C, et al. Video 1 4 8 . Barker A, Maratos EC, Edmonds L, et al. Recurrence rates
assisted thoracoscopic surgery management of spontaneous of video-assisted thoracoscopic versus open surgery in the
pneumothorax-long-term results. Eur J Cardiothorac Surg. prevention of recurrent pneumothoraces: a systematic
20 1 1 ;40: 1 20- 1 23 . review of randomised and non-randomised trials. Lancet.
1 32. Marcheix B, Brouchet L, Renaud C, e t a l . Videothoraco 2007;370:329-335.
scopic silver nitrate pleurodesis for primary spontaneous 149. Dusmet M, Corpataux JM. The axillary minithoracotomy
pneumothorax: an alternative to pleurectomy and pleural is a cost-effective alternative to VATS for bullectomy in
abrasion? EurJ Cardiothorac Surg. 2007;3 1 : 1 1 06-1 1 07. recurrent pneumothorax. Am J Respir Dis Crit Care Med.
1 33 . Yim AP, Liu HP. Video assisted thoracoscopic manage 2000; 1 6 l :A268.
ment of primary spontaneous pneumothorax. Surg Laparosc 1 5 0. Deslauriers J, Beaulieu M, Despres J-P, et al. Transaxillary
Endosc. 1 997;7:236-240. pleurectomy for treatment of spontaneous pneumothorax.
1 34. Cardillo G, Facciolo F, Giunti R, et al. Videothoracoscopic Ann Thorac Surg. 1 980;30 : 5 69-574.
treatment of primary spontaneous pneumothorax: a 6-year 1 5 1 . Baumann MH. Pneumothorax and air travel: lessons learned
experience. Ann Thorac Surg. 2000;69:357-36 1 . from a bag of chips. Chest. 2009 ; 1 36:655-566.
1 3 5 . Waller DA. Video-assisted thoracoscopic surgery for sponta 1 52. Currie GP, Kennedy AM, Paterson E, et al. A chronic pneu
neous pneumothorax-a 7-year learning experience. Ann R mothorax and fitness ro By. Thorax. 2007;62 : 1 87- 1 8 9 .
Coll Surg Engl. 1 999;8 1 :3 87-392. 1 53 . Chambers A , Scarci M . In patients with first-episode primary
1 36. Bertrand PC, Regnard JF, Spaggiari L, et al. Immediate spontaneous pneumothorax is video-assisted thoracoscopic
and long-term results after surgical treatment of primary surgery superior to tube thoracostomy alone in terms of time
spontaneous pneumothorax by VATS. Ann Thorac Surg. to resolution of pneumothorax and incidence of recurrence?
1 996;6 1 : 1 64 1 - 1 64 5 . Interact Cardiovasc Thorac Surg. 2009;9: 1 003- 1 006.
1 37. Berrisford RG, Page RD . Video assisted thoracic surgery 1 54. Weissberg D , Refaely Y. Pneumothorax. Chest. 2000;
for spontaneous pneumothorax. Thorax. 1 996;5 1 (suppl 2): 1 1 7: 1 279- 1 28 5 .
S23-S28. 1 5 5 . Dines D E , Clagett OT, Payne W S . Spontaneous pneumo
138. Margolis M, Gharagozloo F, Tempesta B , et al. Video thorax in emphysema. Mayo Clin Proc. 1 970;45 :48 1 -487.
assisted thoracic surgical treatment of initial spontaneous 1 56. Tanaka F, Itoh M, Esaki H, et al. Secondary spontaneous
pneumothorax in young patients. Ann Thorac Surg. 2003; pneumothorax. Ann Thorac Surg. 1 99 3 ; 5 5 : 372-376.
76: 1 6 6 1 - 1 663. 1 57. Shields TW, Oilschlager GA. Spontaneous pneumothorax in
139. Cardillo G, Carleo F, Giunti R, et al. Videothoracoscopic patients 40 years of age and older. Ann Thorac Surg. 1 966;
talc poudrage in primary spontaneous pneumothorax: a 2:377-383.
single-institution experience in 8 6 1 cases. ] Thorac Cardiovasc 1 5 8 . George RB, Herbert SJ, Shames JM, et al. Pneumothorax
Surg. 2006; 1 3 1 :322-328. complicating pulmonary emphysema. JAMA. 1 975;234:
1 40. Chen JS, Hsu HH, Chen RJ, et al. Additional minocy 389-393.
cline pleurodesis afi:er thoracoscopic surgery for primary 1 5 9 . Slater A, Goodwin M, Anderson KE, et al. COPD can
spontaneous pneumothorax. Am J Respir Crit Care Med. mimic the appearance of pneumothorax on thoracic ultra
2006 ; 1 73 : 548-5 54. sound. Chest. 2006; 1 2 9 : 545-5 5 0 .
C H A PT E R 24 I P N E U M OT H O RAX 399
1 60. Waseem M, Jones J, Brutus S, et al. Giant bulla mimicking pneumonia and pulmonary tuberculosis. Eur Respir ].
pneumothorax. J Emerg Med. 2005;29: 1 5 5- 1 5 8 . 1 997; 1 0 : 1 332-1 335.
1 6 1 . Phillips G D , Trotman-Dickenson B, Hodson ME, e t al. Role 1 83 . Leoung GS, Feigal DW Jr, Montgomery AB, et al. Aerosol
of CT in the management of pneumothorax in patients with ized pentamidine for prophylaxis against Pneumocystis carinii
complex cystic lung disease. Chest. 1 997; 1 1 2 :275-278 . pneumonia. N Engl] Med. 1 990;323:769-775 .
1 62 . Fraser RS , Muller N L , Colman N, e t al. Diagnosis ofDiseases 1 84 . Renzi P M , Corbeil C , Chasse M, e t al. Bilateral pneumotho
of the Chest, Vol 4, 4th ed. Philadelphia, PA: WB Saunders; races hasten mortality in AIDS patients receiving secondary
2000:278 1 -2794. prophylaxis with aerosolized pentamidine. Association with
1 63 . Videm V, Pillgram-Larsen J, Ellingsen 0, et al. Spontaneous a lower DLco prior to receiving aerosolized pentamidine.
pneumothorax in chronic obstructive pulmonary disease: Chest. 1 992; 1 02:49 1 -496.
complications, treatment, and recurrences. Eur J Respir Dis. 185. Cuthbert AC, Wright D, McVerry BA. Pneumothorax in
1 9 87;7 1 :365-37 1 . pentamidine-treated haemophiliacs [Letter] . Lancet. 1 9 9 1 ;
1 64. Ng AW, Chan KW, Lee SK. Simple aspiration of pneumo 337: 9 1 8 .
thorax. Singapore Med]. 1 994;3 5 : 5 0-52. 1 86. Rivero A , Perez-Camacho I, Lozano F, e t al; Andalusian
1 6 5 . Searon D , Yoganathan K, Coady T, et al. Spontaneous pneu Group for the Study of lnfectious Diseases (GAEi) . Etiology
mothorax: marker gas technique for predicting outcome of of spontaneous pneumothorax in 1 0 5 HIV-infected patients
manual aspiration. Br Med]. 1 9 9 1 ;302:262-265 . without highly active antiretroviral therapy. Eur J Radio!.
1 66. Chen C H , Liao WC, Liu YH, e t al. Secondary spontane 2009;7 1 :264-268.
ous pneumothorax: which associated condition benefit from 1 87. Newsome GS, Ward DJ, Pierce PF. Spontaneous pneumo
pigtail catheter treatment? Am J Emerg Med. 2 0 1 0;30:45-50. thorax in patients with acquired immunodeficiency syn
1 67. Wait MA, Escrera A. Changing clinical spectrum of sponta drome treated with prophylactic aerosolized pentamidine.
neous pneumothorax. Am J Surg. 1 992; 1 64:528-5 3 1 . Arch Intern Med. 1 990; 1 5 0:2 1 67-2 1 68 .
1 6 8 . Bourgouin P, Cousineau G , Lemire P, et al. Computed 188. Shanley DJ, Luyckx BA, Haggerty MF, e t al. Spontaneous
tomography used to exclude pneumothorax in bullous lung pneumothorax in AIDS patients with recurrent Pneumocystis
disease. J Can Assoc Radio!. 1 9 8 5 ;36:34 1 -342. carinii pneumonia despite aerosolized pentamidine prophy
1 69 . Onuki T, Murasugi M, Ikeda T, et al. Thoracoscopic surgery laxis. Chest. 1 9 9 1 ;99: 5 02-5 04.
for pneumothorax in older patients. Surg Endosc. 2002; 1 89. Scannell KA. Pneumothoraces and Pneumocystis carinii
1 6: 3 5 5-357. pneumonia in two AIDS patients receiving aerosolized pent
1 70. Lee P, Yap WS, Pek WY, et al. An audit of medical thora amidine. Chest. 1 990;97:479-480.
coscopy and talc poudrage for pneumothorax prevention in 1 90 . Beers MF, Sohn M, Swartz M . Recurrent pneumothorax in
advanced COPD. Chest. 2004; 1 2 5 : 1 3 1 5- 1 320. AIDS patients with Pneumocystis pneumonia. A clinico
1 7 1 . Heffner JE, Huggins JT. Management of secondary spon pathologic report of three cases and review of the literature.
taneous pneumothorax: there's confusion in the air. Chest. Chest. 1 990;98:266-270.
2004; 1 2 5 : 1 1 90-1 1 92. 191. Gerein AN, Brumwell ML, Lawson LM, et al. Surgical man
1 72 . Yankaskas MR, Mallory GB Jr. The Consensus Committee. agement of pneumothorax in patients with acquired immu
Lung transplantation in cystic fibrosis. Chest. 1 99 8; 1 1 3 : nodeficiency syndrome. Arch Surg. 1 99 1 ; 1 26: 1 272- 1 276.
2 1 7-226. 1 92 . Ingram RJ, Call S, Andrade A, et al. Management and out
173. Dusmet M, Winton TL, Kesten S, et al. Previous intrapleu come of pneumothoraces in patients infected with human
ral procedures do not adversely affect lung transplantation. immunodeficiency virus. Clin Infect Dis. 1 996;23:624-627.
J Heart Lung Transplant. 1 996; 1 5 :249-254. 1 93 . Wait MA, Dal Nogare AR. Treatment of AIDS-related spon
1 74. Waller DA, Forty J, Soni AK, et al. Videothoracoscopic oper taneous pneumothorax. Chest. 1 994; 1 06: 693-696.
ation for secondary spontaneous pneumothorax. Ann 7horac 1 94. Wait MA. AIDS-related pneumothorax. Ann 7horac Surg.
Surg. 1 994;57: 1 6 1 2-1 6 1 5 . 1 997;64:290-29 1 .
1 7 5 . Deslauriers J. Th e management o f spontaneous pneumotho 195. Driver AG, Peden JG, Adams HG, et al. Heimlich valve
rax [Editorial] . Can ] Surg. 1 994;37: 1 82. treatment of Pneumocystis carinii-associated pneumothorax.
1 76 . Aihara K, Handa T, Nagai S, et al. Efficacy of blood-patch Chest. 1 99 1 ; 1 00:28 1-282.
pleurodesis for secondary spontaneous pneumothorax in 1 96. Walker WA, Pate JW, Amundson D, et al. AIDS-related
interstitial lung disease. Intern Med. 2 0 1 1 ;5 0 : 1 1 57- 1 1 62 . bronchopleural fistula. Ann 7horac Surg. 1 99 3 ; 5 5 : 1 048.
1 77. Zeng Y, Hong M, Zhang H, et a l . Transbronchoscopic 1 97. Byrnes TA, Brevig JK, Yeoh CB. Pneumothorax in patients
selective bronchial occlusion for intractable pneumothorax. with acquired immunodeficiency syndrome . ] 7horac Cardio
Respirology. 20 1 0; 1 5 : 1 68- 1 7 1 . vasc Surg. 1 990;98 : 5 46-5 5 0 .
1 78 . El-Sameed Y, Waness A, Al Shamsi I , e t al. Endobronchial 1 98 . Crawford B K , Galloway AC, Boyd A D , et al. Treatment o f
valves in the management of broncho-pleural and alveolo AIDS-related bronchopleural fistula b y pleurectomy. Ann
pleural fistulae. Lung. 20 1 2 ; 1 90:347-3 5 1 . 7horac Surg. 1 992;54:2 1 3 .
1 79 . Sepkowitz KA, Telzak EE, Gold JW, et al. Pneumothorax in 1 99 . Horowitz M D , Oliva H . Pneumothorax i n AIDS patients:
AIDS. Ann Intern Med. 1 9 9 1 ; 1 1 4:45 5-4 5 9 . operative management. Am Surg. 1 993;59:200-204.
1 8 0. Coker RJ , Moss F, Peters B, et a l . Pneumothorax in patients 200. Read CA, Reddy VD, O'Mara TE, et al. Doxycycline
with AID S . Respir Med. 1 993;87:43-47. pleurodesis for pneumothorax in patients with AIDS. Chest.
1 8 1 . Vricella LA, Trachiotis GD. Heimlich valve in the manage 1 994; 1 0 5 :823-825 .
ment of pneumothorax in patients with advanced AIDS. 20 1 . Flume PA, Strange C, Ye X , e t al. Pneumothorax i n cystic
Chest. 200 1 ; 1 20: 1 5- 1 8 . fibrosis. Chest. 2005 ; 1 28 :720-728 .
1 82. Tumbarello M , Tacconelli E, Pirronti T, e t al. Pneumotho 202. Flume PA. Pneumothorax i n cystic fibrosis. Chest. 2003;
rax in HIV-infected patients: role of Pneumocystis carinii 1 23:2 1 7-22 1 .
400 PLE U RAL D I S EASES
203. Luck SR, Raffensperger JG, Sullivan HJ, et al. Management 223. Downey DB, Towers MJ, Poon PY, et al. Pneumoperito
of pneumothorax in children with chronic pulmonary dis neum with catamenial pneumothorax. A]R Am J Roentgenol.
ease. J Thorac Cardiovasc Surg. 1 977;74:834-839. 1 990; 1 5 5 :29-30.
204. Flume PA, Mogayzel Jr PJ, Robinson KA, et al. Clinical 224. Dotson RL, Peterson CM, Doucette RC, et al. Medical
Practice Guidelines For Pulmonary Therapies Committee. therapy for recurring catamenial pneumothorax following
Cystic Fibrosis Pulmonary Guidelines: pulmonary complica pleurodesis. Obstet Gynecol. 1 993;82(4 Pt 2 suppl) :656-6 5 8 .
tions: hemoptysis and pneumothorax. Am ] Respir Crit Care 225. Marshall MB, Ahmed Z, Kucharczuk J C , et al. Catamenial
Med. 20 1 0 ; 1 82:298-306. pneumothorax: optimal hormonal and surgical manage
205. Curtis HJ, Bourke SJ, Dark JH, et al. Lung transplantation ment. EurJ Cardiothorac Surg. 2005;27: 662-666.
outcome in cystic fibrosis patients with previous pneumo 226. Alifano M, Trisolini R, Cancellieri A, et al. Thoracic endo
thorax. J Heart Lung Transplant. 2005;24:865-869. metriosis: current knowledge. Ann Thorac Surg. 2006;
206. Wilder RJ, Beacham EG, Ravitch MM. Spontaneous 8 1 :76 1 -769.
pneumothorax complicating cavitary tuberculosis. ] Thorac 227. Bagan P, Le Pimpec Barthes F, Assouad J, et al. Catamenial
Cardiovasc Surg. 1 962;43 : 5 6 1 -573. pneumothorax: retrospective study of surgical treatment.
207. Blanco-Perez ], Bordon J, Pineiro-Amigo L, et al. Pneumo Ann Thorac Surg. 2003;75 :378-3 8 1 .
thorax in active pulmonary tuberculosis: resurgence of an old 228. Stern H , Toole AL, Merino M . Catamenial pneumothorax.
complication? Respir Med. 1 998;92 : 1 269- 1 273. Chest. 1 980;78:480-482.
208. Freixinet JL, Caminero JA, Marchena J, et al. Spontaneous 229. Eckford SD, Westgate J . A cure for pneumothorax during
pneumothorax and tuberculosis. Long-term follow-up. Eur menstruation. Lancet. 1 996;347:734.
Respir J 20 1 1 ;3 8 : 1 26-1 3 1 . 230. Chernick V, Reed MH. Pneumothorax and chylothorax in
209. Urban T, Lazor R, Lacronique ], et al. Pulmonary lymphan the neonatal period. ] Pediatr. 1 970;76:624-632.
gioleiomyomatosis. A study of 69 patients. Medicine (Balti 23 1 . Ogata ES, Gregory GA, Kitterman JA, et al. Pneumo
more). 1 999;78:32 1 -337. thorax in the respiratory distress syndrome: incidence
2 1 0 . Chu SC, Horiba K, Usuki J, et al. Comprehensive evalua and effect on vital signs, blood gases, and pH. Pediatrics.
tion of 35 patients with lymphangioleiomyomarosis. Chest. 1 976; 5 8 : 1 77- 1 83 .
1 999; 1 1 5 : 1 04 1 - 1 052. 2 3 2 . Horbar JD, Badger GJ, Carpenter JH, e t a l . Trends in
2 1 1 . Johnson SR, Tattersfield AE. Clinical experience oflymphan mortality and morbidity for very low birth weight infants,
gioleiomyomatosis in the UK. Thorax. 2000; 5 5 : 1 052- 1 057. 1 99 1 - 1 999. Pediatrics. 2002; 1 1 0 : 1 43-1 5 1 .
2 1 2 . Hagaman JT, Schauer DP, McCormack FX, et al. Screening 233. Karlberg P. Respiratory studies in newborns. II. Pulmonary
for lymphangioleiomyomarosis by high-resolution computed ventilation and mechanics of breathing in the first minutes
tomography in young, nonsmoking women presenting with of life including the onset of respiration. Acta Paediatr.
spontaneous pneumothorax is cost-effective. Am J Respir Crit 1 962;5 1 : 1 2 1 - 1 36.
Care Med. 20 1 0; 1 8 1 : 1 376- 1 382. 234. Adler SM, Wyszogrodski I. Pneumothorax as a function of
2 1 3 . Young LR, Almoosa KF, Pollock-Barziv S, et al. Patient per gestational age: clinical and experimental studies. J Pediatr.
spectives on management of pneumothorax in lymphangi 1 975; 87:77 1 -775 .
oleiomyomatosis. Chest. 2006; 1 29 : 1 267- 1 273. 235. Engdahl MS, Gershan WM . Familial spontaneous pneumo
2 1 4. Mendez JL, Nadrous HF, Vassallo R, et al. Pneumothorax thorax in neonates. Pediatr Pulmonol. 1 998;25 :398-400.
in pulmonary Langerhans cell histiocyrosis. Chest. 2004; 236. Mehrabani D, Gowen CW Jr, Kopelman AE. Association of
1 2 5 : 1 028- 1 032. pneumothorax and hypotension with intraventricular haem
2 1 5 . Lillingron GA, Mitchell SP, Wood GA. Catamenial pneumo orrhage. Arch Dis Child. 1 9 9 1 ;66:48-5 1 .
thorax. ]AMA. 1 972;2 1 9 : 1 328- 1 332. 237. Kuhns LR, Bednarek FJ, Wyman ML, et al. Diagnosis of
2 1 6 . Alifano M, Jablonski C, Kadiri H, et al. Catamenial and pneumothorax or pneumomediastinum in the neonate by
non-catamenial, endometriosis-related or nonendometriosis transillumination. Pediatrics. 1 975;56:35 5-360.
related pneumothorax referred for surgery. Am J Respir Crit 238. Smith J, Schumacher RE, Donn SM, et al. Clinical course
Care Med. 2007; 1 76: 1 048- 1 05 3 . of symptomatic spontaneous pneumothorax in term and late
2 1 7. Korom S, Canyurt H, Missbach A , e t a l . Catamenial pneu preterm newborns: Report from a large cohort. Am J Perina
mothorax revisited: clinical approach and systematic review of tol. 2 0 1 1 ;2 8 : 1 63- 1 6 8 .
the literature. ] Thorac Cardiovasc Surg. 2004; 1 2 8 : 5 02-5 0 8 . 2 3 9 . Rainer C , Gardetto A , Fruhwirth M, e t al. Breast deformity
2 1 8 . Rousset-Jablonski C, Alifano M, Plu-Bureau G, et al. Cata in adolescence as a result of pneumothorax drainage during
menial pneumothorax and endometriosis-related pneu neonatal intensive care. Pediatrics. 2003; 1 1 1 : 8 0-86.
mothorax: clinical features and risk facrors. Hum Reprod. 240. Carlon GC, Kahn RC, Howland WS, et al. Clinical expe
20 1 1 ;26:2322-2329. rience with high frequency jet ventilation. Crit Care Med.
2 1 9 . Joseph J, Sahn SA. Thoracic endometriosis syndrome: new 1 9 8 1 ;9 : 1 -6.
observations from an analysis of 1 1 0 cases. Am J Med. 1 996; 24 1 . Sassoon CSH, Light RW, O'Hara VS, et al. Iatrogenic
1 00: 1 64- 1 70 . pneumothorax: etiology and morbidity. Respiration. 1 992;
2 2 0 . Alifano M , Roth T, Broet SC, et a l . Catamenial pneumotho 59:2 1 5-220.
rax: a prospective study. Chest. 2003; 1 24 : 1 004-1 008. 242. Zhan C, Smith M, Stryer D . Accidental iatrogenic pneumo
22 1 . Maurer ER, Schaal JA, Mendez FL. Chronic recurrent thorax in hospitalized patients. Med Care. 2006;44: 1 82- 1 86.
spontaneous pneumothorax due to endometriosis of the 243 . De Lassence A, Timsit JF, Taffiet M, et al. Pneumothorax in
diaphragm. JAMA. 1 9 5 8 ; 1 68:20 1 3-20 1 4 . the intensive care unit: incidence, risk factors, and outcome.
222. Jablonski C, Alifano M, Regnard JF, e t al. Pneumoperitoneum Anesthesiology. 2006; 1 04: 5- 1 3 .
associated with catamenial pneumothorax in women with 244. Despars JA, Sassoon CSH, Light RW. Significance o f iatro
thoracic endometriosis. Fertil Steril. 2009; 9 1 :930.el 9-e22. genic pneumothoraces. Chest. 1 994; 1 0 5 : 1 1 47- 1 1 50 .
C H A PT E R 24 I P N E U M OT H O RAX 40 1
245 . Vitulo P, Dore R, Cerveri I, et al. The role of functional respi central venous catheter insertion? Am J Respir Crit Care Med.
ratory tests in predicting pneumothorax during lung needle 200 1 ; 1 64:403-405.
biopsy. Chest. 1 9 9 5 ; 1 09 :6 1 2-6 1 5 . 264. Cavanna L, Civardi G, Vallisa D, et al. Ultrasound-guided
246. Collings CL, Westcott JL, Banson NL, e t al. Pneumothorax central venous cathererization in cancer patients improves
and dependent versus nondependent patient position after the success rate of cannulation and reduces mechanical
needle biopsy of the lung. Radiology. 1 999;2 I 0 : 5 9-64. complications: A prospective observational study of 1 ,978
247. Cox JE, Chiles C, McManus CM, et al. Transthoracic needle consecutive catheterizations. World] Surg Oncol. 20 1 0; 8 : 9 1 .
aspiration biopsy: variables that affect risk of pneumothorax. 265. Tyburski JG, Joseph AL, Thomas GA, et al. Delayed pneu
Radiology. 1 999;2 1 2 : 1 65- 1 68 . mothorax after central venous access: a potential hazard.
2 4 8 . Topal U, Ediz B . Transthoracic needle biopsy: factors effect Am Surg. 1 993;5 9 : 5 87-5 8 9 .
ing risk of pneumothorax. Eur J Radio/. 2003;48:263-267. 266. Gordon C E , Feller-Kopman D, Balk E M , et a l . Pneumotho
249. Geraghry PR, Kee ST, Mcfarlane G, et al. CT-guided rrans rax following rhoracentesis: A systematic review and mera
thoracic needle aspiration biopsy of pulmonary nodules: analysis. Arch Intern Med. 20 1 0; 1 70:332-339.
needle size and pneumothorax rate. Radiology. 2003; 267. Jones PW, Moyers JP, Rogers JT, et al. Ultrasound-guided
229:475-48 1 . rhoracentesis. Is it a safer method? Chest. 2003; 1 23 : 4 1 8-423 .
250. Hiraki T, Mimura H , Gobara H , et al. Incidence o f and risk 268. Duncan D R , Morgenthaler T l , Ryu JH, e t al. Reducing
factors for pneumothorax and chest tube placement after CT iatrogenic risk in thoracentesis. Chest. 2009; 1 3 5 : 1 3 1 5- 1 320.
fluoroscopy-guided percutaneous lung biopsy: retrospective 269. Capizzi SA, Prakash UB. Chest roentgenography after
analysis of the procedures conducted over a 9-year period. outpatient thoracentesis. Mayo Clin Proc. I 998;73:948-950.
A]R Am ] Roentgenol. 20 1 0; 1 94 : 809-8 1 4 . 270. Doyle ]], Hnatiuk OW, Torrington KG, et al. Necessiry
25 1 . Berger R , Smith D . Efficacy o f the lateral decubitus position of routine chest roentgenography after thoracentesis. Ann
in preventing pneumothorax after needle biopsy of the lung. Intern Med. 1 996; 1 24 : 8 1 6-820.
South Medj 1 9 88;8 1 : 1 1 40-1 1 43 . 27 1 . De Latorre FJ, Tomasa A, Klamburg J, et al. Incidence of
252. Choi C M , U m SW, Yoo C G , e t a l . Incidence and risk factors pneumothorax and pneumomediastinum in patients with
of delayed pneumothorax after transthoracic needle biopsy aspiration pneumonia requiring ventilatory support. Chest.
of the lung. Chest. 2004; 1 26: 1 5 1 6- 1 52 1 . 1 977;72: 1 4 1 - 1 44 .
2 5 3 . Zaetta JM, Licht MO, Fisher JS, e t al. A lung biopsy tract 272. Weg JG, Anzueto A , Balk RA , et a l . The relation of pneu
plug for reduction of postbiopsy pneumothorax and other mothorax and other air leaks to mortaliry in the acute
complications: results of a prospective, multicenter, random respiratory distress syndrome. N Engl J Med. 1 998;338:
ized, controlled clinical study. J Vtzsc Interv Radio/. 20 I O ; 3 4 1 -346.
2 1 : 1 235-1 243 . 273 . Boussarsar M, Thierry G, Jaber S, et al. Relationship between
254. Petsas T, Siamblis D, Giannakenas C, et al. Fibrin glue ventilatory settings and barotrauma in the acute respiratory
for sealing the needle track in fine-needle percutaneous distress syndrome. Intensive Care Med. 2002;28 :406-4 1 3 .
lung biopsy using a coaxial system: part II-clinical study. 274. Gammon RB, Shin M S , Buchalter S E . Pulmonary baro
Cardiovasc Intervent Radio/. 1 99 5 ; 1 8 :378-382. trauma in mechanical ventilation. Patterns and risk factors.
255. Billich C, Muche R, Brenner G, et al. CT-guided lung Chest. 1 992; I 02:568-572.
biopsy: incidence of pneumothorax after instillation of 275 . Sihoe AD, Wong RH, Lee AT, et al. Severe acute respira
NaCl into the biopsy track. Eur Radio/. 2008; 1 8 : 1 1 46- 1 1 52 . tory syndrome complicated by spontaneous pneumothorax.
256. Tam A , Singh P, Ensor J E , et a l . Air Travel after Biopsy Chest. 2004; 1 2 5 :2345-23 5 1 .
related Pneumothorax: Is it safe to fly? J Vase Interv Radio/. 276. Kao HK, Wang JH, Sung CS, e t al. Pneumothorax and
20 1 1 ;22 : 5 9 5-602. mortaliry in the mechanically ventilated SARS patients: a
257. Lockwood AH . Percutaneous subclavian vein catheteriza prospective clinical study. Crit Care. 2005;9:R440-R445 .
tion. Too much of a good thing? Arch Intern Med. 1 9 84; 277. Poe RH , Israel RH, Utell MJ, et al. Sensitiviry, specificiry,
1 44: 1 407- 1 4 0 8 . and predictive values of closed pleural biopsy. Arch Intern
2 5 8 . Farrell ], Walshe ], Gellens M, et a l . Complications associ Med. 1 984; 1 44:325-32 8 .
ated with insertion of j ugular venous catheters for hemodi 278 . Nour-Eldin N E , Naguib N N , Saeed A S , et al. Risk factors
alysis: the value of posrprocedural radiograph. Am J Kidney involved in the development of pneumorhorax during radio
Dis. 1 997;30:690-692. frequency ablation of lung neoplasms. A]R Am J Roentgenol.
2 5 9 . Damascelli B, Pacelli G, Frigerio LF, et al. Placement of 2009; 1 93:W43-W48.
long-term central venous catheters in outpatients: study of 279. Yoshimatsu R, Yamagami T, Terayama K, et al. Delayed and
1 34 patients over 24, 596 catheter days. A]R Am] Roentgenol. recurrent pneumothorax after RF ablation of lung tumors.
1 997; 1 6 8 : 1 235-1 239. Chest. 2009; 1 3 5 : 1 002- 1 009.
260. Miller JA, Singireddy S, Maldjian P, et al. A reevaluation 280. Frazier WD, Pope TL Jr, Findley LJ. Pneumothorax follow
of the radiographically detectable complications of percu ing transbronchial biopsy. Chest. 1 990;97: 539-540.
taneous venous access lines inserted by four subcutaneous 28 1 . Blasco LH, Hernandez IMS, Garrido W, et al. Safery of
approaches. Am Surg. 1 999;65 : 1 25-130. the transbronchial biopsy in outpatients. Chest. 1 9 9 1 ;99:
26 1 . Ray S, Stacey R, Imrie M, et al. A review of 560 Hickman 5 62-5 6 5 .
catheter insertions. Anaesthesia. 1 996;5 I : 9 8 1-98 5 . 282. Richard HM I I I , Stancato-Pasik A, Salky BA, et al. Pneumo
262. Molgaard 0, Nielsen M S , Handberg B B , e t al. Routine thorax and pneumomediastinum after laparoscopic surgery.
X-ray control of upper central venous lines: is it necessary? Clin Imaging. 1 9 97;2 1 :337-339.
Acta Anaesthesiol Scand. 2004;48 :685-68 9 . 283. Douglas JM, Spaniol S . Prevention of postoperative pneu
2 6 3 . Maury E, Guglielminotti ], Alzieu M, et al. Ultrasonic mothorax in patients undergoing cardiac surgery. Am J Surg.
examination. An alternative to chest radiography after 2002; 1 83 : 5 5 1-553.
402 PLE U RAL D I S EASES
284. Shanti CM, Carlin AM, Tyburski JG. Incidence of pneu injured patients with blunt chest trauma: results of a pro
mothorax from intercostal nerve block for analgesia in rib spective study. ] Trauma. l 997;43:405-4 1 1 .
fractures. J Trauma. 200 1 ; 5 1 :536-539. 304. Ball CG, Hameed SM, Evans D , et al. Occult pneumothorax
285. Tobkes Al, Nord HJ. Liver biopsy: review of methodology in the mechanically ventilated trauma patient. Can J Surg.
and complications. Dig Dis. 1 99 5 ; 1 32:67-74. 2003;46:373-379.
286. Oki M, Saka H, Kitagawa C, et al. Visceral pleural 305. Bridges KG, Welch G, Silver M, et al. CT detection of occult
perforation in two cases of ultrathin bronchoscopy. Chest. pneumothorax in multiple trauma patients. ] Emerg Med.
2005 ; 1 27:227 1 -2273 . 1 993; 1 1 : 1 79- 1 86.
287. Ludwig J, Kienzle GD. Pneumothorax in a large auropsy 306. Neff MA, Monk JS Jr, Peters K, et al. Detection of occult
population. Am ] Clin Pathol. 1 978;70:24-26. pneumothoraces on abdominal computed tomographic
2 8 8 . Paranjpe DV, Wittich GR, Hamid LW, et al. Frequency and scans in trauma patients. J Trauma. 2000;49 :28 1 -2 8 5 .
management of pneumothoraces in heart-lung transplant 307. Jaffer U, McAuley D . Best evidence topic report. Transtho
recipients. Radiology. 1 994; 1 90:255-256. racic ultrasonography to diagnose pneumothorax in trauma.
289. Lee YCG, McGrath GB, Chin WS, et al. Contralateral ten Emerg MedJ 2005;22:5 04-5 0 5 .
sion pneumothorax following unilateral chest tube drain 308. Chan SS. Emergency bedside ultrasound t o detect pneumo
age of bilateral pneumothoraces in a heart-lung transplant thorax. Acad Emerg Med. 2003; 1 0:9 1-94.
patient. Chest. 1 999; 1 1 6: 1 1 3 1 -1 1 33. 309. Ding W, Shen Y, Yang J, et al. Diagnosis of pneumothorax
290. Tocino IM, Miller MH, Fairfax WR. Distribution of pneu by radiography and ultrasonography-a meta-analysis. Chest.
mothorax in the supine and semirecumbent critically ill 20 1 1 ; 1 40 : 8 5 9-866.
adult. A]R Am ] Roentgenol. 1 98 5 ; 1 44:90 1-905. 3 1 0. Soldati G, Testa A, Pignataro G, et al. The ultrasonographic
29 1 . Kollef MH. Risk facrors for the misdiagnosis of pneumo deep sulcus sign in traumatic pneumothorax. Ultrasound
thorax in the intensive care unit. Grit Care Med. 1 9 9 1 ; Med Biol. 2006;32: 1 1 5 7- 1 1 63 .
1 9 : 906-9 1 0 . 311. Blaivas M , Lyon M, Duggal S . A prospective comparison
292. Chung MJ, Goo JM, I m J G , e t al. Value o f high-resolu of supine chest radiography and bedside ultrasound for the
tion ultrasound in detecting a pneumothorax. Eur Radio!. diagnosis of traumatic pneumothorax. Acad Emerg Med.
2005 ; 1 5 :930-93 5 . 2005 ; 1 2: 844-849.
293. Heffner JE, McDonald J, Barbieri C . Recurrent pneumo 312. Rivera L, O'Reilly EB, Sise MJ, et al. Small catheter tube
thoraces in ventilated patients despite ipsilateral chest tubes. thoracostomy: effective in managing chest trauma in stable
Chest. 1 99 5 ; 1 08 : 1 053- 1 0 5 8 . patients. ] Trauma. 2009;66:393-399.
294. Laronga C, Meric F, Truong MT, e t al. A treatment 3 1 3. Carrillo EH, Schmacht DC, Gable DR, et al. Thoracoscopy
algorithm for pneumothoraces complicating central venous in the management of posttraumatic persistent pneumotho
catheter insertion. Am ] Surg. 2000; 1 80:523-527. rax. ] Am Col! Surg. 1 99 8 ; 1 86:636-639.
295. Delius RE, Obeid FN, Horst HM, et al. Catheter aspiration 3 1 4. Carrillo EH, Kozloff M, Saridakis A, et al. Thoracoscopic
for simple pneumothorax. Experience with 1 1 4 patients. application of a topical sealant for the management of
Arch Surg. 1 9 89; 1 24: 833-836. persistent posttraumatic pneumothorax. J Trauma. 2006;
296. Yamagami T, Terayama K, Yoshimatsu R, et al. Role of 60: 1 1 1 - 1 1 4.
manual aspiration in treating pneumothorax after com 315. Knottenbelt JD, van der Spuy JW Traumatic pneumothorax:
puted tomography-guided lung biopsy. Acta Radio!. a scheme for rapid patient turnover. Br JAccident Surg. 1 990;
2009;4 : 1 - 8 . 2 1 :77-80.
2 9 7 . Gurley MB, Richli WR, Waugh KA . Outpatient manage 3 1 6. Ordog GJ, Wasserberger J, Balasubramanium S, et al.
ment of pneumothorax after fine-needle aspiration: eco Asymptomatic stab wounds of the chest. J Trauma. 1 994;
nomic advantages for the hospital and patient. Radiology. 36:680-684.
1 998;209:7 1 7-722. 3 1 7. Garramone RR Jr, Jacobs LM, Sahdev P. An objective
298. Yamagami T, Nakamura T, Iida S, et al. Management of method to measure and manage occult pneumothorax. Surg
pneumothorax after percutaneous CT-guided lung biopsy. Gynecol Obstet. 1 9 9 1 ; 1 73:257-26 1 .
Chest. 2002; 1 2 1 : 1 1 5 9- 1 1 64. 318. Brasel KJ, Stafford RE, Weigelt JA, et al. Treatment of
299. Wagner JM, Hinshaw JL, Lubner MG, et al. CT-guided occult pneumothoraces from blunt trauma. J Trauma.
lung biopsies: pleural blood patching reduces the rate of 1 999;46:987-990.
chest tube placement for postbiopsy pneumothorax. A]R Am 3 1 9. Wolfman NT, Myers WS, Glauser SJ, et al. Validity of CT
J Roentgenol. 20 1 1 ; 1 97:783-788. classification on management of occult pneumothorax: a pro
300. Anderson CLV, Crespo ]CA, Lie TH. Risk of pneumothorax spective study. A]R Am ] Roentgenol. 1 998 ; 1 7 1 : 1 3 1 7- 1 320.
not increased by obstructive lung disease in percutaneous 320. de Moya MA, Seaver C, Spaniolas K, et al. Occult pneu
needle biopsy. Chest. 1 994; 1 0 5 : 1 705-1708. mothorax in trauma patients: development of an objective
30 1 . Macklin MI, Macklin CC. Malignant interstitial emphy scoring system. J Trauma. 2007;63 : 1 3- 1 7 .
sema of the lungs and mediastinum as important occult 32 1 . Collins J C , Levine G, Waxman K. Occult traumatic pneu
complication in many respiratory diseases and other condi mothorax: immediate tube thoracostomy versus expectant
tions: interpretation of clinical literature in light of labora management. Am Surg. 1 992;58 :743-746.
tory experiment. Medicine. 1 944;23:28 1 -356. 322. Wilson H, Ellsmere J, Tallon J, et al. Occult pneumothorax
302. Wintermark M, Schnyder P. The Macklin effect: a frequent in the blunt trauma patient: Tube thoracostomy or observa
etiology for pneumomediastinum in severe blunt chest tion? Injury. 2009;40:928-93 1 .
trauma. Chest. 200 1 ; 1 20:543-547. 323. Mowery NT, Gunter OL, Collier B R, et al. Practice manage
303. Trupka A, Waydhas C, Hallfeldt KK, et al. Value of thoracic ment guidelines for management of hemothorax and occult
computed tomography in the first assessment of severely pneumothorax. ] Trauma. 2 0 1 1 ;70 : 5 1 0-5 1 8 .
C H A PT E R 24 I P N E U M OT H O RAX 403
324. Zierold D, Lee SL, Subramanian S, et al. Supplemental oxy- method to quantify air leakage after lung resection. Eur J
gen improves resolution of injury-induced pneumothorax. Cardiothorac Surg. 2006;29:867-872.
J Pediatr Surg. 2000;3 5 : 9 9 8- 1 00 1 . 348. Bishop MJ, Benson MS, Pierson DJ . Carbon dioxide excre
325. Lin MY, Wu MH, Chan CS, et al. Bronchial rupture caused tion via bronchopleural fistulas in adulc respiratory distress
by blunt chest inj ury. Ann Emerg Med. 1 99 5 ;25 :4 1 2-4 1 5 . syndrome. Chest. 1 9 87;9 1 : 400-402.
326. Kunisch-Hoppe M, Hoppe M, Rauber K, et al. Tracheal rup- 349. Prezant DJ, Aldrich TK, Fell SC, et al. The maintenance
ture caused by blunt chest trauma: radiological and clinical of total ventilatory requirements through a chronic bron
features. Eur Radio/. 2000; 1 0:480-483. chopleural cutaneous fistula. Am Rev Respir Dis. 1 9 87;
327. Sherr HP, Light RW, Merson MH, et al. Origin of pleural 1 36: 1 00 1 - 1 002.
fluid amylase in esophageal rupture. Ann Intern Med. 1 972; 350. Feeley Tw, Keating D , Nishimura T. Independent lung ven
76:98 5-986. tilation using high-frequency ventilation in the management
328. Air Transport Medicine Committee. Aerospace Medical of a bronchopleural fistula. Anesthesiology. l 9 8 8 ;69 :420-422.
Association. Medical guidelines for air travel. Aviat Space 35 1 . Lau KY. Postural management of bronchopleural fistula.
Environ Med. 1 996;67(suppl II) : B l -BS. Chest. 1 9 88;94: 1 1 22.
329. Cheatham ML, Safcsak K. Air travel following traumatic 352. Powner DJ, Cline CD, Rodman GH. Effect of chest-tube
pneumothorax: when is it safe? Am Surg. 1 999;65: 1 1 60-1 1 64. suction on gas flow through a bronchopleural fistula. Crit
330. Douglass RE, Levison MA. Pneumothorax in drug abusers: Care Med. 1 9 8 5 ; 1 3:99- 1 0 1 .
an urban epidemic. Am Surg. 1 986;52:377-3 80. 353. Orlando R III, Gluck EH, Cohen M , e t al. Ulcra-high
33 1 . Woodring JH, Baker MD, Stark P. Pneumothorax ex vacuo. frequency jet ventilation in a bronchopleural fistula model.
Chest. 1 996; 1 1 0 : 1 1 02- 1 1 05 . Arch Surg. 1 98 8 ; 1 23 : 5 9 1-593.
332. Ponrartana S, Laberge JM, Kerlan RK, et al. Management of 354. Albelda SM, Hansen-Flaschen JH, Taylor E, et al. Evaluation
patients with "ex vacuo" pneumothorax after thoracentesis. of high frequency jet ventilation in patients with broncho
Acad Radio!. 2005 ; 1 2: 9 80-986. pleural fistulas by quantitation of the air leak. Anesthesiology.
333. Heidecker J, Huggins JT, Sahn SA et al. Pathophysiology of 1 985;63 : 5 5 1-554.
pneumothorax following ulcrasound-guided thoracentesis. 355. Bishop MJ, Benson MS, Sato P, et al. Comparison of
Chest. 2006; 1 30: 1 1 73- 1 1 84 . high-frequency jet ventilation with conventional mechanical
334. Light RW Tension pneumothorax. Intensive Care Med. ventilation for bronchopleural fistula. Anesth Analg. 1 987;
l 994;20:468-469. 66:833-838.
335. Rutherford RB, Hurt HH, Brickman RD , et al. The patho- 356. Gonzalez F, Harris T, Black P, et al. Decreased gas flow
physiology of progressive, tension pneumothorax. J Trauma. through pneumothoraces in neonates receiving high
1 968;8:2 1 2-227. frequency jet versus conventional ventilation . ] Pediatr. 1 987;
336. Gustman P, Yerger L, Wanner A. Immediate cardiovascu- 1 1 0:464-466.
lar effects of tension pneumothorax. Am Rev Respir Dis. 357. Baumann MH, Sahn SA. Medical management and therapy
1 9 83; 1 27: 1 7 1 - 1 74. of bronchopleural fistulas in the mechanically ventilated
337. Carvalho P, Hilderbrandt J, Charan NB. Changes in bron- patient. Chest. 1 990;97:72 1 -728.
chial and pulmonary arterial blood flow with progressive 358. Lois M, Noppen M . Bronchopleural fistulas: an overview of
tension pneumothorax. J Appl Physiol. 1 996;8 l : 1 664- 1 669. the problem with special focus on endoscopic management.
338. Barton ED, Rhee P, Hutron KC, et al. The pathophysiology Chest. 2005 ; 1 28:3955-3965.
of tension pneumothorax in ventilated swine. J Emerg Med. 359. Kempainen RR, Pierson DJ . Persistent air leaks in patients
1 997; 1 5 : 1 47- 1 53. receiving mechanical ventilation. Semin Respir Crit Care
339. Beards SC, Lipman J. Decreased cardiac index as an indica- Med. 200 1 ;22:675-684.
tor of tension pneumothorax in the ventilated patient. Anaes- 360. Ponn RB . Complications of pulmonary resection. In: Shields
thesia. 1 994;49 : 1 37- 1 4 1 . Tw, LoCicero J III, Ponn RB, et al. eds. General Thoracic
340. Murphy DG, Sloan EP, Hart RG, et al. Tension pneumotho- Surgery, 6th ed. Philadelphia, PA: Lippincott Williams &
rax associated with hyperbaric oxygen therapy. Am ] Emerg Wilkins, 200 5 : 5 54-5 86.
Med. 1 9 9 1 ;9 : 1 76- 1 79 . 36 1 . Rivera C, Bernard A, Falcoz PE, et al. Characterization and
34 1 . Leigh-Smith S , Harris T. Tension pneumothorax-time fo r a prediction of prolonged air leak after pulmonary resection: a
re-think? Emerg Med]. 2005;22:8- 1 6 . nationwide study setting up the index of prolonged air leak.
342. McConaghy PM, Kennedy N . Tension pneumothorax due Ann Thorac Surg. 2 0 1 1 ;92: 1 062- 1 068.
to intrapulmonary placement of intercostal chest drain. 362. Cerfolio RJ, Bass CS, Pask AH, et al. Predictors and treatment
Anaesth Intensive Care. 1 99 5 ;23:496-498. of persistent air leaks. Ann Thorac Surg. 2002;73 : 1 727- 1 730.
343. Mainini SE, Johnson FE. Tension pneumothorax compli- 363. Fabian T, Federico JA, Ponn RB. Fibrin glue in pulmonary
eating small-caliber chest tube insertion. Chest. 1 990;97: resection: a prospective, randomized, blinded study. Ann
75 9-760. Thorac Surg. 2003;75 : 1 5 87- 1 5 92.
344. Castle N, Tagg A, Owen R. Bilateral tension pneumothorax. 364. Wain JC, Kaiser LR, Johnstone DW, et al. Trial of a novel
Resuscitation. 2005;65 : 1 03- 1 0 5 . synthetic sealant in preventing air leaks after lung resection.
345. Warner KJ, Copass MK, Bulger EM. Paramedic use of needle Ann Thorac Surg. 200 1 ;7 1 : 1 623- 1 62 8 .
thoracostomy in the prehospital environment. Prehosp Emerg 365. Cerfolio RJ, Bass C, Katholi C R . Prospective randomized
Care. 2008; 1 2 : 1 62- 1 6 8 . trial compares suction versus water seal for air leaks. Ann
346. Cerfolio RJ . Recent advances in the treatment of air leaks. Thorac Surg. 200 1 ;7 1 : 1 6 1 3- 1 6 1 7.
Curr Opin Pulm Med. 2005 ; 1 1 :3 1 9-323. 366. Brunelli A, Monteverde M, Borri A, et al. Comparison of
347. Anegg U, Lindenmann ], Matzi V, et al. AIRFIX": the water seal and suction after pulmonary lobectomy: a prospec
first digital posroperative chest tube airflowmetry-a novel tive, randomized trial. Ann Thorac Surg. 2004;77: 1 932- 1 937.
404 PLE U RAL D I S EASES
367. Deng B, Tan QY, Zhao YP, et al. Suction or non-suction co 374. Jones DP, David I. Gelfoam occlusion of peripheral bron
the underwater seal drains following pulmonary operation: chopleural fistulas. Ann Thorac Surg. 1 9 86;42:334-335.
meta-analysis of randomised controlled trials. Eur J Cardio 375. Lan R-S, Lee C-H, Tsai Y-H, et al. Fiberoptic bronchial
thorac Surg. 2 0 1 0;38:2 1 0-2 1 5 . blockade in a small bronchopleural fistula. Chest. 1 987;
368. Waller DA, Edwards JG, Rajesh PB. A physiological com 92: 944-946.
parison of flutter valve drainage bags and underwater seal sys 376. Torre M, Quaini E, Ravini M, et al. 1 9 87: endoscopic glu
tems for postoperative air leaks. Thorax. 1 999;54:442-443. ing of bronchopleural fistula. Updated in 1 994. Ann Thorac
369. Shackcloth MJ, Poullis M, Jackson M, et al. Intrapleural Surg. 1 994; 58:9 0 1-902.
instillation of aucologous blood in the treatment of pro 377. Hollaus PH, Lax F, Janakiev D, et al. Endoscopic treatment
longed air leak after lobectomy: a prospective randomized of postoperative bronchopleural fistula: experience with
controlled trial. Ann Thorac Surg. 2006;82: 1 0 52- 1 056. 45 cases. Ann Thorac Surg. 1 998;66:923-927.
370. Achanassiadi K, Bagaev E, Haverich A. Aucologous blood 378. Salmon CJ, Ponn RB, Westcott JL. Endobronchial vascular
pleurodesis for persistent air leak. Thorac Cardiovasc Surg. occlusion coils for control of a large parenchymal broncho
2009;57:476-479. pleural fistula. Chest. l 990;98 :233-234.
37 1 . Asamura H, Naruke T, Tsuchiya R, et al. Bronchopleural 379. Andreecti C, D'Andrilli A, Ibrahim M, et al. Effective treat
fi s tulas associated with lung cancer operations. Univariate ment of post-pneumoneccomy bronchopleural fistula by
and multivariate analysis of risk facrors, management, and conical fully covered self-expandable stent. Interact Cardio
outcome. ] Thorac Cardiovasc Surg. 1 992; 1 04: 1 4 5 6- 1 464. vasc Thorac Surg. 20 1 2 ; 1 4:420-42 5 .
372. Arnold PG, Pairolero PC. Intrathoracic muscle flaps. An 3 8 0 . Conforti S, Torre M , Fieschi S, e t al. Successful treatment
account of their use in the management of 1 00 consecutive of persistent postoperative air leaks following the placement
patients. Ann Surg. 1 990;2 l l :656-662. of an endobronchial one-way valve. Monaldi Arch Chest Dis.
3 73. De Maeseneer N, Van Hee R, Schoofs E, et al. The management 2 0 1 0;73 : 8 8-9 1 .
of bronchopleural fistulas. Acta Chir Belg. 1 987;87:269-274.
H e m ot h o ra x
Hemothorax is the presence of a significant amount seen in a 1 -year period ( 1 ) . The relative incidence of
of blood in the pleural space. Most hemothoraces hemothorax due to penetrating and blunt thoracic
result from penetrating or nonpenetrating chest trauma depends on whether the medical center cares
trauma. An occasional hemothorax results from iatro primarily for victims of automobile accidents or of
genic manipulation such as the placement of central stab and gunshot wounds.
venous catheters percutaneously by the subclavian or There is a high incidence ofhemothorax with blunt
internal j ugular route or from translumbar aortogra trauma. In a retrospective analysis of 5 1 5 cases of blunt
phy. On rare occasions, a hemothorax results from a chest trauma, 1 93 patients (37%) had hemothoraces
medical condition such as pulmonary embolism or (2) . In patients with rib fractures, hemothorax is more
rupture of an aortic aneurysm. common if the fracture is displaced (3) . Pneumothorax
Blood may enter the pleural space from injury to occurring concomitantly with hemothorax is common
the chest wall, diaphragm, lung, or mediastinum. whether the trauma is blunt or penetrating (Fig. 25 . 1 ) .
Blood entering the pleural space coagulates rapidly.
Presumably as a result of physical agitation produced
by movement of the heart and the lungs, the clot may
be defibrinated. Loculation occurs early in the course
of hemothorax, as with empyema.
When a diagnostic thoracentesis in a medical patient
reveals pleural fluid that appears to be pure blood, a
hematocrit should always be obtained on the pleural
fluid. Frequently, although the pleural fluid appears to
be blood, the hematocrit on the pleural fluid is less than
5%. A hemothorax should be considered to be present
only when the pleural fluid hematocrit is equal to or
greater than 50% of the peripheral blood hematocrit.
If a measured hematocrit is not available, a rough esti
mate of the hematocrit can be obtained by dividing the
pleural fluid red blood cell (RBC) count by 1 00,000.
For example, a pleural fluid RBC count of 1 ,000,000
equates with a pleural fluid hematocrit of 1 0%.
405
406 PLE U RAL D I S EASES
In a series of 1 1 4 patients with hemothorax second was obscured or if the fluid was more than 2 cm in
ary to blunt trauma, 71 (62%) also had pneumothorax thickness on the lateral decubitus radiograph ( 1 0) .
(4) . In another series of 373 patients with hemothorax Most patients need tube thoracostomy fo r a relatively
secondary to penetrating trauma, 307 (83%) also had short period. In one study of 1 ,845 patients from
pneumothorax ( 1 ) . South Africa, chest tubes were checked every 6 hours
and were removed when there was no air leak and
there was less than 50 mL drainage in the previous
D i a g nosis
6 hours ( 1 0) . With this protocol, the average drainage
The diagnosis of a traumatic hemothorax should be time was 27. 1 hours and 82% of the patients were
suspected in any patient with penetrating or nonpen discharged in less than 48 hours ( 1 0) . Chest tubes
etrating trauma to the chest. The diagnosis is usually should be removed as soon as they stop draining or
established by the demonstration of a pleural effu cease to function because they can serve as conduits
sion with a chest radiograph or with ultrasound. As for pleural infection.
an initial screening test, surgeon-performed ultraso In the past, it was believed by some that the inser
nography appears to be as sensitive as a supine chest tion of a chest tube would decrease pleural pressure
radiograph in detecting hemothorax. In one study and would thereby augment the pleural bleeding.
of 360 patients, 39 of 40 effusions were detected If the bleeding originates from lacerated pleura,
by ultrasound and 37 were detected by chest radio however, apposition of the pleural surfaces will pro
graph. The performance time for ultrasonography duce a tamponade and will stop the bleeding ( 1 1 ) .
was significantly faster than that for chest radiogra I f the bleeding i s from larger vessels, the slight
phy ( 1 . 3 vs. 1 4.2 minutes) ( 5 ) . In some patients, a decrease in the pleural pressure with a chest tube
hemothorax becomes apparent only after some delay. is insignificant in comparison to the transvascu
In one series of 1 67 hemothoraces from Toronto due lar pressure ( 1 1 ) . The advantages of the immediate
to blunt trauma, the initial chest radiograph revealed institution of tube thoracostomy are as follows : (a) it
no hemothorax in 7 patients, but a hemothorax was allows more complete evacuation of the blood from
subsequently diagnosed 22 hours to 1 6 days later in the pleural space; (b) it stops the bleeding if the
these 7 (4.2%) (6) . All seven patients had multiple rib bleeding is from pleural lacerations; (c) it allows one
fractures, which were displaced in five (6) . to quantitate easily the amount of continued bleed
A case can be made for obtaining chest computed ing; (d) it may decrease the incidence of subsequent
tomography (CT) scans in all patients with severe chest empyema because blood is a good culture medium
injuries. Trupka et al. (7) obtained supine chest radio ( 1 2) ; (e) the blood drained from the pleural space
graphs and chest CT scans in 1 03 patients with severe may be autotransfused ( 1 ) ; and (f) the rapid evacua
chest injuries and reported that the chest radiograph tion of pleural blood decreases the incidence of sub
missed hemothoraces in 2 1 patients, lung contusion sequent fibrothorax ( 1 3) .
in 33 patients, and pneumothorax in 27 patients. In a Large-bore chest tubes (size 2 4 t o 3 6 F ) should
second study of 93 patients with blunt trauma to the be inserted in patients with hemothorax because the
chest, 25 had normal supine chest radiographs and the blood frequently clots ( 1 4) . Beall et al. ( 1 2) recom
chest CT scan showed multiple injuries in 1 3 patients mend inserting the chest tube high (fourth or fifth
including two aortic lacerations, three hemothoraces, intercostal space) in the midaxillary line because the
and one pericardia! effusion (8) . diaphragm may be elevated by the trauma. Imme
diate thoracotomy or thoracoscopy is indicated for
suspected cardiac tamponade, vascular injury, pleu
Treatment
ral contamination, debridement of devitalized tissue,
The treatment of choice for patients with traumatic sucking chest wounds, or major bronchial air leaks
hemothorax is the immediate insertion of a chest ( 1 5 ) . Vascular inj ury is suggested if the initial chest
tube. Obviously, if there is only a very small hemo tube output is more than 1 , 500 mL.
thorax, tube thoracostomy is not necessary. An occult Continued pleural hemorrhage is another indi
hemothorax is one that is seen on CT scan but not cation for immediate thoracotomy or video-assisted
apparent on the supine chest radiograph. Tube tho thoracic surgery (VATS) . There is no precise crite
racostomy is not necessary for most patients with rion for the amount of pleural bleeding that should
occult hemothorax (9) . In one study, tube thoracos serve as an indication for surgery, because each case
tomy was performed if either diaphragmatic dome must be considered individually ( 1 1 ) ; however, if the
C H APT E R 2 5 / H E M OT H O RAX 407
bleeding is at a rate of more than 200 ml/hour and received the antibiotic (24) . However, in a more
shows no signs of slowing, thoracotomy or VATS recent study, Maxwell et al. (2 5), in a randomized,
should be seriously considered. Approximately 1 0% double-blind study, gave cefazolin for 24 hours, for
to 20% of patients with hemothorax require thora the duration of tube thoracostomy or placebo to
cotomy or VATS ( 1 ,3,4, 1 0, 1 2) . 224 patients. The use of antibiotics did not signifi
One must ensure that the bleeding i s not from a cantly affect the incidence of empyema or pneumo
misplaced central venous catheter ( 1 6, 1 7) . Mattox nia (25 ) . However, only 1 .3% of the patients who
and Fisher ( 1 6) reported seven patients with a trau received antibiotics developed empyema while 5 .6%
matic hemothorax in whom continued bleeding of the patients receiving no antibiotics developed
originated from a misplaced central venous catheter. empyema. A longer duration of tube thoracostomy
This diagnosis is readily established by examining the and a higher thoracic trauma score were associated
appearance of the pleural drainage when the character with a higher incidence of empyema (25 ) .
of the infusion fluid is changed. If blood is obtained I t appears that prehospital autotransfusion has a
when fluid is withdrawn from the central catheter, role in the management of life-threatening hemotho
the catheter may still be misplaced in the pleural rax. Barriot et al. (26) developed a system by which
space ( 1 8) . autotransfusions could be administered in ambu
VATS is replacing thoracotomy in some patients lances. The system consists of a 28-to-30 F plastic
with traumatic hemothorax who otherwise would chest tube and an autotransfusion device. The latter
have been subj ected to thoracotomy. Thoracotomy is basically a 75 0-mL bag with filters. The blood
rather than VATS should be performed if there is drains by gravity into the collection bag and is then
exsanguinating hemorrhage through the chest tubes reinfused without anticoagulation into a central line.
( 1 9,20) . However, in the hemodynamically stable They reported the use of their system on 1 8 patients
patient with persistent bleeding VATS is very effective in Paris with life-threatening traumatic hemothorax.
(2 1 ) . Villavicencio et al. (22) in a literature review During transfer to the hospital, the patients received
found that VATS was effective in controlling the 4. 1 ± 0.6 L of autotransfused blood, without antico
bleeding in 33 of 40 (82%) such cases. VATS was agulation. Thirteen of the 1 8 patients (72%) survived,
effective in controlling the bleeding when the bleeding and there were no complications. They believed that
arose from intercostal vessels and lung lacerations. the 1 3 patients would have died had it not been for
Manlulu et al. (20) reported that VATS was the the autotransfusions.
definitive procedure in all 1 9 patients in a single insti
tution on whom it was attempted. Indications for
Com p l icati o n s of H e m o t h o ra x
VATS in this study included continued hemorrhage
in six, retained hemothorax in six, and suspected The four main pleural complications of traumatic
diaphragmatic injury in three (20) . hemothorax are the retention of clotted blood in the
An alternative approach to the patient with persis pleural space, pleural infection, pleural effusion, and
tent bleeding is to perform a contrast-enhanced CT fibrothorax.
scan and then perform transcatheter arterial emboli
Reten tion of Clotted Blood
zation in patients who exhibit contrast extravasation.
Hagiwara et al. (23) used this approach in 6 patients Although most traumatic hemothoraces are managed
who were draining > 200 ml/hour from their chest nonoperatively by tube thoracostomy, some of these
tube and reported that 5 had contrast extravasation blood collections remain only partially drained. The
and the bleeding was controlled in all five via trans residual blood, already potentially contaminated by
catheter arterial embolization. the insertion of the chest tube, may be the nidus for
The efficacy of prophylactic antibiotics for the significant complications such as empyema or fibro
prevention of empyema in patients with tube tho thorax. Surgical evacuation of retained hemothoraces
racostomy for hemothorax is unclear. In an early decreases this risk (27) . In one study from California,
study, Brunner et al. (24) randomly allowed 90 such 20 of 703 patients (3%) treated with tube thoracos
patients to receive cefazolin or nothing immediately tomy for hemothorax were found to have retention
before and then every 6 hours until tube removal. of clotted blood (27) . The chest radiograph may be
They reported that there were six empyemas and misleading in patients with hemothorax who are sus
three pneumonias in the control group, but only pected of having retention of clotted blood. Velmahos
one pneumonia and no empyema in the group that et al. (27) obtained chest CT scans on 58 consecutive
408 PLE U RAL D I S EASES
patients with hemothorax who had opacification hospitalizations (3 1 ,32) . In addition, there has been
extending above the costophrenic angle on standard one case in which hypoxemic respiratory failure
chest radiograph 1 day after tube thoracostomy for developed after the intrapleural administration of
hemothorax. They found that the prediction of Buid fibrinolytics (33) . Nevertheless, ifVATS is not readily
from the chest radiograph was correct in less than available, consideration can be given to intrapleural
50% of cases by both the chest surgeon and the fibrinolytics or thoracotomy.
radiologist (27) . This study suggests that a CT scan
Posttrauma tic Empyema
should be obtained before an operative procedure is
undertaken to remove the clotted blood. The second complication following hemothorax is
When a patient is identified as having retention empyema, occurring in 3% to 4% of cases (34) . This
of clotted blood in the pleural space, three ques complication can be minimized by using meticu
tions need to be asked: (a) Does the clot need to lously sterile technique while inserting thoracostomy
be removed? (b) If so, when should it be removed? tubes and by ensuring good apposition of the pleural
(c) How should it be removed? If the residual clot surfaces so that no space remains for the accumula
occupies at least one third of the involved hemithorax tion of Buid or blood. The risk of empyema increases
48 to 72 hours or more after the initial tube thoracos with the presence of persistent bronchopleural fis
tomy, it should probably be removed (28 ,29) . The clot tula, pulmonary contusions, and residual clotted
is most easily disrupted and removed by thoracos hemothoraces. As mentioned in the preceding text,
copy 48 to 72 hours after the initial injury. After 7 the administration of antibiotics to patients with
to 9 days, the clot adheres to the lung and pleura, hemothorax who are treated with tube thoracostomy
making thoracoscopic removal difficult or impossible significantly reduces the subsequent development
and increasing the incidence of complications, such of empyema and possibly pneumonia (24) . Patients
as retained collections, persistent pleural drainage, who are admitted in shock are more likely to develop
or air leaks (28) . Therefore, the optimal time to remove empyema, as are those with gross contamination of
the clotted blood appears to be between 48 and the pleural space at the time of the original injury.
96 hours (29) . Empyema is also more common with associated
The optimal method for removal of the clotted abdominal injuries ( 1 3) and with prolonged pleural
blood appears to be VATS (22,28,29) . Villavicencio drainage ( 1 4) . At times, the pathogenesis of the
et al. (22) reviewed eight reports with a total of empyema is rupture of a pulmonary abscess (3 5 ) . The
99 patients who were subj ected to VATS for retained treatment of empyema complicating hemothorax is
hemothoraces. Evacuation of the retained hemo similar to that of any bacterial infection of the pleural
thorax was successful in 89 (90%) of the cases. An space (see Chapter 1 2) .
alternative approach to the patient with a retained
Pleural Effusion
hemothorax is to insert a second chest tube. A ran
domized study in which 24 patients received a second The third complication of hemothorax is the occur
chest tube and 1 5 patients underwent VATS demon rence of a pleural effusion when the chest tubes are
strated that hospitalizations were significantly shorter removed. In the series reported by Wilson et al. ( 1 5) ,
(5.4 vs. 8 . 1 days) and the hospital costs were sig 37 of 290 patients ( 1 3%) with n o residual hemotho
nificantly less in the group that received VATS (30) . rax developed pleural effusions after removal of the
Of the 24 patients who received a second chest chest tubes, and 40 of 1 1 8 patients (34%) with resid
tube, 1 0 (42%) eventually required thoracoscopy or ual hemothorax had pleural effusions at the time of
thoracotomy (30) . discharge from the hospital. Of these 77 patients with
Some authors have recommended the intra a pleural effusion after tube thoracostomy, 20 (26%)
pleural injection of fibrinolytic agents for the treat had empyema, but the pleural effusions resolved in
ment of retained hemothorax (3 1 ) . Inci et al. (3 1 ) the other 57, leaving no or minimum residual disease
treated 2 4 patients with either 250,000 I U strepto ( 1 5 ) . This series indicates that pleural effusions are
kinase or 1 00,000 IU urokinase, and each patient common after tube thoracostomy for hemothorax.
received a mean of five doses. Although complete When such effusions occur, a diagnostic thoracentesis
response was demonstrated in 62. 5 % of the patients, should be performed to rule out the possibility of a
this treatment is not recommended because it is pleural infection. If no pleural infection is present,
more expensive than thoracoscopy owing to the the pleural effusion usually clears by itself and leaves
expense of the fibrinolytic agents and the longer no residual disease.
C H APT E R 25 I H E M OT H O RAX 409
IATRO G E N I C H E M OT H O RAX
D i a g nosis a n d Treatm ent
When a hemothorax is discovered, the possibility
of iatrogenic origin should be considered. The most When bloody-appearing pleural Buid is obtained
common causes of iatrogenic hemothorax are the during a diagnostic thoracentesis, the hematocrit of
perforation of a central vein by a percutaneously the pleural Buid should be determined. If the hema
inserted catheter ( 1 6, 1 7 ,36) or leakage from the aorta tocrit of the pleural Buid is greater than 50% that
after translumbar aortographic study (37) . Iatrogenic of the peripheral blood, the patient has a hemotho
hemothorax can also follow thoracentesis or pleural rax. Regardless of how bloody the pleural Buid looks,
biopsy. Insertion of a Swan-Ganz catheter is occa a hematocrit should be obtained because pleural Buid
sionally associated with rupture of the pulmonary with a hematocrit of less than 5% may appear to be
artery with a resulting hemothorax, and, in such blood. If a hematocrit is not available, a pleural Buid
instances, an immediate thoracotomy is necessary if RBC count that is more than 50% of the peripheral
the patient is going to survive (38) . Iatrogenic hemo RBC count indicates that the patient has a hemotho
thoraces have also been reported after many other rax. A chest tube should be inserted into patients with
procedures, including percutaneous lung aspiration a spontaneous hemothorax to evacuate the blood and
or biopsy, transbronchial biopsy, and sclerotherapy to assess the rate of continued bleeding. Thoracotomy
for esophageal varices (39) . In the intensive care or VATS should be considered if brisk bleeding
unit, hemothorax is more common in patients after (> 1 00 ml/hour) persists. If the bleeding is from an
invasive procedures if they have chronic renal failure intercostal artery, it may be controlled with selective
(40) . Patients with iatrogenic hemothorax should angiographic embolization (54) .
be managed with chest tubes for the same reasons as
are patients with traumatic hemothorax. H e mothorax Co m p l i cati n g
Anticoag u l ant T h e ra py
N O NTRAU M AT I C H E M OTH O RAX
Twenty cases of hemothorax complicating anticoagu
Nontraumatic hemothoraces are distinctly uncom lant therapy had been reported by 200 8. (5 5 ) . The
mon. The most common cause, not associated with anticoagulation was for pulmonary embolism in 1 8 or
pneumothorax, is metastatic malignant pleural disease the 20 cases (55). Usually, the hemothorax becomes
(4 1 ) . The most common tumors are schwanommas apparent 4 to 7 days after anticoagulant therapy
of von Recklinghausen disease and soft tissue tumors is initiated, but it may occur only after several months
such as sarcomas, angiosarcomas, and hepatocellular (55). Most of the patients were receiving heparin and
carcinomas. The second most common cause is a warfarin and the coagulation parameters were supra
complication of anticoagulant therapy for pulmo therapeutic in six (55). Hemothoraces have also been
nary emboli (42) , and the third leading cause is observed with enoxaparin therapy (56) and systemic
probably catamenial hemothorax (43). Other causes therapy with thrombolytics (57) . The spontaneous
of spontaneous hemothorax include a complication hemothorax is almost always on the side of the original
410 PLE U RAL D I S EASES
23. Hagiwara A, Yanagawa Y, Kaneko N, et al. Indications for 42. Ali HA, Lippmann M, Mundathaje U, et al. Spontaneous hemo
transcatheter arterial embolization in persistent hemothorax thorax: a comprehensive review. Chest. 2008 ; 1 34: 1 0S6- 1 06S.
caused by blunt trauma. ] Trauma. 2008;6S : S 8 9-S94. 43. Shepard MK, Mancini MC, Campbell GD, et al. Right-sided
24. Brunner RG, Vinsant GO, Alexander RH , et al. The role of hemothorax and recurrent abdominal pain in a 34-year-old
antibiotic therapy in the prevention of empyema in patients woman. Chest. 1 993; 1 03 : 1 239- 1 240.
with an isolated chest injury (ISS 9-1 O): a prospective study. 44. Slind RO, Rodarte JR. Spontaneous hemothorax in an other
] Trauma. 1 990;30: 1 1 48- 1 1 S 3. wise healthy young man. Chest. 1 974;66: 8 1 .
2 S . Maxwell RA, Campbell DJ, Fabian TC, et al. Use of presumptive 4 S . Cochran JW Pancreatic pseudocyst presenting as massive
antibiotics following cube thoracostomy for traumatic hemopneu hemothorax. Am J Gastroenterol. l 978;69: 84-87.
mothorax in the prevention of empyema and pneumonia 46. Dippel WF, Dory DB, Ehrenhaft JL. Tension hemothorax due
a multi-center trial. J Trauma. 2004; S7:742-749 . to patent ductus arceriosus. N Engl] Med. 1 973;28 8 : 3 S 3-3S4.
2 6 . Barriot P, Riou B, Viars P. Prehospital autotransfusion i n life 47. DeFrance JH, Blewecc JH Jr, Ricci JA, et al. Massive hemotho
threatening hemothorax. Chest. 1 9 88;93: S22-S26. rax: two unusual cases. Chest. 1 974;66: 82-84.
27. Velmahos GC, Demetriades D, Chan L, et al. Predicting 48. Spear BS, Sully L, Lewis CT. Pulmonary arcerio-venous
the need for thoracoscopic evacuation of residual traumatic fistula presenting as spontaneous hemothorax. Thorax. l 97S;
hemothorax: chest radiograph is insufficient. ] Trauma. 30:3SS-3 S6.
1 999;46:6S-70 . 49. Martinez FJ, Villanueva AG, Pickering R, et al. Spontaneous
2 8 . Carrillo EH, Richardson JD. Thoracoscopy in the manage hemothorax. Report of 6 cases and review of the literature.
ment of hemothorax and retained blood after trauma. Curr Medicine. 1 992;7 1 :3S4-368.
Opin Pulm Med. l 998;4:243-246. SO. Smith PR, Manjoney DL, Teitcher JB, et al. Massive hemo
29. Mowery NT, Gunter OL, Collier BR, et al. Practice manage thorax due to intrathoracic extramedullary hematopoiesis in a
ment guidelines for management of hemothorax and occult patient with thalassemia intermedia. Chest. 1 9 8 8;94:603-608.
pneumothorax. ] Trauma. 2 0 1 1 ;70: S l 0-S 1 8 . S l. Rodriguez E, Martinez MJ, Javaloyas M, et al. Haemothorax
3 0 . Meyer DM, Jessen ME, Wait MA , e t al. Early evacuation in the course of chickenpox. Thorax. 1 986;4 l :49 1 .
of traumatic retained hemothoraces using thoracoscopy: a S2. Harrison NK, Wilkinson J , O 'Donohue J , et al. Osteochon
prospective, randomized trial. Ann Thorac Surg. 1 997;64: droma of the rib: an unusual cause of haemothorax. Thorax.
1 396- 1 400. 1 994;49 :6 1 8-6 1 9 .
3 1 . lnci I, Ozcelik C, Ulku R, et al. Intrapleural fibrinolytic treatment S3. Laurin S, Aronson S, Schuller H, e t al. Spontaneous hemo
of traumatic clotted hemothorax. Chest. 1 998; 1 1 4: l 60-l 6S. thorax from bronchopulmonary sequestration. Pediatr Radio!.
32. Oguzkaya F, Akcali Y, Bilgin M . Videothoracoscopy versus 1 9 80; 1 0 : S4-S 6.
intrapleural streptokinase for management of post traumatic S4. Kessel B, Alfici R, Ashkenazi I, et al. Massive hemothorax
retained haemothorax: a retrospective study of 6S cases. Injury. caused by intercostal artery bleeding: selective embolization
200S;36: S26-S29. may be an alternative to thoracotomy in selected patients. Tho
33. Frye MD, Jarratt M, Sahn SA. Acute hypoxemic respiratory rac Cardiovasc Surg. 2004;S2 :234-236.
failure following intrapleural thrombolytic therapy for hemo SS. Ali HA, Lippmann M , Mundathaje U, et al. Spontaneous hemo
thorax. Chest. 1 994; 1 0 S : l S 9 S-l S96. thorax: a comprehensive review. Chest. 2008; 1 34: 1 0S6- 1 06S.
34. Battistella FD, Benfield JR. Blunt and penetrating injuries of S6. Mrug M , Mishra PV, Lusane HC, et al. Hemothorax and ret
the chest wall, pleura and lungs. In: Shields TW, LoCicero J III, roperitoneal hematoma after anticoagulation with enoxaparin.
Ponn RB, eds. General Thoracic Surgery. Philadelphia, PA: South MedJ 2002;9S: 936-938.
Lippincott Williams & Wilkins; 2000 : 8 1 S-83 l . S7. Varnholt V, Ringe H, Nietsch L, et al. Hemothorax under
3 S . O'Connor JV, Chi A, Joshi M , et al. Post-traumatic empyema: thrombolytic therapy with recombinant tissue plasminogen
aetiology, surgery and outcome in l 2S consecutive patients. activator (re-PA) in a 1 6-year-old girl. Eur ] Pediatr 1 999;
Injury. 20 1 2 ahead of print 1 S 8 (suppl. 3) : S l 40-S l 42 .
36. Krauss D, Schmidt GA. Cardiac tamponade and contralat S 8 . Rostand RA, Feldman RL, Block E R . Massive hemothorax
eral hemothorax after subclavian vein catheterization. Chest. complicating heparin anticoagulation for pulmonary embolus.
1 9 9 1 ; 9 9 : S l 7-S 1 8 . South Med] 1 977;70: 1 1 28-1 1 30 .
37. Bilbrey GL, Hedberg CL. Hemorrhagic pleural effusion sec S 9 . Joseph J, Sahn SA. Thoracic endometriosis syndrome: new
ondary to aorcography: a case report. ] Thorac Cardiovasc Surg. observations from an analysis of 1 1 0 cases. Am ] Med. 1 996;
1 967; S 4 : 8 S-8 9 . 1 00: 1 64- 1 70.
3 8 . KearneyTJ, Shabot MM. Pulmonary artery rupture associated 60. Hwong TM, Ng CS, Lee TW; et al. Video-assisted thoracic
with the Swan-Ganz catheter. Chest. 1 99 S ; l 0 8 : 1 349- 1 3S2. surgery for primary spontaneous hemopneumothorax. Eur J
39. Hussain A, Raja AJ. Occurrence of hemothorax (unilateral) Cardiothorac Surg. 2004;26: 8 93-8 96.
after sclerotherapy. Am ] Gastroenterol. 1 99 1 ; 86: 1 S S 3- 1 SS4. 6 1 . Wu YC, Lu MS, Yeh CH, et al. Justifying video-assisted tho
40. Chen CY, Hsu CL, Chang CH, et al. Hemothorax in a medi racic surgery for spontaneous hemopneumothorax. Chest.
cal intensive care unit: incidence, comorbidiry and prognostic 2002; 1 22: 1 844- 1 847.
factors. ] Formos Med Assoc. 20 1 0 ; 1 09:S74-S 8 1 . 62. Hsu NY, Shih CS, Hsu CP, et al. Spontaneous hemopneumo
4 1 . Berliner K. Hemorrhagic pleural effusion: an analysis of 1 20 thorax revisited: clinical approach and systemic review of the
cases. Ann Intern Med. 1 94 1 ; 1 4:2266-2284. literature. Ann Thorac Surg. 200S;80: l 8 S 9-l 863.
C h y l ot h o ra x a n d Pse u d oc h y l ot h o ra x
At times, pleural fluid is milky or at least turbid. When the descending aorta in close proximity to the esoph
the milkiness or turbidity persists after centrifu agus and the pericardium. At the level of the fourth
gation, it is almost always due to a high lipid content to sixth thoracic vertebrae, the duct crosses to the left
of the pleural fluid. High levels of lipid accumulate of the vertebral column and continues cephalad to
in the pleural fluid in two situations. First, when the enter the superior mediastinum between the aortic
thoracic duct is disrupted, chyle can enter the pleural arch and the subclavian artery and the left side of the
space to produce a chylous pleural effusion. In this esophagus.
situation, the patient is said to have a chylothorax Once the thoracic duct passes the thoracic inlet, it
( 1 ) . Second, in long-standing pleural effusions, large arches 3 to 5 cm above the clavicle and passes anterior
amounts of cholesterol or lecithin-globulin com to the subclavian artery, vertebral artery, and thyro
plexes can accumulate in the pleural fluid to produce cervical trunk to terminate in the region of the left
a chyliform pleural effusion. The patient is then said jugular and subclavian veins. Wide anatomic varia
to have a pseudochylothorax. It is important to differ tions may exist in all portions of the thoracic duct.
entiate these two conditions because their prognosis More than one thoracic duct may leave the cisterna
and management are completely different. chyli. The duct may continue on the right side of
the vertebral column to enter the veins in the right
subclavian region. Multiple anastomoses usually exist
C H Y LOT H O RAX
between various lymphatic channels, and direct lym
A chylothorax is formed when the thoracic duct is phaticovenous communications with the azygos vein
disrupted and chyle enters the pleural space. may be present (2,3).
The drainage from the thoracic duct is called chyle.
Chyle appears grossly as a milky, opalescent fluid that
Path ophysiolog i c Featu res
usually separates into three layers upon standing: a
Dietary fats in the form oflong-chain triglycerides are creamy uppermost layer containing chylomicrons,
transformed into chylomicra and very-low-density a milky intermediate layer, and a dependent layer
lipoproteins. These are secreted into the intestinal containing cellular elements, most of which are small
lacteals and lymphatics and are then conveyed to lymphocytes (4) . If the patient has not eaten, how
the cisterna chyli, which overlies the anterior surface ever, chyle may be only slightly turbid because its
of the second lumbar vertebra, posterior to and to lipid content will be reduced. Chyle is bacteriostatic
the right of the aorta. Usually, one major lymphatic and does not become infected even when it stands
vessel, the thoracic duct, leaves the cisterna chyli at room temperature for several weeks (5) . Lampson
and passes through the esophageal hiatus of the dia (5) reported that Escherichia coli and Staphylococcus
phragm into the thoracic cavity. The thoracic duct aureus were unable to grow in 1 00% chyle. Chyle
ascends extrapleurally in the posterior mediastinum that is extravasated into the pleural cavity is not irri
along the right side of the anterior surface of the ver tating and usually does not evoke the formation of a
tebral column and lies between the azygos vein and pleural peel or a fibroelastic membrane.
412
C HAPT E R 26 I C HYLOTH O RAX A N D P S E U DOCHYLOTHO RAX 413
Chylothorax has also been reported as a complication and reported that 9 (24%) had a chylothorax. We
of coronary artery bypass surgery when the internal reviewed cases of the superior vena caval syndrome
mammary artery is harvested (22) , heart transplant at my previous hospital over a 6-year period and
(23), high translumbar aortography (24) , sclerother found that chylothorax complicated 4 of 76 cases
apy for esophageal varices (25 ) , thoracolumbar fusion (3 1 ) . Chylothorax can also complicate innominate
for correction of kyphosis (26) , and cervical node vein (32) or left subclavian vein thrombosis (33) . Cir
dissection (27) . rhosis is a relatively common cause of chylothorax.
Of course, penetrating trauma to the chest or neck Romero et al. (34) analyzed 24 cases of chylothorax
such as gunshot or knife wounds can also sever the occurring at their institution and reported that 5
thoracic duct and may lead to chylothorax. Trauma (2 1 %) were secondary to cirrhosis. Interestingly,
in which the spine is hyperextended or a vertebra is the mean protein level in these five chylothoraces
fractured is most likely to cause chylothorax, particu was only 1 .7 g/dL (compared with 4. 1 g/dL in the
larly if the inj ury occurs after the recent ingestion of other chylous effusions) , the mean lactate dehydroge
a fatty meal (27) . A chylothorax secondary to closed nase (LDH) level was only 96 IU/L (compared with
trauma is usually on the right side, and the site of 3 5 1 IU/L in the other chylous effusions) and asci
rupture is most commonly in the region of the 9th tes was present in three of the five patients (34) . On
or 1 0th thoracic vertebra (27) . Such trauma includes rare occasions, a chylothorax is associated with heart
falls from a height, motor vehicle accidents, compres failure or the nephrotic syndrome and the effusion
sion injuries to the trunk, heavy blows to the back or is also a transudate in these instances (3 5 ) . In most
stomach, and childbirth (28) . The injury may be less patients with the nephrotic syndrome and a chylo
impressive, and chylothoraces have been attributed thorax, the chylothorax is secondary to chylous asci
to coughing, vomiting, and weight lifting. In one tes, but on occasions, it can be secondary to superior
well-documented case report, an episode of vigorous vena caval thrombosis (36) .
stretching while yawning was followed by swelling in Many other causes of chylothorax have been
the left supraclavicular fossa and the development of reported, but even when all are grouped together,
bilateral chylothoraces (29 ) . they account for only a small percentage of chylo
Another leading cause of chylothorax is malig thoraces. The most interesting of these is pulmo
nancy. The most common malignancy to cause a chy nary lymphangioleiomyomatosis (LAM), which has
lothorax is a lymphoma, and lymphomas accounted associated interstitial parenchymal infiltrates, and is
for 75% of the chylothoraces due to malignan discussed later in this chapter. Other causes include
cies are listed in Table 26. 1 . In the series from the Gorham syndrome (37) (also discussed later in this
Mayo Clinic ( 1 3) , lymphomas (20 non-Hodgkin chapter) , Kaposi sarcoma in patients with acquired
lymphoma and 3 Hodgkin disease) accounted for immunodeficiency syndrome (AIDS) (38,39), the
68% of the chylothoraces due to malignancy. Other yellow nail syndrome ( 1 4) , filariasis, paragonimiasis
malignancies producing a chylothorax in the Mayo (40), giant lymph node hyperplasia (Castleman dis
Clinic series included chronic lymphocytic leuke ease) (4 1 ) , lymphangiomatosis (42) , familial lymph
mia 5, metastatic disease 5, and lung cancer 1 ( 1 3) . edema (43), lymphangitis of the thoracic duct,
Chylothorax may b e the presenting symptom of lym obstruction of the superior vena cava secondary to
phoma (7, 1 1 , 1 5) . Therefore, a nontraumatic chy Behi;:et syndrome (44,45 ) , tuberculosis (46) , sarcoid
lothorax is an indication for a diligent search for a osis involving the intrathoracic lymph nodes (47) ,
lymphoma. In the series of Roy et al. ( 1 1 ) before the aneurysms of the thoracic aorta that erode the duct,
availability of computed tomographic (CT) scans, the abnormalities of the lymphatic vessels such as intes
diagnosis of lymphoma was not established until 6 tinal lymphangiectasis (48) or reticular hyperplasia
to 1 2 months after the appearance of the chylothorax ( 1 2,49) , radiation-induced mediastinal fibrosis (50),
in four patients. and hypothyroidism (5 1 ) .
The third category of chylothorax is the miscel Th e fourth category o f chylothorax i s idiopathic,
laneous category. Thrombosis of the superior vena including most cases of congenital chylothorax. One
cava or the subclavian vein is becoming one of the should exclude lymphoma as a cause of the chylo
more common causes of chylothorax. Berman et al. thorax before it is labeled idiopathic. Most cases of
(30) reviewed the case histories of 37 infants and idiopathic chylothorax in the adult are probably due
children with thrombosis of their superior vena to minor trauma, such as coughing or hiccupping,
cava in a newborn and pediatric intensive care unit after the ingestion of fatty meals.
C HAPT E R 26 I C HYLOTH O RAX A N D P S E U DOCHYLOTHO RAX 415
Chylothorax is the most common form of pleural Th e mediastinal pleura eventually ruptures, chyle gains
effusion encountered in the first few days of life (52) . access to the pleural space, and dyspnea is produced
The fetal pleural effusion discussed in Chapter 20 is by the chyle compressing the lung. At times, hypoten
probably also a chylothorax. Neonatal chylothorax sion, cyanosis, and extreme dyspnea occur when the
is relatively uncommon; during a 22-year period, chyloma ruptures into the pleural space. The ruptured
12 cases were diagnosed at the Hospital for Sick chyloma is no longer visible radiographically.
Children, which is a large pediatric tertiary care With nontraumatic chylothorax, the onset of
center (53) . The babies are usually born at full term symptoms is usually gradual. In congenital chylo
after normal labor and delivery. The etiology of con thorax, the infant develops respiratory distress in the
genital chylothorax is unknown (54) . Abnormalities first few days of life; 50% of patients have symptoms
of the thoracic duct have not been found in most within the first 24 hours, whereas 75% have symp
babies who have undergone exploratory thoracotomy toms by the end of the first week (52) . The chyle pro
(52) . Several cases of generalized pleural oozing have duction in a neonate may exceed 250 mL/day (54) .
been described during surgery (53) . It is possible that The main threat to life from chylothorax is mal
birth trauma may result in a tear of a major lymphatic nutrition and a compromised immunologic status.
channel in at least some individuals. Most neonatal Because the thoracic duct carries 2,500 mL of fluid
chylothoraces represent pleural effusions that had daily that contains substantial amounts of protein,
been present antenatally. In some cases, a congenital fats, electrolytes, and lymphocytes, the patient can
chylothorax is associated with Turner's syndrome, become cachectic rapidly if this amount of chyle is
Noonan's syndrome, or Down's syndrome (54) . Con removed daily through chest tubes or repeated thora
genital chylothorax is also more common in infants centesis. In addition, the patients develop lymphope
who are hydropic or who have polyhydramnios (53). nia and a compromised immunologic status because
Interestingly, mice that lack the integrin a 9 {3 1 of the removal of large numbers of lymphocytes with
appear normal at birth but then develop respiratory the chyle (58). In one case report, one patient had
failure and die between 6 and 1 2 days of age ( 5 5 ) . The more than 35 L of fluid withdrawn over a 1 4-day
respiratory failure is caused by bilateral chylothoraces. period that contained 2.3 kg of fat and 0.7 kg of pro
Although the thoracic duct appears normal grossly, tein. During this period, the peripheral lymphocyte
microscopic examination reveals edema and lym count dropped from 1 ,665 to 264/mm3 (8) . Indeed,
phocytic infiltration in the chest wall ( 5 5 ) . Members until Lampson initially described successful liga
of the integrin family of adhesion receptors mediate tion of the thoracic duct in 1 948 (5), the mortality
both cell-cell and cell-matrix interactions and have rate from chylothorax was 50%. When managing
been shown to play vital roles in embryonic develop a patient with chylothorax, one should abandon
ment, wound healing, and other biologic processes. conservative treatment before the patient becomes
It has been postulated that integrin a 9 {3 1 deficiency too malnourished and immunocompromised.
could be one cause of congenital chylothorax (55) .
D ia g nosis
C l i n ical M a n ifestations
The diagnosis of chylothorax is usually not difficult
The initial symptoms of chylothorax are usually because chyle typically has a distinctive white, odorless,
related to the presence of the space-occupying fluid milky appearance. When such fluid is found, the main
in the thoracic cavity, and, therefore, patients have differentiation is between chylothorax, empyema, and
dyspnea. On rare occasions, the patient can develop a a pseudochylothorax. The milkiness with empyema
tension chylothorax with compromise of the systemic is caused by the suspended white blood cells (WBCs)
circulation (56) . Pleuritic chest pain and fever are rare and debris, and if such fluid is centrifuged, the super
because chyle is not irritating to the pleural surface. natant is clear. The cloudiness of the chyliform pleu
With traumatic chylothorax, a latent period of 2 to ral effusion from a pseudochylothorax is also caused
1 0 days usually occurs between the trauma and the by high lipid levels and either cholesterol or lecithin
onset of the pleural effusion (27) . There is one case globulin complexes. Chylous and chyliform pleural
report in which the latent period was 1 1 weeks (57) . fluids remain opaque after centrifugation.
Lymph collects extrapleurally in the mediastinum If cholesterol crystals are responsible for the turbidity,
after the initial thoracic duct disruption, forms a chy they may be easily demonstrated by examination of the
loma, and produces a posterior mediastinal mass (3) . pleural fluid sediment (Fig. 26. 1 ) . If the turbidity is
416 PLE U RAL D I S EASES
The best way to establish the diagnosis of chylothorax At times, one would like to demonstrate the loca
is by measuring the triglyceride and cholesterol levels tion of the disruption of the thoracic duct. In other
C HAPT E R 26 I C HYLOTH O RAX A N D P S E U DOCHYLOTHO RAX 417
patients, visualization of the thoracic duct is desired tube into the subclavian vein for a total of 1 8 days.
in patients with chylothorax for which there is no There have been no large series evaluating this pro
ready explanation. In the past, this has most com cedure. In the first half of the last century, a couple
monly been done by injecting a dye in the dorsum of patients died from "anaphylaxis" soon after chyle
of the feet, which is subsequently absorbed in the infusion was started (69) .
lymphatics. When the dye is conveyed to the chest, When a postoperative chylothorax is managed
the thoracic duct can be visualized. However, most initially with tube thoracostomy, efforts should be
medical centers no longer perform lymphangiograms. made to decrease the Bow of chyle through dietary
There have been two recent reports that have dem manipulation. The Bow of chyle is minimized if all
onstrated that imaging of the thoracic duct can be nourishment by mouth is halted and if the patient's
performed by lymphoscintigraphy (66,67) . With this gastrointestinal tract is maintained as empty as possible
procedure, Technetium-99m human serum albumin by constant gastric suction (70) . The patient's nutri
is injected into the dorsum of the foot or hand and tional status can be maintained with intravenous
subsequently the thoracic duct is imaged in nuclear hyperalimentation (7 1 ) . In the past, attempts have
medicine. There is one paper that reported the suc been made to decrease the lymph Bow by providing
cessful imaging of the thoracic duct after the oral the fat calories in the diet with medium-chain triglyc
ingestion of iodine- 1 23-labeled 1 5- (4-iodophenyl)- erides (7 1 ) . These triglycerides have 1 0 or fewer car
3-(R,S)-methyl-pentadecanoic acid (I- 1 2 3 BMIPP) bon atoms and are absorbed directly into the portal
(68) . The thoracic duct can be imaged approximately vein and thereby gain entrance to the circulatory
80 minutes after the ingestion of this tracer (68) . The system without ever entering the thoracic duct (72) .
advantage of this procedure is that it does not require Because they are relatively unpalatable and hyperali
the difficult injections into the foot. mentation decreases the Bow of chyle much more,
hyperalimentation rather than medium-chain triglyc
erides is recommended when one wishes to reduce
Treatment
the Bow of chyle. The Bow of chyle is also decreased if
The main danger to patients with chylothorax is the patient stays in bed because any lower extremity
that they become malnourished and immunocom movement increases the Bow of lymph (72). There
promised because of the removal of large amounts is one report in which the Bow of chyle was reduced
of protein, fat, electrolytes, and lymphocytes from markedly with inhaled nitrous oxide in a 4 1 -week-old
the body with repeated thoracentesis or chest tube infant with a postoperative chylothorax (73) . The
drainage. In the past, the mortality rate from chylo reduction in the Bow of chyle was thought to be due
thorax approached 50%. When managing a patient to alleviation of the central venous hypertension (73) .
with chylothorax, one must treat the chylothorax One must question the efficacy of this approach as
definitively, such as with thoracic duct ligation or there have been no subsequent articles on this subject.
pleuroperitoneal shunt implantation, before the The defect in the thoracic duct frequently closes
patient becomes too cachectic to tolerate the opera spontaneously in traumatic chylothorax. If the tho
tion. Because the management of chylothorax differs racic duct is transected in dogs, chyle ceases to enter
for traumatic, nontraumatic, and congenital chylo the pleural space within 1 0 days as collateral lymphatic
thoraces, treatment regimens are described separately. channels are formed (9) . In one series of 22 children
with postoperative chylothorax, 1 9 (86.4%) closed
Traumatic Chylothorax
spontaneously when the patients were treated con
The general aims in treating the patient with trau comitantly with total parenteral nutrition or low-fat
matic chylothorax are relief of dyspnea by removal enteral diets (74) . The average duration of drainage of
of the chyle, prevention of dehydration, maintenance these 1 9 patients was 1 3 .7 days, with a range of 7 to
of nutrition, and a reduction in the rate of chyle 30 days (74) . In another series of 47 adults with post
formation. When a postoperative chylothorax is dis operative chylothorax from the Mayo Clinic (75 ) , the
covered, tube thoracostomy should be performed to leak closed spontaneously in 7 of 36 patients ( 1 9%)
remove the chyle and relieve the dyspnea. In this situ who received central hyperalimentation and in 6 of
ation, consideration should be given to recycling the 1 1 patients (55%) who were treated with medium
chyle to prevent malnutrition and immunosuppres chain triglyceride diets (75 ) .
sion. Thomson and Simms (69) reported one case in Chylothoraces that occur after pulmonary resec
which the chyle was reinfused directly from the chest tions are usually small and resolve with medical
418 PLE U RAL D I S EASES
therapy (76) . In one series, 20 of 26 cases resolved reports are of one or two cases. Sometimes the results
with only medical therapy (76) . The probable expla are impressive as the chyle seems to cease entering the
nation for these excellent results is that the injury is chest soon after the medication is started. In other
usually to a tributary of the thoracic duct and not to cases, the chylothorax resolves but the relationship
the thoracic duct itself ( 1 7,76) . between the administration of the drug and the reso
If large amounts of chyle continue to drain for lution of the chylothorax is unclear. Some patients do
more than several days postoperatively, a procedure not respond at all ( 1 8 , 8 8 ,93) . Until better data are
should be performed to treat the chylothorax defini available, it seems prudent to use octreotide in per
tively. The alternatives at this j uncture are (a) to sistent chylothoraces before proceeding to more inva
administer somatostatin or its analogue octreotide, sive procedures (9 1 ) . If there is no large decrease in
(b) to insert a pleuroperitoneal shunt, (c) to percutane the lymphatic flow within a few days, then one should
ously embolize the thoracic duct using a transabdom proceed to a more invasive procedure.
inal approach, (d) to attempt to create a pleurodesis The side effects of octreotide are usually mild
to obliterate the pleural space through tube thoracos and transient. The most serious side effect was necro
tomy, (e) to perform thoracoscopy with pleural abra tizing enterocolitis in an infant treated with intrave
sion or partial pleurectomy to create a pleurodesis, nous octreotide for postoperative chylothorax (94) .
(f) to perform thoracoscopy with attempted ligation One case of hypothyroidism was attributed to soma
of the thoracic duct, or (g) to perform a thoracotomy tostatin administration (9 5 ) .
with ligation of the thoracic duct.
P l e u roperito n e a l S h u nt
Som atostati n (Octreoti d e)
A good method to remove the chyle and alleviate the
In the last decade, there have been numerous reports dyspnea with chylothorax appears to be the insertion
on treating chylothoraces with somatostatin or octreo of a pleuroperitoneal shunt (96-98) (see the discus
tide. These agents have been used for congenital chylo sion on pleuroperitoneal shunt in Chapter 1 0) . The
thoraces (77,78), postoperative chylothoraces in both primary advantage of the pleuroperitoneal shunt is
children (79-8 5) and adults (86-88), spontaneous that the lymph is not removed from the body, and,
chylothoraces (88), chylothorax due to the yellow nail therefore, the patient does not become malnourished
syndrome (89), or non-Hodgkin lymphoma (90) . or immunocompromised. When the chyle is shunted
Octreotide is a somatostatin analog and offers the to the peritoneal cavity, it is absorbed without creat
advantage of subcutaneous administration whereas ing significant ascites (96) . A second advantage is that
somatostatin requires continuous intravenous infu the pleuroperitoneal shunt can be inserted with local
sion (9 1 ) . The mechanism of action of octreotide is anesthesia as opposed to general anesthesia, which is
unknown. Its administration decreases lymphatic required for thoracic duct ligation. When a pleuro
flow and triglyceride absorption in dogs (92) presum peritoneal shunt is implanted, the defect closes spon
ably by decreasing the blood flow in the splanchnic taneously in most cases and the shunt can be removed
circulation and by decreasing gastrointestinal motil 30 to 90 days after its insertion (99) . When the chyle
ity (9 1 ) . Alternatively, the lymphatic vessels may have ceases to enter the pleural space, the shunt is removed.
somatostatin receptors and their stimulation could The shunt should not be inserted if chylous ascites
result in decreased lymphatic flow (84) . The primary is present. Little et al. (96) inserted pleuroperitoneal
side effects are related to suppressive actions of gas shunts in two patients with chylothoraces postopera
trointestinal motility and secretion and include loose tively and reported that the patients had complete res
stools, malabsorption, nausea, and flatulence. The olution of their effusions with subsequent removal of
usual starting dose of somatostatin is 3 . 5 µ,g/kg/hour, the shunts. In the largest series using the pleuroperito
which can be increased to 1 0 µ,g/kg/hour. The usual neal shunt, 1 6 infants were treated with the shunt and
dose for octreotide in the adult is 50 mg q 8 hours excellent results were obtained in 12 (97) . The only
(9 1 ) . The usual dose for octreotide in children has patients who did not have favorable outcomes were
ranged from 0 . 3 to 1 . 0 µ,g/kg/hour (77,8 1 ) . those whose chylothoraces were due to central venous
Th e actual efficacy o f somatostatin o r octreotide thrombosis (97) . Murphy et al. (97) recommend plac
in the management of chylothorax is difficult to ing the shunt if the drainage persists beyond 5 days.
evaluate. There are no controlled studies comparing An alternative approach to patients with postoperative
the efficacy of octreotide with anything. Most cited chylothorax is to insert the pleuroperitoneal shunt as
C HAPT E R 26 I C HYLOTH O RAX A N D P S E U DOCHYLOTHO RAX 419
soon as the diagnosis is made without ever resorting been rare. In a review of 78 patients receiving this
to tube thoracostomy. It is not known how important procedure with a mean follow-up of 34 months, four
it is to slow the flow of lymph if a pleuroperitoneal patients had chronic leg swelling and six had chronic
shunt is inserted. At present, I place no dietary restric diarrhea ( 1 03) . There was a report of one patient
tions on my patients who have chylothoraces and who developed chylous ascites ( 1 04) . This proce
who are being treated with pleuroperitoneal shunts. It dure appears to be the best available option in those
should be noted that since the pleuroperitoneal shunt facilities where it is available.
is being used less for malignant pleural effusions, it is
also being used less commonly to treat chylothorax:.
Pleu rodesis t h ro u g h a Ch est Tu be
It is not recommended, however, because the intra Probably, the best method is to inject Evans blue
pleural administration of talc can lead to the develop dye at a dose of 0.7 to 0 . 8 mg/kg, the total dose not
ment of the acute respiratory distress syndrome (see exceeding 25 mg, into the subcutaneous tissue of the
Chapter 1 O) . An alternative to thoracoscopy with talc leg. Within 5 minutes, the chyle will be stained blue
insuffiation is thoracoscopy with pleural abrasion (70) . The patient may also be given butter or cream
( 1 1 4) or partial parietal pleurectomy. However, if one to eat 3 to 4 hours preoperatively. The objection to
is going to perform thoracoscopy, ligation of the tho this method is that the stomach may not be empty
racic duct is a better procedure. before the induction of anesthesia, although stomach
contents may be removed by nasogastric suction (70) .
It is probably preferable to place a tube in the duode
Ligation of the Thoracic D u ct t h ro u g h
num and inject the cream in this location ( 1 1 8) . If for
Tho racoto my
some reason the thoracic duct cannot be successfully
A definitive treatment for postoperative chylothorax ligated at thoracotomy, a parietal pleurectomy should
is ligation of the thoracic duct. Lampson first demon be performed to obliterate the pleural space (7) .
strated that a chylothorax could be controlled by liga If the chylous drainage from the chest tubes
tion of the thoracic duct (5). Ligation of the thoracic persists and the nutritional status of the patient is
duct causes no ill effects, probably on account of the deteriorating, one must not delay thoracotomy too
multiple anastomoses among various lymphatic chan long. In one series, all three patients with traumatic
nels and direct lymphaticovenous communications chylothorax who underwent thoracotomy died in the
(6,7) . In one recent series ( 1 1 5) , ligation of the tho postoperative period ( 1 1 ) . These deaths were attrib
racic duct controlled the chylothorax in 2 1 of 22 cases uted to the debilitation of the patients by the time the
of chylothorax following surgery in adults (9 5%). Nath operation was performed.
et al. ( 1 1 6) performed thoracic duct ligation in 20 pedi
atric patients with chylothorax after cardiothoracic sur
Ligation of the Thoracic D u ct t h ro u g h
gery. They reported that the procedure was successful in
Thoracoscopy
1 6 patients (80%), but noted that patients with throm
bosis of upper body venous vessels or prolonged chest With the advent of video-assisted thoracic surgery
tube drainage were more likely to fail and/or die ( 1 1 6) . (VATS) , one would expect that ligation of the tho
These researchers recommend thoracic duct ligation racic duct would be tried with the videothoracoscope.
within 2 weeks of recognizing the chylothorax ( 1 1 6) . Thoracoscopy permits the entire pleural space to be
Usually, the side o f the chylothorax is ipsilateral t o the visualized, as well as allowing direct suture of a lym
original surgery, and reoperation can be performed phatic leak. Fahimi et al. ( 1 1 9) performed VATS on
through the original thoracotomy incision (75) . If the six adult patients with postoperative chylothorax
chylothorax is unilateral, the thoracotomy should be and reported that the site of the thoracic duct injury
performed on the side of the fluid (70) . If the chylo could be identified and treated in four patients. In the
thorax is bilateral, a right thoracotomy should be per remaining two patients, the site of injury could not
formed because the duct is more readily approached be identified and either fibrin glue or talc insuffiation
from that side (70) . An alternative approach has been was performed. The chylothoraces were cured in all
suggested by Mason et al. ( 1 1 7) , who recommend that six patients ( 1 1 9) . Wurnig et al. ( 1 20) reported the
the thoracic duct be ligated j ust below the diaphragm successful use of VATS with ligation of the thoracic
through an intraabdominal approach. duct immediately above the diaphragm in four cases of
It has been recommended that a preoperative lym adults with chylothoraces of varying etiology. None of
phangiogram be obtained in every case of chylothorax the patients had pleural drainage for more than 7 days
that does not respond to nonoperative management postoperatively ( 1 20) . Similar results were reported
because the site of leak or obstruction can usually be by Christodoulou et al. ( 1 2 1 ) . Pego-Fernandez et al.
demonstrated by this technique (75 ) . At the time of ( 1 22) performed thoracic duct ligation via VATS in
operation, one should attempt to find the actual point 1 4 children with chylothorax post cardiac surgery and
of leakage from the duct and ligate the duct on both reported that it was successful in 1 2 (86%) .
sides of the leak (70) . In many instances, however, the
Nontra uma tic Chylothorax
leak cannot be located, and, therefore, the thoracic
duct should be ligated. Several aids for identifying the In general, the goals of management of nontraumatic
thoracic duct intraoperatively have been suggested. chylothorax are the same as for traumatic chylothorax.
C HAPT E R 26 I C HYLOTH O RAX A N D P S E U DOCHYLOTHO RAX 42 1
With nontraumatic chylothorax, however, one must be administered. Roy et al. ( 1 1 ) reported that radiation
also attempt to establish a cause. If the characteristics therapy to the mediastinum adequately controlled
of the chyle are transudative, that is, there is a low the chylothorax for the remainder of the patient's life
protein and a low LDH , then the three most likely in 68% of those with lymphoma and in 50% of those
etiologies are cirrhosis, congestive heart failure, and with metastatic carcinoma. If radiotherapy or chemo
the nephrotic syndrome (34) . therapy does not control the chylothorax in patients
The possibility of lymphoma should be consid with known lymphoma or metastatic carcinoma,
ered in all patients with a nontraumatic chylothorax exploratory thoracotomy is probably not indicated
because it is the most common cause of nontraumatic in view of these patients' dismal prognosis ( 1 1 ) . If
chylothorax (Table 26. 1 ) . Frequently, the patient with these patients are symptomatic from the chylothorax,
lymphoma and chylothorax has no evidence of lym however, one should insert a pleuroperitoneal shunt
phoma outside the thorax. CT studies of the medi unless the patient also has ascites. An alternative to
astinum should be performed in all patients with the pleuroperitoneal shunt is medical thoracoscopy
nontraumatic chylothorax to ascertain whether medi with the instillation of talc. Mares and Mathur ( 1 24)
astinal lymphadenopathy is present. In women with reported their experience with this procedure for 24
chylothorax and parenchymal infiltrates, another pos hemithoraces in 1 9 patients with lymphoma-related
sibility is pulmonary LAM (see the section later in chylothorax in whom chemotherapy or radiation
this chapter) . therapy had failed. They reported that none of their
Another test that should be obtained on all patients patients had a recurrence of their chylothorax dur
with nontraumatic chylothoraces is a lymphangio ing the 90-day follow-up period. Other alternatives
gram ( 1 23) . With the lymphangiogram, a total or are VATS with ligation of the thoracic duct or the
partial obstruction of lymph flow and the position implantation of an indwelling catheter ( 1 25 ) . I am
of the obstacle can be demonstrated. In addition, the hesitant to insert an indwelling catheter because of
lymphangiogram can demonstrate the presence of the risk of malnutrition and immunosuppression.
enlarged lymph nodes or lymphangiectasis that may There are special aspects to the treatment of chylo
give a clue as to the etiology of the chylothorax. How thorax associated with some entities. Because patients
ever, as mentioned previously, lymphangiograms are with chylothorax secondary to the nephrotic syndrome
not available in many medical centers. are likely to have chylous ascites, it is important not
The initial management of a patient with a nontrau to place a pleuroperitoneal shunt or ligate the thoracic
matic chylothorax should be similar to that of a patient duct until this possibility is evaluated ( 1 26) . The chylo
with a traumatic chylothorax. In most cases, a pleuro thorax associated with sarcoidosis is likely to disappear
peritoneal shunt should be inserted. If tube thoracos if the patient is treated with corticosteroids (99) . The
tomy is performed, the gastrointestinal tract should be chylothorax associated with thrombosis of the superior
put at rest and the patient's nutritional status should vena cava resolved in one case with the use of throm
be maintained by parenteral hyperalimentation. If the bolytic therapy and placement of a self-expanding
chylothorax is due to minor trauma, these measures are metallic stent ( 1 27) . The chylothorax associated with
usually curative within 1 week. If CT examination of cirrhosis has been successfully treated by a transjugular
the mediastinum reveals no lymphadenopathy or other intrahepatic portosystemic shunt ( 1 28) .
masses and the chylothorax is controlled, no further
Congenital Chylothorax
treatment is indicated. In one series of 3 5 patients with
chylothorax due to tumors, none were successfully As mentioned earlier in this chapter, chylothorax is
managed with chest tubes or repeated pleural aspira the most common type of pleural effusion in infants.
tion ( 1 1 ) . If the chylothorax is not controlled with the Chylothorax in infancy can be fatal. The mortality
pleuroperitoneal shunt, or if the CT study of the rate was 30% in an earlier series of 1 0 patients with
mediastinum is positive, the patient should undergo congenital chylothorax ( 1 29). The three deaths in this
a videothoracoscopy or an exploratory thoracotomy. series were all ascribed to malnutrition and second
The mediastinum should then be carefully examined ary infection, and the babies who died were the only
for masses, with lymphoma in mind. In addition, the ones in the series who were subjected to more than 1 4
thoracic duct should be ligated. thoracenteses. On the other hand, five o f the babies
If the patient is known to have lymphoma or (50%) had no recurrence of their chylothorax after
metastatic carcinoma, then radiation should be one to three thoracenteses, and all seven babies who
given to the mediastinum or chemotherapy should survived were apparently normal ( 1 29). In a more
422 PLE U RAL D I S EASES
recent study of 19 infants with congenital chylotho with chylothorax at the Mayo clinic during this time
rax, the mortality was zero ( 1 30) . In this latter series, ( 1 39). In a series of 50 patients with LAM from the
polyhydramnios was present in 1 0 of the pregnancies United Kingdom, 1 1 (22%) had a chylothorax ( 1 40) .
( 1 30) . All 1 9 patients had respiratory distress, 1 5 at The proliferation of smooth muscle in the perivas
birth, and the others at a subsequent time. In this cular spaces may obstruct the pulmonary venules and
series, 1 6 of the patients required mechanical ventila may produce pulmonary hemorrhage, hemoptysis, and
tion ( 1 30) . pulmonary hemosiderosis. The proliferation of peri
The recommended management of congenital bronchial smooth muscles can partially or completely
chylothorax is as follows. Initially, the baby should be obstruct the airways to cause air trapping, cyst and bul
treated conservatively with repeated thoracenteses. lae formation, and a high incidence of pneumothorax
In addition, the nutrition for the infant should be ( 1 4 1 ) . Approximately one third of patients have renal
provided by total parenteral nutrition ( 1 3 1 ) . An alter angiomyolipoma, which is a benign mesenchymal
native to TPN is to feed the baby with its mother's tumor ( 1 36) . Usually, the renal angiomyolipoma is
fat-free milk ( 1 32) . If the chylothorax recurs after discovered before the diagnosis of LAM is made.
the third pleural aspiration, a trial of octreotide It should be noted that there is now an International
should be given. Although there are no randomized LAM Registry (lamregistry.org) ( 1 42) . There is also a
controlled trials of octreotide, a review ( 1 33) of the LAM foundation in the United States (http://www.
literature in 20 1 0 found 1 9 case reports of 20 neo thelamfoundation.org) . ( 1 43) . The telephone number
nates and this treatment was successful in 1 4 (70%) . for the LAM Foundation is 1 -877-CURELAM .
However, Horvers et al. ( 1 34) treated 7 infants with
octreotide and concluded that no clear and consistent
C l i n ical M a n ifestations
effect of octreotide was identified. If the infant does
not respond to the octreotide, a pleuroperitoneal LAM occurs almost exclusively in women of repro
shunt should be inserted ( 1 3 5 ) . Milson et al. ( 1 3 5 ) ductive age ( 1 36) . In one series of 69 patients, how
implanted pleuroperitoneal shunts in seven infants, ever, 5 patients were postmenopausal at disease onset
one of whom was 7 days old, and reported that the ( 1 36) and there is a case report of LAM occurring in a
shunt cured the chylothorax in six of the seven 3-year-old girl ( 1 44) . The onset of symptoms can
patients. Thoracic duct ligation is indicated if the occur from age 3 to 70, but most patients are between
pleuroperitoneal shunting fails. The advantage of the the ages of 25 and 5 0 . Most patients have increasing
shunt over the thoracic duct ligation is that it is a shortness of breath and/or cough, but hemoptysis,
much simpler procedure. pneumothorax, or an incidentally discovered chylo
thorax can be the presenting manifestation. During
the course of their disease, almost all patients have
P U L M O N A RY
parenchymal infiltrates, 1 5 % to 30% have chylo
LY M P H A N G O L E I O M Y O M ATO S I S
thorax, and approximately 70% have pneumothorax
Pulmonary LAM is a rare condition characterized by ( 1 36, 1 37) .
widespread proliferation of immature smooth muscle Pulmonary LAM is at times part of the syndrome
throughout the peribronchial, perivascular, and peri of pulmonary tuberous sclerosis (TSC) . TSC is an
lymphatic regions of the lung ( 1 36, 1 37) . The lymph uncommon, genetically transmitted disease with
nodes in the mediastinum and retroperitoneal space the classic triad of seizures, adenoma sebaceum, and
may be infiltrated with immature smooth muscle mental retardation. In one study from the Mayo
cells impairing lymphatic flow. The thoracic duct Clinic, 20 of 76 patients (26%) with TSC had evi
may be either obliterated or dilated ( 1 3 8 ) . The peri dence of LAM ( 1 45 ) . The radiographic and patho
lymphatic proliferation of smooth muscle results in logic findings in the lung are similar in both disorders.
lymphatic obstruction. The preceding three factors Lymph nodes are less commonly involved, and
result in a significant incidence of chylothorax with chylothorax occurs less commonly with pulmonary
LAM . When two recent series with a total of 1 04 TSC, however ( 1 3 8 , 1 46) .
patients are combined, chylothorax occurred in 28 of Spontaneous LAM and the LAM associated with
the patients (27%) ( 1 36, 1 37) . However, only 8 of 79 TSC are linked genetically ( 1 47) . TSC patients
patients ( 1 0 . 1 %) seen at the Mayo Clinic with LAM have germline mutations in the TSC genes whereas
between 1 976 and 2000 had chylothoraces ( 1 39) . patients with spontaneous LAM do not have germ
These eight patients represented 3 . 5 % of all patients line mutations in the TSC genes, but rather have
C HAPT E R 26 I C HYLOTH O RAX A N D P S E U DOCHYLOTHO RAX 423
FIGURE 26.2 • Posteroa nterior rad i o g r a p h from FIGURE 26.3 • H i g h-reso l ut i o n com p uted
a 37-yea r-o ld wom a n with p u l mo n a ry lym p h a n g i tomogra p h i c sca n of 48-ye a r- o l d wo m a n with
o l e iomyomatos is. N ote the reticu lonod u l a r pattern lym p h a n g i o l e io myom atosis. N ote the n u m erous
and the wide d ista nce between the r i bs suggesti n g t h i n-wa l l e d cysts, ro u n ded and u n iform i n sha pe,
hyperi nflati o n . The p l e u r a l effu s i o n o n the l eft with n o rm a l i nterve n i n g l u n g p a renchym a .
was a chylothorax. (Courtesy of Dr Harry Sassoon.)
is a monoclonal antibody with specific immunoreactiv was much overlap between the groups ( 1 5 6) . Urban
ity for malignant melanoma ( 1 50) . LAM cell dusters et al. ( 1 36) reported that hormonal therapy was
can also be demonstrated in pleural fluid and they stain administered in 57 patients but that only 4 had a
positively with HMB45 ( 1 5 1 ) . The availability of this greater than 1 5 % improvement in their FEV1 • The
test will probably facilitate the use of transbronchial most common treatment regimen is medroxyproges
biopsy in making the diagnosis of LAM ( 1 52) . terone acetate, 400 mg intramuscularly each month.
Measurement of the serum vascular endothelial However, doses this suprapharmacologic are probably
growth factor-D (VEGF-D) appears useful in the diag not needed as the primary goal is to suppress serum
nosis of LAM. Young et al. ( 1 53) measured the VEGF estrogen production ( 1 47) . Norethindrone acetate,
D levels in 56 patients with LAM and 44 patients with 1 0 mg PO daily or twice a day is recommended
other cystic lung diseases and reported that the serum ( 1 47) . Estrogen-containing medications may have
VEGF-D levels in the patients with LAM were (median adverse effects and should be discontinued ( 1 47) .
1 , 1 75 [interquartile range (IQR) : 780-2,0 1 3] pg/mL) One therapy that should be considered for patients
were significantly higher than those in the other patients with LAM is sirolimus. The theory behind sirolimus
(median 2 8 1 [IQR 203-3 5 1 ] pg/ml) . Patients with is that it inhibits mammalian target of rapamycin
TSC-LAM had significantly higher levels than patients (mTOR) which is upgraded due to mutations of the
with TSC only ( 1 53). These results were confirmed TSC l and TSC2 gene and leads to abnormal cellular
in a subsequent study ( 1 54) , but in the second study growth ( 1 57) . Taveira-DaSilva et al. ( 1 57) treated 1 9
VEGF-D levels were not increased in patients that had patients with rapidly progressing LAM with sirolimus
LAM involving only the lung. and reported that the mean declines in the FEV1 and
DLCO were 2 . 8 % and 4.8% per year, respectively,
before treatment and both actually increased after
Treatment
treatment to 1 . 8% and 0 . 8 % per year, respectively.
In the past, it was believed that the prognosis of In this study, 1 1 patients had a chylothorax for at
women with pulmonary LAM was dismal because least 1 . 5 years before therapy with sirolimus and the
most patients died within 1 0 years of the onset of chylothorax disappeared completely in 9 with siro
symptoms ( 1 46) . More recent reports suggest that the limus. This study suggests that sirolimus should be
prognosis may be better to some extent. In a study considered in patients with LAM with chylothorax.
from France, the Kaplan-Meier plot showed survival One possible therapy for LAM is lung trans
probabilities of 9 1 % after 5 years, 79% after 1 0 years, plantation. In 2009, Benden et al. ( 1 58) reviewed
and 71 % after 1 5 years of disease duration ( 1 36) . 6 1 patients who received lung transplantation for
Patients die primarily of progressive respiratory insuf LAM at 30 centers in Europe between 1 997 and
ficiency ( 1 52) . Characteristics associated with a poor 2007. Prior to surgery, the mean FEV1 and FVC
prognosis include a reduced forced expiratory volume were 27% and 52% of the predicted, respectively.
in 1 second (FEV 1 ) and forced vital capacity (FVC) , They reported that the actuarial Kaplan-Meier sur
an increased TLC, and a predominantly cystic type vival was 79% at 1 year and 73% at 3 years post
of histology as opposed to a predominantly smooth transplant ( 1 5 8 ) . Recurrence of LAM occurred in 4
muscle type of histology ( 1 5 5 ) . recipients ( 1 5 8 ) . The results of Benden et al. ( 1 58)
Several researchers have suggested that hormonal are better than those reported by Boehler et al. ( 1 5 8a)
manipulation may be of value in treating this disease in 1 996 in which actuarial survival calculated by
( 1 36) . The almost exclusive occurrence of the disease the Kaplan-Meier method was 69% after 1 year and
in women of reproductive age has suggested that 5 8 % after 2 years. Given the relatively good progno
the smooth muscle proliferation may be hormon sis of the patients in the study by Urban et al. ( 1 36),
ally dependent. The results, however, with hormonal lung transplantation is not recommended until the
manipulation have not been impressive. Johnson and patient becomes debilitated from her disease.
Tattersfield ( 1 5 6) retrospectively analyzed the rate of The chylothorax associated with LAM should be
decline in pulmonary function of 43 patients with managed as are chylothoraces secondary to other
LAM . They reported that the mean drop in FEV1 and medical conditions. If the chylothorax is small and
diffusing capacity was significantly less in patients is asymptomatic, it need not be treated. If it is larger,
receiving progesterone than in those not receiving it is probably best treated with thoracoscopic liga
progesterone, and in postmenopausal women as tion of the thoracic duct or the implantation of
opposed to premenopausal women, although there a pleuroperitoneal shunt.
C HAPT E R 26 I C HYLOTH O RAX A N D P S E U DOCHYLOTHO RAX 42 5
Because the visceral pleura is usually thickened, the or she has been treated previously or has received the
underlying lung contributes minimally to the total bacille Calmette-Guerin vaccine.
ventilation, and the patient may have dyspnea on exer If the patient's exercise capacity is limited by short
tion. Chyliform pleural effusions are usually unilateral. ness of breath, a therapeutic thoracentesis should
be performed. Hillerdal reported that the removal
D i a g nosis of several hundred milliliters of pleural fluid from
patients with pseudochylothorax resulted in a mark
The diagnosis of chyliform pleural effusion is not
edly improved exercise tolerance ( 1 63) . Decortication
usually difficult. When a patient with a long-standing
should be considered if the patient is symptomatic
pleural effusion is found to have turbid or milky
and the underlying lung is believed to be functional
pleural fluid, the two other diagnostic possibilities are
( 1 67) . The decortication may result in a markedly
empyema and chylothorax. In an empyema, centrif
improved functional status for the patient ( 1 67) .
ugation results in a clear supernatant. The differen
tiation between chylothorax and pseudochylothorax
is not usually difficult; the patient with chylothorax
REFERENCES
has an acute pleural effusion and normal pleural sur
faces, whereas the patient with pseudochylothorax 1 . Romero S. Nontraumatic chylothorax. Curr Opin Pulm Med.
has a chronic pleural effusion and a thickened or 2000;6:287-29 1 .
calcified pleura ( 1 66) . On occasion, fat-fluid or fat 2 . Miller JI. Diagnosis and management of chylothorax. Chest
Surg Clin North Am. 1 996;6: 1 39- 1 4 8 .
calcium levels can be seen on the CT scan in patients 3 . Hillerdal G . Chylothorax and pseudochylothorax. Eur Respir
with a pseudochylothorax ( 1 66). J 1 997; 1 0 : 1 1 57- 1 1 62.
Analysis of the pleural fluid is useful in the differ 4. Merrigan BA, Winter DC, O 'Sullivan GC. Chylothorax. Br J
entiation of chylothorax and pseudochylothorax. If Surg. 1 997;84: 1 5-20 .
5 . Lampson R S . Traumatic chylothorax: a review of the literature
cholesterol crystals are seen on smears of the sediment,
and report of a case treated by mediastinal ligation of the tho
the patient has a chyliform pleural effusion. The cho racic duct. J Thorac Surg. 1 948; 1 7:778-79 1 .
lesterol crystals give a distinct, satin-like sheen to the 6 . Bower GC. Chylothorax: observations in 2 0 cases. Dis Chest.
pleural fluid. Microscopically, the cholesterol crystals l 964;46:464-468.
present a typical rhomboid configuration (Fig. 26. 1 ) . 7. Williams KR, Burford TH. The management of chylothorax.
Ann Surg. 1 964; 1 60 : 1 3 1 - 1 40 .
I f n o cholesterol crystals are seen, the patient may still
8 . Breaux JR, Marks C. Chylothorax causing reversible T-cell
have a chyliform effusion. Pleural fluid cholesterol depletion. J Trauma. 1 9 88;28 :705-707.
levels above 200 mg/dL strongly suggest a chyliform 9. Hodges CC, Fossum TW, Evering W. Evaluation of thoracic
effusion ( 1 60) . The cholesterol levels in the pleural duct healing after experimental laceration and transection.
fluid are elevated in high-lipid pleural effusions owing Vet Surg. 1 9 93;22:43 1 -43 5 .
1 0 . Staats BA, Ellefson RW, Budahn LL, e t al. Th e lipoprotein
to high numbers of cholesterol crystals or lecithin
profile of chylous and non-chylous pleural effusions. Mayo
globulin complexes (59), but cholesterol levels may Clin Proc. 1 9 80;55:700-704.
also be elevated in chylous pleural effusions ( 1 0) . 1 1 . Roy PH, Carr DT, Payne WS . The problem of chylothorax.
Lipoprotein analysis should b e performed i f any Mayo Clin Proc. 1 967;42:457-467.
doubt exists as to whether the fluid is chylous or pseu 1 2 . Strausser JL, Flye MW. Management of non-traumatic chylo
thorax. Ann Thorac Surg. 1 98 1 ;3 1 : 520-526.
dochylous because only chylous pleural fluid contains 13. Doerr CH, Allen MS, Nichols FC III, et al . Etiology of chylo
chylomicrons ( 1 0,64) . Some chyliform effusions have thorax in 203 patients. Mayo Clin Proc. 2005;80: 867-870.
high ( > 2 5 0 mg/dL) triglyceride levels (64) , so this 14. Huggins JT. Chylothorax and cholesterol pleural effusion.
finding is not diagnostic of chylothorax. Semin Respir Crit Care Med. 2 0 1 0;3 1 :743-750.
15. Johnstone DW. Anatomy of the thoracic duct and chylotho
rax. In: Shields TW, Locicero J III, Reed CE, Feins RH eds.
Treatment General Thoracic Surgery, 7th ed. Philadelphia, PA: Wolters
Kluwer/Lippincott Williams & Wilkins; 2007:827-833.
When a patient is diagnosed as having a chyliform
16. Maloney JV, Spencer FC. The non-operative treatment of
pleural effusion, the possibility of tuberculosis should traumatic chylothorax. Surgery. 1 956;40: 1 2 1 - 1 2 8 .
always be entertained. If the patient has a history of 1 7 . Shimizu K , Yoshida ], Nishimura M, e t al. Treatment strategy
tuberculosis and has never been treated with antitu for chylothorax after pulmonary resection and lymph node
berculous therapy, isoniazid and rifampin should be dissection for lung cancer. ] Thorac Cardiovasc Surg. 2002;
1 24:499-502.
given for at least 9 months. Similarly, if the patient
1 8 . Chan EH, Russell JL, Williams WG, et al. Postoperative chy
has a positive purified protein derivative test, he or lothorax after cardiothoracic surgery in children. Ann Thorac
she should be treated with these drugs unless he Surg. 2005;80: 1 864- 1 870.
C HAPT E R 26 I C HYLOTH O RAX A N D P S E U DOCHYLOTHO RAX 427
1 9 . Shah RD, Lukecich JD, Schuchert MJ, et al. Poscesophagec 42. Faul JL, Berry GJ, Colby TV, et al. Thoracic lymphangio
tomy chylochorax: incidence, risk factors, and outcomes. Ann mas, lymphangiectasis, lymphangiomatosis, and lymphatic
Thorac Surg. 2 0 1 2;93 : 897-904. dysplasia syndrome. Am J Respir Crit Care Med. 2000;
20. Dougenis D, Walker WS, Cameron EW, et al. Management of 1 6 1 : 1 037- 1 046.
chylochorax complicating extensive esophageal resection. Surg 43. Feldman L, Kleiner-Baumgarten A, Maislos M . Large peri
Gynecol Obstet. 1 992; 1 74 : 5 0 1 -506. cardia! and pleural effusions associated with familial lymph
2 1 . Lai FC, Chen L, Tu YR, et al. Prevention of chylochorax edema. lsr Med Assoc]. 200 1 ;3:769-770.
complicating extensive esophageal resection by mass ligation 44. Konishi T, Takeuchi H, Iwata J, et al. Behyet's disease with
of thoracic duce: a random control study. Ann Thorac Surg. chylochorax-case report. Angiology. 1 9 8 8;39:68-7 1 .
20 1 1 ; 9 1 : 1 770-1 774. 4 5 . Coplu L , Emri S , Selcuk ZT, e t al. Life threatening chylous
22. Zaidenstein R, Cohen N, Dishi V, et al. Chylothorax follow pleural and pericardia! effusion in a patient with Behyet's
ing median scernotomy. Clin Cardiol. 1 996; 1 9 : 9 1 0-9 1 2 . syndrome. Thorax. 1 992;47:64-6 5 .
2 3 . Twomey CR. Chylothorax i n the adult heart transplant 4 6 . Vennera M C , Moreno R , C o t J, e t al. Chylothorax and tuber
patient: a case report. Am ] Crit Care. 1 994;3:3 1 6-3 1 9 . culosis. Thorax. l 983;38:694-695.
24. Dupont PA. Chylothorax after high cranslumbar aortography. 47. Parker JM, Torrington KG, Phillips YY. Sarcoidosis compli
Thorax. 1 975;30: 1 1 0- 1 1 2 . cated by chylothorax. South Med]. l 994;87:860-862.
2 5 . Nygaard SD, Berger HA, Fick R B . Chylochorax a s a complica 48. Dagenais F, Ferraro P, Duranceau A. Spontaneous chylochorax
tion of oesophageal sclerotherapy. Thorax. 1 992;47: 1 34- 1 3 5 . associated with primary lymphedema and a lymphangioma
2 6 . Nagai H, Shimizu K , Shikata J, e t al. Chylous leakage after malformation. Ann Thorac Surg. 1 999;67: 1 480- 1 482.
circumferential thoracolumbar fusion for correction of 49. Bresser P, Kromhout JG, Reekers JA, et al. Chylous pleural effu
kyphosis resulting from fracture. Report of three cases. Spine. sion associated with primary lymphedema and lymphangioma
1 997;22:2766-2769 . like malformations. Chest. 1 993; 1 03 : 1 9 1 6- 1 9 1 8 .
2 7 . Thorne PS. Traumatic chylothorax. Tubercle. l 9 5 8 ;39 :29-34. 5 0 . Lee YC, Tribe AE, Musk AW Chylochorax from radiation
28. Cammarata SK, Brush RE Jr. , Hyzy RC. Chylothorax after induced mediastinal fibrosis. Aust N Z J Med. 1 998;28:
childbirth. Chest. 1 9 9 1 ;99: 1 539-1 540. 667-668.
29. Reilly KM, Tsou E. Bilateral chylothorax: a case report follow 5 1 . Kollef MH. Recalcitrant chylothorax and chylous ascites asso
ing episodes of screeching. JAMA. 1 975 ;233 : 536-537. ciated with hypothyroidism. Milit Med. 1 993; 1 5 8 :63-6 5 .
30. Berman W Jr, Fripp RR, Yabek SM, et al. Great vein and right 52. Chernick V, Reed MH. Pneumothorax and chylothorax i n the
atrial thrombosis in critically ill infants and children with cen neonatal period. ] Pediatr 1 970;76:624-632.
tral venous lines. Chest. 1 9 9 1 ;99 : 963-967. 53. Van Aerde J, Campbell AN, Smyth JA, et al. Spontaneous
3 1 . Rice Tw, Rodriguez RM, Barnette R, et al. The prevalence chylothorax in newborns. Am] Dis Child. 1 984; 1 3 8:96 1-964.
and characteristics of pleural effusions in superior vena cava 54. van Scraacen HL, Gerards LJ, Krediet TG. Chylochorax in the
syndrome. Respirology. 2006; 1 1 :299-305. neonatal period. Eur J Pediatr. 1 993; 1 52:2-5 .
32. Thomas R, Christopher DJ, Roy A, et al. Chylothorax fol 5 5. Huang XZ, Wu JF, Ferrando R, et al. Fatal bilateral chylotho
lowing innominate vein thrombosis-a rare complication rax in mice lacking the integrin a9 {3 1 • Mo! Cell Biol. 2000;
of transvenous pacemaker implantation. Respiration. 2005; 20:5208-52 1 5 .
72:546-54 8 . 5 6 . Chamberlain M , Ratnatunga C . Late presentation o f tension
33. Van Veldhuizen PJ, Taylor S . Chylochorax: a complication chylochorax following blunt chest trauma. Eur J Cardiothorac
of a left subclavian vein thrombosis. Am ] Clin Oncol. 1 996; Surg. 2000; 1 8 :3 57-3 5 9 .
1 9 :99-1 0 1 . 57. Milano S, Maroldi R , Vezzoli G, e t al. Chylothorax after blunt
34. Romero S , Marcin C , Hernandez L , et al. Chylochorax in chest trauma: an unusual case with a long latent period. Thorac
cirrhosis of the liver: analysis of its frequency and clinical char Cardiovasc Surg. 1 994;42: 1 87- 1 90 .
acteristics. Chest. 1 998 ; 1 1 4: 1 54- 1 5 9 . 5 8 . Wasmuth-Pietzuch A , Hansmann M, Bartmann P, et al. Con
3 5 . Villena V, d e Pablo A , Marcin-Escribano P. Chylochorax genital chylothorax: lymphopenia and high risk of neonatal
and chylous ascites due to heart failure. Eur Respir ]. 1 99 5 ; infections. Acta Paediatr 2004;93:220-224.
8 : 1 235-1 236. 59. Hughes RL, Mintzer RA, Hidvegi DF, et al. The management
36. Hanna J, Truemper E, Burton E. Superior vena cava throm of chylochorax. Chest. 1 979;76:2 1 2-2 1 8 .
bosis and chylothorax: relationship in pediatric nephrotic 60. Maldonado F, Hawkins FJ, Daniels CE, et al. Pleural fluid char
syndrome. Pediatr Nephrol. 1 997; 1 1 :20-22. acteristics of chylochorax. Mayo Clin Proc. 2009;84: 1 29- 1 33 .
37. Tie MLH, Poland GA, Rosenow EC III. Chylothorax in 6 1 . Diaz-Guzman E , Culver DA, Stoller J K . Transudative chylo
Gorham's syndrome. A common complication of a rare chorax: report of two cases and review of the literature. Lung.
disease. Chest. 1 994; 1 05 :208-2 1 3 . 2005 ; 1 83: 1 69- 1 7 5 .
3 8 . Pennington DW, Warnock ML, Stulbarg M S . Chylochorax 6 2 . Vaz MAC, Teixeira LR, Vargas FS, et al. Relationship between
and respiratory failure in Kaposi's sarcoma. WestJ Med. 1 990; pleural fluid and serum cholesterol levels. Chest. 200 1 ;
1 5 2:42 1 -422. 1 1 9 :204-2 1 0 .
39. Judson MA, Postic B . Chylothorax in a patient with AIDS 63. Wolthuis A, Landewe RB, Theunissen PH, et al. Chylothorax
and Kaposi's sarcoma. South Med]. l 990;83:322-324. or leakage of total parenteral nutrition? Eur Respir ]. 1 998;
40. Wright RS, Jean M, Rochelle K, et al. Chylochorax caused by 1 2: 1 233- 1 23 5 .
Paragonimus westermani in a native California. Chest. 20 1 1 ; 6 4 . Coe JE, Aikawa JK. Cholesterol pleural effusion. Arch Intern
1 40 : 1 064- 1 066. Med. 1 96 1 ; 1 08 :763-774.
4 1 . Blankenship ME, Rowlett J, Timby JW, et al. Giant lymph 65. Klepser RG, Berry JF. The diagnosis and surgical manage
node hyperplasia (Castleman's disease) presenting with chylous ment of chylothorax with the aid of lipophilic dyes. Dis Chest.
pleural effusion. Chest. 1 997; 1 1 2 : 1 1 32-1 1 33. 1 9 54;25 :409-426.
428 PLE U RAL D I S EASES
66. Stavngaard T, Mortensen J, Brenoe J, et al. Lymphoscintigraphy 8 9 . Makrilakis K, Pavlatos S, Giannikopoulos G, et al. Successful
using technetium-99m human serum albumin in chylothorax. octreotide treatment of chylous pleural effusion and lymph
Thorac Cardiovasc Surg. 2002;50:250-252. edema in the yellow nail syndrome. Ann Intern Med. 2004;
67. Restrepo JM, Caride VJ. Lymphoscintigraphy and radionuclide 1 4 1 :246-247.
venography in chylothorax. Clin Nucl Med. 2004;29 :440-44 1 . 90. Demos NJ, Kozel J, Scerbo JT. Somatostatin in the treat
6 8 . Qureshy A, Kubota K, Ono S , et al. Thoracic duct scintig ment of chylothorax. Chest. 200 1 ; 1 1 9 : 964-966.
raphy by orally administered I- 1 23 BMIPP: normal findings 9 1 . Kalomenidis I. Octreotide and chylothorax. Cur Opin Pul
and a case report. Clin Nucl Med. 200 1 ;26:847-8 5 5 . Dis. 2006; 1 2:264-267.
6 9 . Thomson IA , Simms M H . Postoperative chylothorax: a case 92. Nakabayashi H, Sagara H, Usukura N, et al. Effect of soma
for recycling? Cardiovasc Surg. 1 993; 1 :384-3 8 5 . tostatin on the flow rate and triglyceride levels of thoracic
70. Ross J K . A review o f the surgery of the thoracic duct. Thorax. duct lymph in normal and vagotomized dogs. Diabetes. 1 98 1 ;
1 96 1 ; 1 6 : 1 2-2 1 . 30:440-445 .
7 1 . Valentine VG, Raffin TA. The management o f chylothorax. 93. Mikroulis D, Didilis V, Bitzikas G, e t al. Octreotide i n the
Chest. 1 992; 1 02:5 86-5 9 1 . treatment of chylothorax. Chest. 2002; 1 2 1 :2079-2080.
72. Lichter I , Hill GL, Nye ER. The use of medium-chain trigly 94. Maayan-Metzger A, Sack J, Mazkereth R. Somatostatin
cerides in the treatment of chylothorax in a child. Ann Thorac treatment of congenital chylothorax may induce transient
Surg. 1 968;4:352-3 5 5 . hypothyroidism in newborns. Acta Paediatr. 2005;94:
7 3 . Berkenbosch J W, Withington D E . Management o f postop 78 5-789 .
erative chylothorax with nitric oxide: a case report. Crit Care 9 5 . Mohseni-Bod H, Macrae D , Slavik Z. Somatostatin ana
Med. 1 999;27: 1 022- 1 024. log (octreotide) in management of neonatal postoperative
74. Nguyen OM, Shum-Tim D, Dobell AR, et al. The manage chylothorax: is it safe? Pediatr Crit Care Med. 2004 ; 5 :
ment of chylothorax/chylopericardium following pediatric 356-3 57.
cardiac surgery: a I O-year experience. ] Cardiac Surg. 1 99 5 ; 96. Little AG, Kadowaki MH, Ferguson MK, et al. Pleuroperi
1 0 : 302-308. toneal shunting: alternative therapy for pleural effusions.
7 5 . Cerfolio RJ, Allen MS, Deschamps C, et al. Postoperative Ann Surg. 1 9 88;208 :443-450.
chylothorax. J Thorac Cardiovasc Surg. 1 996; 1 1 2 : 1 3 6 1 - 1 36 5 . 97. Murphy MC, Newman BM, Rodgers BM. Pleuroperitoneal
76. L e Pimpec-Barthes F, D'Attellis N, Dujon A, e t al. Chylothorax shunts in the management of persistent chylothorax. Ann
complicating pulmonary resection. Ann Thorac Surg. 2002;73 : Thorac Surg. 1 989;48 : 1 95-200.
1 7 1 4- 1 7 1 9 . 9 8 . Rheuban KS , Kron IL, Carpenter MA, et al. Pleuroperito
77. Goto M, Kawamata K , Kitano M, e t al. Treatment o f chylo neal shunts for refractory chylothorax after operation for
thorax in a premature infant using somatostatin. ] Perinatol. congenital heart disease. Ann Thorac Surg. l 992; 5 3 : 8 5-87.
2003;23 : 5 63-5 64. 99. Engum SA, Rescorla FJ, West Kw; et al. The use of pleuro
78. Rasiah S, Oei J, Lui K. Octreotide in the treatment of con peritoneal shunts in the management of persistent chylotho
genital chylothorax. ] Paediatr Child Health. 2004;4 : 5 8 5-5 8 8 . rax in infants. ] Pediatr Surg. l 999;34:286-290.
79. Goyal A , Smith NP, Jesudason E C , e t al. Octreotide for treat 1 00. Itkin M, Kucharczuk JC, Kwak A, et al. Nonoperative tho
ment of chylothorax after repair of congenital diaphragmatic racic duct embolization for traumatic thoracic duct leak: expe
hernia. ] Pediatr Surg. 2003;3 8 : E l 9-E20 . rience in I 09 patients. ] Thorac Cardiovasc Surg. 20 1 O ; 1 39 :
8 0 . Tibballs J, Soto R , Bharucha T. Management of newborn 5 84-5 8 9 .
lymphangiectasia and chylothorax after cardiac surgery with 1 0 1 . Boffa DJ, Sands MJ, Rice Tw; et a l . A critical evaluation
octreotide infusion. Ann Thorac Surg. 2004;77:22 1 3-22 1 5 . of a percutaneous diagnostic and treatment strategy for
8 1 . Rosti L , D e Battisti F, Butera G , e t al. Octreotide i n the chylothorax after thoracic surgery. Eur J Cardiothorac Surg.
management of postoperative chylothorax. Pediatr Cardiol. 2008;33 :43 5-439 .
2005;26:440-443. 1 02. Itkin M, Krishnamurthy G, Nairn MY, et a l . Percutaneous
82. Hamdan MA, Gaeta ML. Octreotide and low-fat breast milk thoracic duct embolization as a treatment for intrathoracic
in postoperative chylothorax. Ann Thorac Surg. 2004;77: chyle leaks in infants. Pediatrics. 20 1 1 ; 1 28:e237-e24 1 .
22 1 5-22 1 7 . 1 03 . Laslett D , Trerotola SO, Itkin M . Delayed complications
8 3 . Pratap U, Slavik Z, O fo e V D , e t al. Octreotide t o treat post following technically successful thoracic duct embolization.
operative chylothorax after cardiac operations in children. ] Vase Interv Radio!. 2 0 1 2;23:76-79.
Ann Thorac Surg. 200 1 ;72: 1 740- 1 742. 1 04. Gaba RC, Owens CA, Bui JT, et al. Chylous ascites: a rare
84. Buettiker V, Hug Ml, Burger R, et al. Somatostatin: a new complication of thoracic duct embolization for chylothorax.
therapeutic option for the treatment of chylothorax. Intensive Cardiovasc Intervent Radio!. 2 0 1 0;34 (suppl 2):S2 1 8-S223.
Care Med. 200 1 ;27: 1 083- 1 086. 1 0 5 . Shimizu J, Hayashi Y, Oda M, et al. Treatment of postop
85. Caverly L, Rausch CM, Da Cruz E, et al. Octreotide treat erative chylothorax by pleurodesis with the streptococcal
ment of chylothorax in pediatric patients following cardiotho preparation OK-432. Thorac Cardiovasc Surg. 1 994;42:
racic surgery. Congenit Heart Dis. 2 0 1 0 ; 5 : 5 73-578. 233-236.
86. Gomez-Caro Andres A, Marron Fernandez C, Moradiellos 1 06. Adler RH, Levinsky L. Persistent chylothorax. ] Thorac
Diez FJ, et al. Octreotide for conservative management Cardiovasc Surg. 1 978;76 : 8 5 9-863.
of postoperative chylothorax. Arch Bronconeumol. 2004;40: 1 07. Akaogi E, Mitsui K, Sohara Y, et al. Treatment of postop
473-475 . erative chylothorax with intrapleural fibrin glue. Ann Thorac
87. Gabbieri D, Bavutti L, Zaca F, et al. Conservative treatment Surg. 1 989;48 : 1 1 6- 1 1 8 .
of postoperative chylothorax with octreotide. Ital Heart ]. 1 0 8 . Meurer MF, Cohen DJ . Current treatment o f chylothorax: a
2004;5 :479-482. case series and literature review. Texas Med. 1 990;86: 82-8 5 .
88. Demos NJ . Octreotide in the treatment of chylothorax. Chest. 1 09 . Robinson CLN. Th e management o f chylothorax. Ann Thorac
2002; 1 2 1 :2080-208 1 . Surg. 1 98 5 ;39 : 90-9 5 .
C HAPT E R 26 I C HYLOTH O RAX A N D P S E U DOCHYLOTHO RAX 429
1 1 0. Huang PM, Lee YC. A new technique of continuous pleural 1 3 1 . Fernandez Alvarez JR, Kalache KD, Grauel EL. Manage
irrigation with minocycline administration for refractory ment of spontaneous congenital chylothorax: oral medium
chylothorax. Thorac Cardiovasc Surg. 2 0 1 1 ; 5 9 :436--43 8 . chain triglycerides versus total parenteral nutrition. Am J
1 1 1 . Weissberg D, Ben-Zeev I. Talc pleurodesis. Experience with Perinatal. 1 999; 1 6 : 4 1 5-420.
360 patients. ] Thorac Cardiovasc Surg. 1 993; 1 06:689-695 . 1 32 . Chan GM, Lechtenberg E. The use of fat-free human milk
1 1 2 . Vargas F S , Milanez JRC, Filomeno LTB, e t al. Intrapleural in infants with chylous pleural effusion. J Perinatal. 2007;
talc for the prevention of recurrence in benign or undiag 27:434--436.
nosed pleural effusions. Chest. 1 994; 1 06: 1 77 1 - 1 775. 1 33 . Das A, Shah PS. Octreotide for the treatment of chylothorax
1 1 3 . Graham DD, McGahren ED, Tribble CG, et al. Use of in neonates. Cochrane Database Syst Rev. 2 0 1 0;9 :CD0063 8 8 .
video-assisted thoracic surgery in the treatment of chylotho 1 34 . Horvers M, Mooij CF, Antonius TA. I s octreotide treatment
rax. Ann Thorac Surg. 1 994;57: 1 5 07- 1 5 1 1 . useful in patients with congenital chylothorax? Neonatology.
1 1 4. Katanyuwong P, Dearani J , Driscoll D . The role of 2 0 1 2 ; 1 0 1 :225-23 1 .
pleurodesis in the management of chylous pleural effusion 1 3 5 . Milson JW, Kron IL, Rheuban KS, et al. Chylothorax: an
after surgery for congenital heart disease. Pediatr Cardiol. assessment of current surgical management. J Thorac Cardio
2009;30: 1 1 1 2- 1 1 1 6. vasc Surg. 1 98 5 ; 89:22 1 -227.
1 1 5 . Paul S, Altorki NK, Port JL, et al. Surgical management of 1 36. Urban T, Lazor R, Lacronique J, et al. Pulmonary lymphan
chylothorax. Thorac Cardiovasc Surg. 2009;57:226-228. gioleiomyomatosis. A study of 69 patients. Medicine (Balti
1 1 6. Nath DS, Savla J, Khemani RG, et al. Thoracic duct ligation more). 1 999;78:32 1 -337.
for persistent chylothorax after pediatric cardiothoracic sur 1 37. Chu SC, Horiba K, Usuki J, et al. Comprehensive evaluation
gery. Ann Thorac Surg. 2009;88:246-25 1 . of35 patients with lymphangioleiomyomatosis. Chest. 1 999;
1 1 7. Mason PF, Ragoowansi RH, Thorpe JA. Post-thoracotomy 1 1 5 : 1 04 1 - 1 052.
chylothorax-a cure in the abdomen? Eur J Cardiothorac 138. Silverstein EF, Ellis K, Wolff M, et al. Pulmonary lymphan
Surg. 1 997; 1 1 : 5 67-570. giomyomatosis. A]R Am ] Roentgenol. 1 974; 1 20 : 832-8 5 0 .
1 1 8 . Yagihara M, Miyabe M, Mizutani T, et al. Intraduodenal milk 1 3 9 . Ryu JH, Doerr CH, Fisher SD, et al. Chylothorax i n
injection after induction of general anesthesia is safe and lymphangioleiomyomatosis. Chest. 2003; 1 23:623-627.
useful during surgical treatment for intractable chylothorax. 1 40 . Johnson SR, Tattersfield AE. Clinical experience of
Anesth Analg. 200 5 ; 1 0 1 : 1 8 9 1 . lymphangioleiomyomatosis in the UK. Thorax. 2000; 5 5 :
1 1 9 . Fahimi H , Casselman F P, Mariani MA, e t al. Current man 1 052- 1 057.
agement of postoperative chylothorax. Ann Thorac Surg. 1 4 1 . Carrington CB, Cugell DW, Gaensler EA, et al. Lymphangi
200 1 ;7 1 :448--450. oleiomyomatosis. Am Rev Respir Dis. 1 977; 1 1 6:977-9 9 5 .
1 20. Wurnig PN, Hollaus PH, Ohtsuka T, et al. Thoracoscopic 1 42 . Nurok M, Eslick I, Carvalho CR, et al. The international
direct clipping of the thoracic duct for chylopericardium and LAM registry: a component of an innovative web-based cli
chylothorax. Ann Thorac Surg. 2000;70: 1 662- 1 66 5 . nician, researcher, and patient-driven rare disease research
1 2 1 . Christodoulou M , Ris HB, Pezzetta E. Video-assisted platform. Lymphat Res Biol. 2 0 1 0 ; 8 : 8 1 -87.
right supradiaphragmatic thoracic duct ligation for non 143. Sullivan EJ, Beck GJ, Peavy HH, et al. Lymphangioleiomyo
traumatic recurrent chylothorax. Eur J Cardiothorac Surg. matosis registry. Chest. 1 999; 1 1 5 : 30 1 .
2006;29 : 8 l 0-8 1 4 . 1 44 . Nagy B , Nabrady Z , Nemes Z . Pulmonary lymphangio
1 22 . Pego-Fernandes P M , Nascimbem MB, Ranzani OT, e t al. myomatosis in a preadolescent girl. N Engl J Med. 1 99 8 ;
Video-assisted thoracoscopy as an option in the surgical 338:473--474 .
treatment of chylothorax after cardiac surgery in children. 1 4 5 . Costello L C , Hartman T E , Ryu JH. High frequency of pul
J Bras Pneumo!. 20 1 1 ;37:28-3 5 . monary lymphangioleiomyomatosis in women with tuber
1 2 3 . Pui M H , Yueh TC. Lymphoscintigraphy i n chyluria, chyloperi ous sclerosis complex. Mayo Clin Proc. 2000;75 : 5 9 1 -594.
toneum and chylothorax. J Nucl Med. 1 998;39: 1 292-1 296. 1 46. Corrin B, Liebow AA, Friedman PJ . Pulmonary lymphan
1 24. Mares DC, Mathur PN. Medical thoracoscopic talc giomyomatosis. Am J Pathol. 1 975 ;79:348-367.
pleurodesis for chylothorax due to lymphoma: a case series. 1 47. McCormack FX, Sullivan EJ. Lymphangioleiomyomatosis.
Chest. 1 998 ; 1 1 4:73 1 -735. In Mason RJ, Broaddus VC, Murray JF, et al. eds. Murray
125. Jimenez CA, Mhatre AD, Martinez CH, et al. Use of indwell and Nadel's Textbook of Respiratory Medicine, 4th ed, Phila
ing pleural catheter for the management of recurrent chylo delphia, PA: Elsevier Science; 2005 : 1 702- 1 70 5 .
thorax in patients with cancer. Chest. 2007; 1 32 : 1 584- 1 590. 1 4 8 . Fraser R S , Muller N L , Colman N, e t a l . Diagnosis of Dis
1 26. Moss R, Hinds S, Fedullo AJ . Chylothorax: a complication eases of the Chest, 4th ed. Philadelphia, PA: WB Saunders;
of the nephrotic syndrome. Am Rev Respir Dis. 1 9 89; 1 40: 2000: 679-686.
1 436- 1 437. 1 4 9 . Taylor JR, Ryu J, Colby TY, et al. Lymphangioleio
1 27. Veroux P, Veroux M, Bonanno MG, et al. Long-term suc myomatosis: clinical course in 32 patients. N Engl ] Med.
cess of endovascular treatment of benign superior vena cava 1 990;323 : 1 254-1 260.
occlusion with chylothorax and chylopericardium. Eur 1 5 0 . Tanaka H, Imada A, Morikawa T, et al. Diagnosis of pulmo
Radial. 2002; 1 2 (suppl 4):S l 8 1 -S l 84. nary lymphangioleiomyomatosis by HMB45 in surgically
128. Vignaux 0, Gouya H, Dousset B, et al. Refractory chylotho treated spontaneous pneumothorax. Eur Respir J 1 99 5 ; 8 :
rax in hepatic cirrhosis: successful treatment by transjugular 1 879- 1 882.
intrahepatic portosystemic shunt. ] Thorac Imaging. 2002; 1 5 1 . Hirama M, Atsuta R, Mitani K, et al. Lymphangioleiomyo
1 7 :233-236. matosis diagnosed by immunocytochemical and genetic
1 2 9 . Perry RE, Hodgman J, Cass AB. Pleural effusion in the analysis lymphangioleiomyomatosis cell clusters found in
neonatal period. ] Pediatr. l 963;62:838-843 . chylous pleural effusion. Intern Med. 2007;46: 1 593- 1 596.
1 30 . Al-Tawil K, Ahmed G, Al-Hathal M, et al. Congenital 1 52 . Miller WT, Cornog JL, Sullivan MA . Lymphangiomyoma
chylothorax. Am ] Perinatal. 2000; 1 7 : 1 2 1 - 1 26. tosis. A]R Am J Roentgenol. 1 972; 1 1 1 : 5 6 5-572.
43 0 PLE U RAL D I S EASES
1 53 . Young LR, Vandyke R, Gulleman PM, er al. Serum vascular 1 5 9 . Garcia-Zamalloa A, Ruiz-Irastorza G, Aguayo FJ, er al. Pseu
endothelial growth facror-D prospectively distinguishes dochylothorax. Report of2 cases and review of the literature.
lymphangioleiomyomarosis from other diseases. Chest. Medicine (Baltimore). 1 999;78:200-207.
20 1 0; 1 3 8 :674-68 1 . 1 60. Hamm H, Pfalzer B, Fabel H . Lipoprotein analysis in a
1 54. Glasgow CG, Avila NA, Lin J P, er al. Serum vascular endo chyliform pleural effusion: implications for pathogenesis
thelial growth factor-D levels in patients with lymphangi and diagnosis. Respiration. 1 9 9 1 ; 5 8 :294-300.
oleiomyomatosis reflect lymphatic involvement. Chest. 2009; 1 6 1 . Shen PU, Blair JL. Cholesterol crystals causing falsely
1 3 5 : 1 293- 1 300. elevated automated cell count. Am J Clin Pathol. 2006; 1 2 5 :
1 5 5 . Kiraichi M, Nishimura K, Itoh H, er al. Pulmonary lymphan 3 5 8-363.
gioleiomyomatosis: a report of 46 patients including a 1 62. Ferguson GC. Cholesterol pleural effusion in rheumatoid
clinicopathologic study of prognostic factors. Am J Respir lung disease. Thorax. 1 966;2 1 : 577-5 82.
Crit Care Med. 1 99 5 ; 1 5 1 : 527-5 33. 1 63 . Hillerdal G . Chyliform (cholesterol) pleural effusion. Chest.
1 56. Johnson SR, Tattersfield AE. Decline in lung function in 1 9 85;86:426-428.
lymphangioleiomyomatosis: relation to menopause and 1 64. Wrightson JM, Stanton AE, Maskell NA, et al. Pseudochylo
progesterone treatment. Am J Respir Crit Care Med. 1 999; rhorax without pleural thickening: time to reconsider patho
1 60:628-633. genesis? Chest. 2009; 1 36: 1 1 44- 1 147.
1 57. Taveira-Dasilva AM, Hathaway 0, Srylianou M, et al. 165. Johnson RJ, Johnson JR. Paragonimiasis in Indo-Chinese
Changes in lung function and chylous effusions in patients refugees: roentgenographic findings with clinical correla
with lymphangioleiomyomatosis treated with sirolimus. Ann tions. Am Rev Respir Dis. 1 983; 1 2 8 : 534-53 8 .
Intern Med. 20 1 1 ; 1 54:797-80 5 . 1 66. Song JW, I m JG, G o o J M , e r al. Pseudochylous pleural
1 5 8 . Benden C, Rea F, Behr J, e t a l . Lung transplantation fo r effusion with far-fluid levels: report of six cases. Radiology.
lymphangioleiomyomatosis: the European experience. J Heart 2000;2 1 6:478-480.
Lung Transplant. 2009;28 : 1 -7. 1 67. Goldman A, Burford TH. Cholesterol pleural effusion: a
l 5 8a. Boehler A, Speich R, Russi EW, er al. Lung transplanta report of three cases with a cure by decortication. Dis Chest.
tion for lymphangioleiomyomatosis. N Engl J Med. 1 996; 1 9 50; 1 8 : 5 86-594.
335: 1 275- 1 280.
Ot h e r P l e u ra l D i se a ses
43 1
432 PLE U RAL D I S EASES
The prevalence of pleural plaques was 22% in 91 ele pleural calcification in Finland and demonstrated
vator construction workers who probably had been that almost all such subjects lived near open asbestos
exposed to low levels of asbestos in their work (7) . pits. Hillerdal (6) reported that 88% of 1 , 596 adults
The incidence of pleural plaques at autopsy has varied older than 40 with pleural plaques had an occupa
from 0. 5% to 5 8 % (8,9) . When 1 6 separate studies tional exposure to asbestos. Many patients who have
with a total of 7,08 5 routine autopsies were combined, pleural plaques at autopsy have a work history in
the prevalence of pleural plaques was 1 2.2% (8) . The which asbestos exposure would be expected ( 1 5 , 1 6) .
standard chest radiograph identifies berween 50% and Most pleural plaques contain many submicroscopic
80% of the pleural plaques that are actually present (8) . asbestos fibers that can be demonstrated by transmis
Pleural plaques slowly develop in patients exposed sion electron microscopic examination, selective area
to asbestos. Epler et al. ( 1 0) reviewed the chest radio electron diffraction, and microchemical analysis of
graphs of 1 , 1 3 5 patients who had been exposed to particles ( 1 7, 1 8) . In a large study ( 1 9) from France,
asbestos and reported that none of the patients 5,545 patients with a history of asbestos exposure had
developed pleural plaques during the 1 0 years after a HRCT and 882 ( 1 6%) had pleural plaques. Patients
the initial exposure, and the incidence was still only who had pleural plaque had a longer time since their
approximately 1 0% 20 years after the initial exposure first exposure and a greater cumulative exposure.
( 1 0) . Forty years after the initial exposure, however, Ferruginous bodies (asbestos bodies) , long consid
more than 50% of the patients had radiologically ered the histologic hallmark of exposure to asbestos
visible pleural plaques. The mean duration berween (20), consist of fibers coated by complexes ofhemosid
the initial exposure to asbestos and the development erin and glycoproteins and are believed to be formed
of pleural plaques was 33 years in the series of Hill by macrophages that have phagocytized the particles.
erdal (6) . These plaques usually calcify within several Although these bodies have been shown to form from
years of becoming evident radiologically. Calcifica foreign inorganic and organic fibers of many differ
tion of the pleural plaques rarely occurs within the ent types, ferruginous bodies in most human lungs
first 20 years of initial exposure to asbestos, but by have asbestos as a core and are commonly known
40 years, more than one third of these individuals as asbestos bodies (20) . Patients with pleural plaques
have calcified pleural plaques ( 1 0) . have higher numbers of asbestos bodies in their lungs
Pleural plaques can also develop i n individu than do patients without pleural plaques ( 1 5 , 2 1 ,22) .
als who are not occupationally exposed to asbestos. Similarly, the higher the number of asbestos bodies
Kilburn et al. ( 1 1 ) reported that the prevalence of in the lungs, the more likely the presence of pleural
pleural abnormalities was 5 .4% in the chest radio plaques (20,22) . It should be noted, however, that
graphs of 280 wives of asbestos workers who were the number of asbestos fibers that are uncoated or
initially exposed to asbestos at least 20 years previ bare (and visible only on electron microscopy) exceeds
ously. Churg and DePaoli ( 1 2) reported four cases the number of asbestos bodies, which are visible by
of pleural plaques found at autopsy in individuals light microscopy, by 5- to 1 0,000-fold (23) .
who resided in or near the chrysotile mining town It appears that the various types of asbestos fibers
of Thetford Mines, Quebec, but who did not work differ in their ability to induce pleural plaques. Expo
with asbestos. Mineral analysis of the lungs revealed sure to crocidolite is most frequently associated with
that the individuals with pleural plaques had higher the production of pleural plaques. In North America,
levels of tremolite but comparable levels of chrysotile pleural plaques are more likely to result from tremo
than did the lungs of nine control subjects without lite than from chrysotile exposure. Churg et al. (24)
pleural plaques. Constantopoulos et al. ( 1 3) reported correlated the presence of pleural plaques with the
that the prevalence of pleural calcification was 47% fiber type, fiber concentration, and fiber size as deter
in 6 8 8 inhabitants of the Metsovo area in northwest mined by analytic electron microscopy in 94 long
Greece, an area where a solution containing tremolite term chrysotile miners. They found that patients
was used to whitewash the houses. with pleural plaques had a significantly higher
length-width ratio for the tremolite fibers than did
those without plaques (24) . It is believed by some
Path ogenesis
that tremolite rather than chrysotile is responsible for
Convincing evidence links pleural plaques to previ pleural plaques in the asbestos miners in Canada (25 ) .
ous asbestos exposure. Kiviluoto ( 1 4) reviewed the Not all pleural plaques are due t o asbestos expo
place of residence of all individuals with bilateral sure (26) . Zeolite minerals are aluminum silicates that
C H A PT E R 27 I OTH E R PLE U RAL D I S EASES 433
are widespread in the earth's crust. Erionite is a zeolite this hypothesis is the following: the location of black
that is found in old volcanic sites such as in Turkey, spots in the parietal pleura that represent areas where
New Zealand, areas of Japan, and in southwestern organic and inorganic material are sequestered in the
United States. Erionite is also present in gravel pits pleural lymphatics do not correspond to locations
in North Dakota, and pleural plaques were present where pleural plaques are found (33) . A third hypoth
in 3 of 1 5 gravel pit or road maintenance workers esis for the pathogenesis of pleural plaques is that the
(20%) in one study (27) . In a few villages in Turkey, microfibrils embolize to the parietal pleura by either
the mineral has been used in buildings and for road the parenchymal lymphatic plexus or through the cos
construction, and a large percentage of the popula tal vascular supply. Then once present in the parietal
tion has fiber-related pleural changes (28) . One case pleura, the fiber itself or agents carried by the fiber
of diffuse pleural thickening has been attributed to appear to be responsible for initiating and promoting
this fiber in Nevada (29) . Wollastonite, a silicate the inflammatory response. Bernstein et al. (34) have
that can be fibrous and is used in ceramics, has been demonstrated that in rats the inhalation of amosite
reported to cause pleural plaques (30) . Talc, another asbestos for 5 days results in fibers penetrating the vis
mineral that is a flaky silicate, has been reported to ceral pleural wall and within the parietal pleura within
be associated with plaque formation, but this mineral 7 days with a concomitant inflammatory response by
is often contaminated with amphiboles, so the rela 1 4 days. In contrast, no inflammatory response was
tionship remains to be proved (3 1 ) . Pleural plaques, found when chrysotile asbestos was inhaled (34) .
which may or may not be calcified, occur in other If asbestos is injected intratracheally, it migrates
pneumoconioses including those caused by mica, to the pleura. In one study in rats, the asbestos fibers
Bakelite, calcimine, tin, barite, silica, and kaolin (26) . appeared in the pleural space within 3 days of intra
They also occur after exposure to manmade vitreous tracheal injection (3 5 ) . Over a 30-day period, there
fibers (26) . However, concomitant exposure to asbes were two peaks in the appearance of the asbestos
tos is sometimes responsible for the pleural plaques fibers in the pleural space. The first peak occurred on
seen with such diseases ( 1 7) . day 7, at which time the mean length of the fiber was
Th e mechanism b y which asbestos fibers produce 1 .2 µ,m. The second peak occurred on day 2 1 when
pleural plaques is unknown. Kiviluoto ( 1 4) proposed the mean length of the fiber was only 0 . 3 µm (3 5 ) .
that pleural plaques are formed in response to inflam The intrapleural injection of either crocidolite o r
mation of the parietal pleura. When an asbestos fiber chrysotile asbestos fibers leads t o the development
is inhaled, it passes toward the periphery of the lung. of a pleural effusion (36,37) . Sahn and Antony (37)
Kiviluoto ( 1 4) suggested that the fiber pierces the vis injected chrysotile asbestos fibers into normal rabbits,
ceral pleura and then rubs against and irritates the which developed exudative pleural effusions within
parietal pleura during respiratory movements. The 4 hours. Over the next 1 20 hours, there was increas
resulting parietal pleural inflammation then gradu ing metabolic activity in the pleural fluid, as evidenced
ally evolves into the hyaline plaque, which eventu by a falling pH and an increasing Pco 2 The animals
ally calcifies. If this theory were correct, however, one developed pleural plaques that were evident by 7 days
would expect to find adhesions between the visceral and developed completely by 1 month. Interestingly,
and parietal pleura in the areas of pleural plaques, as if the rabbits were made neutropenic, they still devel
well as long asbestos fibers in the parietal pleura. oped the pleural effusion but subsequently developed
Hillerdal (32) has suggested that the short sub marked pleural fibrosis and did not develop pleu
microscopic fibers are primarily responsible for the ral plaques. The neutropenic rabbits did not have a
pleural plaques because these fibers can be demon macrophage influx as did the normal rabbits. These
strated in the plaques. He proposes that these short workers concluded that the pleural macrophage is
fibers reach the pleural space by penetrating the pul important in localizing the asbestos fiber and in the
monary parenchyma and the visceral pleura. These ultimate formation of the pleural plaque. When a
fibers are then removed from the pleural space, as is critical number of macrophages is not present, disor
all particulate matter, by the lymphatic vessels that ganization and widespread fibrosis occur (37) .
lie in the parietal pleura. Some fibers are caught in Several studies have demonstrated that the exposure
the lymphatic vessels, however, and the presence of of mesothelial cells in cell culture to asbestos particles
the fiber, in conjunction with the appropriate inflam can induce the cells to produce substances associated
matory cell, causes pleural plaques to form over with the development of fibrosis. If rat pleural meso
many years. One observation that does not support thelial cells are exposed to crocidolite or chrysotile
434 PLE U RAL D I S EASES
FIGURE 27.5 • Posteroa nte rior (PA) (A) a n d l atera l (B) chest ra d i ographs of a 69-ye a r-o l d m a n who
susta i n ed a g u n s h ot wo u n d to the l eft chest 30 yea rs previ o u s l y s h ow a l a rg e i rreg u l a r d e nse opacity
obsc u r i n g m u ch of the l eft l u n g on the PA view a n d both l u n g s on the l atera l vi ew. On the l atera l vi ew,
n ote the d e n se ca l cific opacity conform i n g to the outl i n e of the p l e u r a l space.
pleura, and the calcification occurs in the inner aspect plaques to some extent, and they should be encour
of the pleural thickening (Fig. 27. 5 ) . aged to stop smoking.
The presence of pleural plaques alone probably
results in a small decrease in the pulmonary func
S i g n ifica n ce
tion test values when there is no parenchymal asbes
Bilateral pleural plaques or calcifications are signifi tosis and when smoking is taken into consideration.
cant as an index of previous exposure to asbestos. It Schwartz et al. (54) performed spirometry on 1 ,2 1 1
appears that the presence of pleural plaques is not sheet metal workers. They reported that the forced
associated with the development of pleural meso vital capacity (FVC) in the 2 5 8 individuals with cir
thelioma when the level and duration of exposure cumscribed plaque was 3.75 L compared with an
to asbestos are taken into consideration (50) . It is FVC of 4.09 in 877 workers without pleural plaques.
controversial whether patients with pleural plaques In a subsequent study, this same group was able to
have an increased risk of lung cancer when the level demonstrate a significant relationship between the
of smoking is taken into consideration (5 1 ) . Weiss volume of the pleural fibrosis as computed from the
(52) reviewed the literature published in English three-dimensional reconstruction of the HRCT scan
in 1 993 and concluded that the weight of evidence and the total lung capacity (TLC) ( 5 5 ) . The mean
favors the conclusion that persons with asbestos TLC, however, of 24 patients with pleural fibrosis
related pleural plaques do not have an increased risk was 1 06% of that predicted ( 5 5 ) . In a large study Clin
of lung cancer in the absence of parenchymal asbes et al. (56) reported that the mean TLC (98 . 1 % pre
tosis. In 1 994, Hillerdal (53) reviewed the incidence dicted) and FVC (96.6% predicted) in 403 patients
of bronchial carcinoma and mesothelioma in 1 , 596 with pleural plaques and no parenchymal abnor
men with pleural plaques initially detected between malities were significantly lower than the mean TLC
1 963 and 1 98 5 . He found that 50 bronchial carci ( 1 0 1 .2% predicted) and FVC ( 1 00.4% predicted) in
nomas occurred whereas 32. 1 were expected, and 1 , 802 patients with asbestos exposure but normal CT
that 9 mesotheliomas occurred whereas only 0 . 8 were scans. Certainly, the functional abnormalities pro
expected. The risk of cancer and mesotheliomas is duced by pleural plaques alone are not sufficient to
therefore probably increased in patients with pleural produce symptoms. Shih et al. (57) demonstrated the
43 8 PLE U RAL D I S EASES
maximal work capacity to be 9 1 .4% of that predicted Although bromocriptine and asbestos exposure
in 20 patients with pleural plaques and no asbestosis can each lead to the development of diffuse pleural
of the lung on chest radiograph. fibrosis, it appears that the administration of bro
mocriptine and a history of asbestos exposure act
synergistically to produce diffuse pleural fibrosis (63) .
DIFFUSE THICKENING
Hillerdal et al. (63) reported on a series of 1 5 patients
In addition to the occurrence of parietal pleural who had a history of asbestos exposure and who
plaques, exposure to asbestos may be followed by developed diffuse bilateral pleural thickening after
the development of diffuse pleural fibrosis. Although taking bromocriptine for Parkinson's disease for 1 to
some authors consider this diffuse pleural fibrosis 1 0 years. The patients complained of malaise, often
to be part of the spectrum of parenchymal asbesto associated with weight loss, dyspnea, and a disturbing
sis ( 1 8) , it appears to be a distinct entity (6, 5 8-60) . cough. When the bromocriptine was withdrawn, the
In contrast to pleural plaques, diffuse pleural fibrosis patients improved clinically. However, in most cases,
commonly involves the costophrenic angles, is asso the diffuse pleural fibrosis and the restrictive lung
ciated with involvement of the visceral pleura with function defect persisted (63) .
pleural symphysis (Fig. 27.6) , and sometimes involves The pathogenesis of diffuse pleural thickening is
a marked loss of pulmonary function that can lead to unknown. However, its locale and strong association
hypercapnic respiratory failure ( 5 8-6 1 ) . with interstitial fibrosis suggest that it may be a direct
Th e incidence o f diffuse pleural fibrosis is much extension of parenchymal fibrosis to the visceral
lower than that of pleural plaques . Hillerdal ( 5 8 ) , pleura (8) . Subpleural interstitial fibrosis has been a
in surveying a group of asbestos workers, found constant feature in the limited studies using HRCT
827 individuals with pleural plaques but only 27 in subjects with diffuse pleural disease (64) . This does
with progressive pleural thickening. Schwartz et al. not explain the observation that the diffuse pleural
(54) reviewed the chest radiographs of 1 ,2 1 1 sheet fibrosis associated with asbestos exposure frequently
metal workers and reported that 260 had circum follows a benign asbestos pleural effusion (4 1 ) (see
scribed plaques, whereas 74 had diffuse thickening. Chapter 23) . Epler et al. ( 1 0) reviewed the chest
The prevalence of diffuse pleural thickening is higher radiographs of l , 1 3 5 asbestos workers and found that
with longer durations of exposure and higher inten of the 44 patients with diffuse thickening greater than
sity of exposure (6 1 ) . One report suggested that the 5 mm, almost 50% had had a previous asbestos pleu
development of pleural fibrosis was more common ral effusion. Of the 35 workers with asbestos effusion,
with human leukocyte antigen (HLA) phenotype 54% had residual diffuse pleural thickening. Hillerdal
DQ2 (62) . (58) documented that the initiating event in 4 of 27
patients with progressive pleural thickening was a
benign pleural effusion. Diffuse pleural thickening
secondary to asbestos exposure almost always involves
the costophrenic angle and invariably becomes bilat
eral, although it may be unilateral at first. This diffuse
pleural thickening starts at the bases and progresses
at a variable rate. Thickening of the pleural cap may
be considerable (58). Although routine radiographs
do not demonstrate pleural calcification in most
patients, CT scanning often demonstrates pleural
calcification (59). Many patients with diffuse pleural
fibrosis have no evidence of intrapulmonary fibrosis
on a CT scan (59) .
Patients with diffuse pleural thickening tend to
have symptoms from their pleural disease. In one
study, 61 of 64 patients (95%) complained of sig
FIGURE 27.6 • Poste roa nte r i o r chest radio nificant breathlessness on exertion (65) . In the same
g ra p h d e m o n strati n g d iffuse p l e u ra l thicke n i n g series, 56% of the patients complained of chest pain,
a n d b l u nt i n g o f the l eft costo p h r e n i c a n g l e from
which was more frequently precipitated by exer
a patient with a h i story of asbestos expos u re .
tion than by deep inspiration. Six of the patients
C H A PT E R 27 I OTH E R PLE U RAL D I S EASES 439
A B
FIGURE 27.7 • Asbestos-re l ated p l e u ra l d isease a n d ro u n ded ate l ectasis i n a 62-yea r-o l d m a n with a 20-ye a r
h i story o f asbestos expos u re . A: P A chest ra d i o g ra p h sh ows a l a rg e r i g ht p l e u r a l effusion (short arrows) a n d
a right l ower l o b e " m ass" (long arrows). B: Com puted to m o g ra p h i c (CT) sca n w i t h i ntrave n o u s contrast
d e m o n strates e n h a ncement of the p a r i eta l p l e u ra (short arro ws), i n d icat i n g a c h r o n i c p l e u ra l eff u s i o n . The
p a renchym a l " m ass" (long arrows), in contact with the visce r a l p l e u ra l su rfa ce, rep resents co l l a psed l u n g .
T h e atelectatic l u n g h a s a rou nded s h a pe owi n g t o f i b ro u s a d hesions a n d i n fo l d i n g o f t h e viscera l p l e u r a . A i r
bronchogra ms a re seen with i n t h e co l l a psed l u ng (arrowhead). C: C T sca n ( l u n g w i n d ows) shows t h e com et
ta i l s i g n or the vac u u m c l e a n e r effect, both descript i o n s of how the vessels l ea d i n g towa rd the atel ectatic
lung d iverge and a rc a ro u n d the u nd e rs u rface of the atel ectatic lung before m e rg i n g with it.
metabolic activity (79) . Fine-needle biopsy can be reviewed the chest radiographs of 765 coal miners
performed easily because the lesion is pleural based, without significant parenchymal abnormalities and
but its utility is limited because malignancy cannot be reported that 45 (5 .9%) had some pleural abnormal
excluded (75 ) . Thoracotomy is definitive, but should ity. Patients who had heavier exposures to silica were
rarely be necessary (3 1 ) . more likely to have pleural changes (80) . The most
common abnormality was obliteration of the costo
phrenic sinus (80) . A second study assessed pleural
P L E U RA L D I S EA S E R E LAT E D TO
changes in 1 1 0 individuals with a history of silica
M I N E RALS OTH E R T H A N A S B E STO S
exposure, a CT scan in the 2 years before death and
On occasion, exposure to minerals other than asbes an autopsy (8 1 ) . They reported that 28 of 62 patients
tos can produce pleural changes. Orriols et al. (80) (45 .2%) with complicated silicosis and 1 0 of 48
C H A PT E R 27 I OTH E R PLE U RAL D I S EASES 44 1
patients (20. 8%) with uncomplicated silicosis had a spaces may narrow, the size of the involved hemitho
pleural effusion on the chest radiograph or CT scan rax may diminish, and the mediastinum may be dis
(8 1 ) . Alternative explanations for the pleural effusion placed ipsilaterally. Radiologically, a peel of uniform
were present in 26 patients, but there was no other thickness surrounds the lung. Calcification occurs
explanation in 1 0 patients. Moreover, 37% of those frequently on the inner aspect of the peel (Fig. 27.5)
with uncomplicated and 77% of those with compli and provides an indicator by which the thickness of
cated silicosis had pleural thickening at autopsy (8 1 ) . the peel may be accurately measured (89) . The three
I n another study o f 1 ,008 workers using refrac main causes of fibrothorax are hemothorax, tubercu
tory ceramic fibers, pleural changes were found in losis, and bacterial lung infection (89) ; but pancre
27 (2 .7%) (82) . In these workers, the changes were atitis (90) , collagen vascular disease (9 1 ) , and uremia
mainly pleural plaques (82) . Pleural plaques have also (92) can also lead to fibrothorax. In a few instances,
been reported after exposure to pumice, an amor no etiology is ever discovered (93) .
phous complex silicate (83), and talc (84) . In the lat
ter instance, the pleural changes may well have been
C l i n ical M a n ifestations
due to concomitant exposure to asbestos (84) .
Pulmonary function is severely compromised in
fibrothorax. The degree of functional abnormality is
D I F F U S E B I LAT E RA L P LE U RA L
much greater than one would expect from the degree
T H I C K E N I N G U N R E LATE D T O
of pleural disease (94) . Pleural thickening in the cos
AS B E STO S
tophrenic angle can cause profound alterations in
Although asbestos exposure accounts for most cases the ventilation of and blood flow to the entire lung.
of diffuse bilateral pleural thickening, there are other Routine pulmonary function testing reveals mild
causes. These include drugs, particularly ergot alkaloid to-severe restrictive ventilatory dysfunction. Sur
drugs such as bromocriptine (85) (see Chapter 22) ; prisingly, the blood flow is reduced more than the
collagen vascular disease (86) (see Chapter 2 1 ) ; and ventilation of the affected side (9 5 ) . In a study of 1 27
infectious diseases, which usually produce unilat patients (95), the mean oxygen uptake on the affected
eral pleural thickening. Nevertheless, there are some side was 1 9% of the total, whereas the mean ventila
cases for which no etiology is apparent. Buchanan et tion was 33% of the total. This finding is in contrast
al. (87) described four patients with bilateral pleural to parenchymal diseases, in which the oxygen uptake
effusions progressing to diffuse pleural thickening for and ventilation are reduced to the same degree (9 5 ) .
which there was no evidence of an infective, embolic, I n severe disease, there i s n o ventilation or perfusion
or occupational cause. Histology showed that in all to the affected side (9 5 ) .
cases, both layers of the pleura were thickened by
fibrous tissue and frequently the pleural space was
Treatment
obliterated. Interestingly, all four cases were HLA-B44
positive. Pleural decortication was successful in the The only treatment available for fibrothorax is decor
three patients on whom this procedure was attempted tication, which involves removing the fibrous peel
(87) . Pleural fibrosis can be familial. Azoulay et al. from the visceral pleura. The functional improvement
(88) reported three sisters with bilateral isolated api following decortication has been variable (89,94,9 5 ) .
cal pleural fibrosis that progressed to produce severe The most important clinical factor i s the extent o f
bilateral fibrosis. Two of the sisters died of respiratory the disease i n the underlying lung (94,95) . The vital
failure, and the third received a lung transplant. capacity may improve more than 50% following
decortication if no underlying parenchymal disease is
present, but the vital capacity may even decrease fol
F I B ROTH O RAX
lowing decortication in patients with extensive paren
When pleural inflammation 1s mtense, its resolu chymal disease. Even in patients with long-standing
tion may be associated with the deposition of a thick fibrothorax, decortication can still lead to functional
layer of dense fibrous tissue on the visceral pleura. improvement. One case report noted a marked sub
The patient is then said to have a fibrothorax. As a jective improvement in a patient who had had a fibro
result of the marked pleural thickening, the hemitho thorax for 44 years (94) .
rax becomes contracted, and its mobility is reduced Which patients should have decortication?
(89) . As the fibrothorax progresses, the intercostal Patients with recent hemothorax (see Chapter 25),
442 PLE U RAL D I S EASES
recent empyema in which the infection is controlled may be 1 5 years or more after the spleen was injured.
(see Chapter 1 2) , or recent tuberculous pleuritis (see A clue to the diagnosis is the absence of Howell-Jolly
Chapter 1 3) should not have a decortication because bodies, pitted erythrocytes, and siderocytes in the
the pleural thickening frequently resolves by itself peripheral blood of asplenic individuals. Normally
over several months. Therefore, decortication should asplenic individuals have these abnormalities in their
be considered only if the pleural thickening has been peripheral blood smear. However, if there is func
stable or progressive over at least a 6-month period. tional splenic tissue elsewhere, such as in the chest,
If the pleural thickening has been present for several these cells will be absent. Technetium-99 labeled sul
months and if the patient's way oflife is compromised fur colloid radionuclide scanning can identify resid
by exertional dyspnea, decortication should probably ual splenic tissue. However, it appears that scanning
be performed unless previous chest radiographs dem with heat-damaged red blood cells is more sensitive
onstrated extensive parenchymal disease. Decortica (98) . The diagnosis can also be made by fine-needle
tion is a major surgical procedure and should not be aspiration (99) , but this is usually unnecessary with
performed on patients debilitated by other diseases. the typical history and radiographic findings. If the
In one series of 1 4 1 patients, the mortality rate with patient is asymptomatic, no therapy is indicated.
decortication was 3 . 5 % (73) .
T H O RACOLITH IAS I S
P L E U R O PA R E N C H Y M A L Thoracolithiasis is a rare benign condition character
F I B RO E LASTO S I S ized by the presence of a calcified or noncalcified loose
Pleuroparenchymal fibroelastosis is a rare condition body in the pleural space. They are usually an inciden
characterized by predominantly upper lobe pleural tal finding on imaging or at surgery ( 1 00) and do not
and subjacent parenchymal fibrosis (96) . The paren require any treatment or intervention. Through 2009,
chymal fibrosis is intra-alveolar with accompanying only 22 cases had been reported in the literature
elastosis of alveolar walls. There is also dense fibrous ( 1 00) . Seventeen of the 22 thoracoliths were on the
thickening of the visceral pleura. The intra-alveolar left side. They may be related to degenerated lipomas,
and pleural changes have a striking upper-zone pre old tuberculosis, phagocytosis of dust, or dropping of
dominance. Most cases are idiopathic, although a pleural or pericardia! fat into the pleural space. They
few cases are familial and some have been reported in frequently migrate from throughout the pleural space.
association with previous bone marrow transplanta
tion (96) . Patients present frequently with shortness
of breath and/ or cough (96) . In one series (96) of 1 2 REFERENCES
patients, 3 (25%) developed a spontaneous pneumo 1 . Mutsaers SE, Prele CM, Brody AR, et al. Pathogenesis of pleu
thorax. The etiology of this disease is unknown (96) . ral fibrosis. Respirology. 2004;9:428-440.
The natural history of this disease is progression. In 2 . Jantz MA, Antony VB . Pleural fibrosis. Clin Chest Med.
2006;27: 1 8 1 - 1 9 1 .
one series (96) of 1 0 patients, 5 patients died and
3 . Guo YB, Kalomenidis I, Hawthorne M , e t al. Pleurodesis is
the time between diagnosis and death ranged from inhibited by anti-vascular endothelial growth factor antibody.
4 months to 2 years. There is no evidence that any Chest. 2005 ; 1 28 : 1 790- 1 797.
treatment changes the course of the disease (96) with 4. Cugell DW, Kamp DW. Asbestos and the pleura: a review.
the exception of lung transplantation. Chest. 2004; 1 2 5 : 1 1 03- 1 1 1 7.
5 . Nishimura SL, Broaddus VC. Asbestos-induced pleural dis
ease. Clin Chest Med. 1 99 8 ; 1 9 : 3 1 1-329.
6. Hillerdal G. Pleural plaques in the general population. Ann N
I NTRAT H O RACIC S P L E N O S I S YAcad Sci. 1 9 9 1 ;643:430-437.
Splenosis is defined as the autotransplantation of 7. Bresnitz EA, Gilman MJ, Gracely EJ, et al. Asbestos-related
radiographic abnormalities in elevator construction workers.
splenic tissue, usually after rupture of the spleen. Am Rev Respir Dis. 1 993; 1 47: 1 3 4 1 - 1 344.
Most commonly, it is discovered as innumerable 8 . Schwartz DA. New developments in asbestos-induced pleural
purple nodules coating the mesentery, omentum, and disease. Chest. 1 99 1 ;99: 1 9 1 - 1 9 8 .
peritoneal surfaces of the abdominal cavity. When the 9 . Karjalainen A , Karhunen PJ, Lalu K , e t al. Pleural plaques and
exposure to mineral fibres in a male urban necropsy popula
diaphragm and spleen are lacerated simultaneously,
tion. Occup Environ Med. 1 994; 5 1 :456-460.
seeding of the pleural cavities can occur. 10. Epler GR, McLoud TC, Gaensler EA. Prevalence and
Intrathoracic splenosis can present with solitary or incidence of benign asbestos pleural effusion in a working
multiple pleural-based nodules (97) . The presentation population. JAMA. l 982;247: 6 1 7-622.
C H A PT E R 27 I OTH E R PLE U RAL D I S EASES 443
1 1 . Kilburn KH, Warshaw R, Thornton JC. Asbestos diseases 32. Hillerdal G. The pathogenesis of pleural plaques and
and pulmonary symptoms and signs in shipyard work pulmonary asbestosis: possibilities and impossibilities. Eur
ers and their families in Los Angeles. Arch Intern Med. ] Respir Dis. 1 980;6 1 : 1 2 9- 1 3 8 .
1 986; 1 46:22 1 3-2220. 33. Mitchev K , Dumortier P, D e Vuyst P. "Black spots" and hya
12. Churg A, DePaoli L. Environmental pleural plaques in line pleural plaques on the parietal pleura of 1 5 0 urban nec
residents of a Quebec chrysotile mining town. Chest. ropsy cases. Am J Surg Pathol. 2002;26: 1 1 98-1 206.
1 9 88;94:58-69 . 34. Bernstein DM, Rogers RA, Sepulveda R, et al. Quantification
1 3 . Constantopoulos SH, Theodoracopoulos P, Dascalopoulos G, of the pathological response and fate in the lung and pleura
et al. Tremolite whitewashing and pleural calcifications. Chest. of chrysotile in combination with fine particles compared to
1 9 87;92:709-7 1 2. amosite-asbestos following short-term inhalation exposure.
1 4 . Kiviluoto R. Pleural calcification as a roentgenologic sign Inhal Toxicol. 20 1 1 ;23:372-39 1 .
of non-occupational endemic anthophyllite-asbestosis. Acta 3 5 . Viallat JR, Raybuad F, Passarel M , et al. Pleural migration of
Radio!. 1 960; 1 94(suppl 1 ) : 1 -67. chrysotile fibers after intratracheal injection in rats. Arch Envi
1 5 . Hourihane DO, Lessof L, Richardson PC. Hyaline and calci ron Health. 1 9 86;4 l :282-286.
fied pleural plaques as an index of exposure to asbestos: a study 36. Shore BL, Daughaday CC, Spilberg I. Benign asbestos pleu
of radiological and pathological features of 1 00 cases with a risy in the rabbit. Am Rev Respir Dis. 1 9 83; 1 28:48 1 -4 8 5 .
consideration of epidemiology. Br MedJ 1 966; 1 : 1 069-1 074. 37. Sahn SA, Antony V B . Pathogenesis of pleural plaques: rela
1 6 . Hillerdal G, Lindgren A. Pleural plaques: correlation of tionship of early cellular response and pathology. Am Rev
autopsy findings to radiographic findings and occupational Respir Dis. 1 984; 1 30 : 8 84-8 87.
history. Eur] Respir Dis. 1 9 80;6 1 : 3 1 5-3 1 9 . 3 8 . Kuwahara M, Kuwahara M, Verma K, et al. Asbestos expo
1 7 . Sargent EN, Jacobson G, Gordonson JS. Pleural plaques: sure stimulates pleural mesothelial cells to secrete the fibro
a signpost of asbestos dust inhalation. Semin Roentgenol. blast chemoattractant, fibronectin. Am J Respir Cell Mo! Biol.
1 977; 12:287-297. 1 994; 1 0 : 1 67- 1 76.
1 8 . Becklake MR. Asbestos-related diseases of the lung and ocher 39. Robledo R, Mossman B . Cellular and molecular mechanisms
organs: their epidemiology and implications for clinical prac of asbestos-induced fibrosis. ] Cel!Physiol. 1 999; 1 80: 1 58-1 66.
tice. Am Rev Respir Dis. 1 976; 1 1 4 : 1 87-227. 40. Light RW, Cheng D-S, Lee YC, et al. A single intrapleural
1 9 . Paris C, Thierry S, Brochard P, et al. Pleural plaques and asbes injection of transforming growth factor-/32 produces excel
tosis: dose and time-response relationships based on HRCT lent pleurodesis in rabbits. Am ] Respir Crit Care Med.
data. Eur Respir j 2009;34:72-79. 2000; 1 62:98-1 04.
20. Craighead JE, Mossman BT. The pathogenesis of asbestos 4 1 . Peacock C, Copley SJ, Hansell DM. Asbesros-related benign
associated diseases. N Engl] Med. 1 9 82;306 : 1 446- 1 4 5 5 . pleural disease. Clin Radio!. 2000;55:422-432.
2 1 . Roberts GH. The pathology of parietal pleural plaques. J Clin 42. Withers BF, Ducatman AM, Yang WN . Roentgenographic
Pathol. 1 97 1 ;24:348-3 5 3 . evidence for predominant left-sided location of unilateral
22. Kishimoto T, O n o T, Okada K , et a l . Relationship between pleural plaques. Chest. 1 9 89;9 5 : 1 262- 1 264.
number of asbestos bodies in autopsy lung and pleural plaques 43. Hu H, Beckett L, Kelsey K, et al. The left-sided predomi
on chest x-ray film. Chest. 1 9 89;95 : 549-5 52. nance of asbestos-related pleural disease. Am Rev Respir Dis.
23. Cowie RL, Murray J, Becklake MR. Pneumoconioses. In: 1 993; 1 4 8 : 9 8 1-984.
Mason RJ, Broaddus VC, Murray JF, et al. eds. Murray and 44. Gallego JC. Absence of left-sided predominance in
Nadels Textbook of Respiratory Medicine. Philadelphia, PA: asbestos-related pleural plaques: a CT study. Chest.
Elsevier Science; 2005: 1 748- 1 782. 1 998; 1 1 3: 1 034-1 036.
24. Churg A, Wright JL, Veda! S . Fiber burden and patterns of 4 5 . Gevenois PA, De Vuyst P, Dedeire S, et al. Conventional and
asbestos-related disease in chrysotile miners and millers. Am high-resolution CT in asymptomatic asbestos-exposed work
Rev Respir Dis. 1 993; 1 48:25-3 1 . ers. Acta Radio!. 1 994;35 :226-229 .
2 5 . Gibbs GW. Etiology of pleural calcification: a study of 46. Whitman GJ, Niklason LT, Pandit M , et a l . Dual-energy digi
Quebec chrysotile miners and millers. Arch Environ Health. tal subtraction chest radiography: technical considerations.
1 979;34:76-83. Curr Prob! Diagn Radio!. 2002;3 l :48-62.
26. Clarke CC, Mowat FS, Kelsh MA, et al. Pleural plaques: a 47. Sargent EN, Boswell WD Jr, Ralls PW, et al. Subpleural fat
review of diagnostic issues and possible nonasbestos factors. pads in patients exposed to asbestos: distinction from non
Arch Environ Occup Health. 2006;6 1 : 1 83- 1 92. calcified pleural plaques. Radiology. 1 9 84; 1 52:273-277.
27. Ryan PH, Dihle M, Griffin S, et al. Erionite in road gravel 48. Lee YCG, Runnion CK, Pang SC, et al. Increased body mass
associated with interstitial and pleural changes-an occupa index is related to apparent circumscribed pleural thick
tional hazard in western United States. J Occup Environ Med. ening on plain chest radiographs. Am J Ind Med. 200 1 ;3 9 :
20 1 1 ; 5 3 : 8 92-89 8 . 1 1 2- 1 1 6 .
2 8 . Baris I, Simonato L , Artvinli M, e t al. Epidemiological and 4 9 . Friedman AC, Piel S B , Fisher M S , e t al. Asbestos-related pleu
environmental evidence of the health effects of exposure to ral disease and asbestosis: a comparison of CT and chest radi
erionite fibres: a four-year study in the Cappadocian region of ography. A]R A m ] Roentgenol. 1 988; 1 5 0:269-275 .
Turkey. Int] Cancer. 1 9 87;39 : 1 0- 1 7 . 50. Reid A, de Klerk N, Ambrosini G, et al. The additional risk of
29. Casey KR, Shigeoka JW, Rom WN , e t al. Zeolite exposure malignant mesothelioma in former workers and residents of
and associated pneumoconiosis. Chest. 1 98 5 ;87: 837-840. Wittenoom with benign pleural disease or asbestosis. Occup
30. Huuskonen MS, Tossavainen A, Koskinen H, et al. Wollaston Environ Med. 2005;62:66 5-669.
ite exposure and lung fibrosis. Environ Res. 1 983;30:29 1 -304. 5 1 . Edelman DA. Asbestos exposure, pleural plaques and
3 1 . Hillerdal G . Nonmalignant pleural disease related to asbestos the risk of lung cancer. Int Arch Occup Environ Health.
exposure. Clin Chest Med. 1 985;6: 1 4 1 - 1 52. 1 9 88;60:389-393.
444 PLE U RAL D I S EASES
52. Weiss W. Asbestos-related pleural plaques and lung cancer. 74. Hillerdal G, Ozesmi M . Benign asbestos pleural effusion: 73
Chest. 1 993; 1 03 : 1 854- 1 8 5 9 . exudates in 60 patients. Eur] Respir Dis. 1 987;7 1 : 1 1 3-1 2 1 .
5 3 . Hillerdal G . Pleural plaques and risk for bronchial carcinoma 7 5 . Szydlowski G W, Cohn HE, Steiner RM , e t al. Rounded
and mesothelioma. Chest. 1 994; 1 0 5 : 1 44-1 50. atelectasis: a pulmonary pseudotumor. Ann 7horac Surg.
54. Schwartz DA, Fuortes LJ, Galvin JR, et al. Asbestos-induced 1 992;53 : 8 1 7-82 1 .
pleural fibrosis and impaired lung function. Am Rev Respir Dis. 76. Horita Y, Noguchi M , Miyazaki M , et al. Prognosis of patients
1 990; 1 4 1 :32 1 -326. with rounded atelectasis undergoing long-term hemodialysis.
55. Schwartz DA, Galvin JR, Yagla SJ, et al. Restrictive lung Nephron. 200 1 ;8 8 : 87-92.
function and asbestos-induced pleural fibrosis. A quantitative 77. Bloom CI, Wilson GE. Rounded atelectasis and respira
approach. ] Clin Invest. 1 993;9 1 :268 5-2692. tory compromise secondary to pergolide use. Respirology.
56. Clin B, Paris C, Ameille J, et al. Do asbestos-related pleural 2009; 1 4 : 906-907.
plaques on HRCT scans cause restrictive impairment in the 78. McLoud TC, Flower CD. Imaging the pleura: sonography, CT,
absence of pulmonary fibrosis? Thorax. 20 I 1 ;66:985-99 1 . and MR imaging. AJRAm] Roentgenol. 1 99 1 ; 1 5 6: 1 1 4 5- 1 1 53 .
57. Shih J-F, Wilson JS, Broderick A, et al. Asbestos-induced 79. Ludeman N, Elicker B M , Reddy GP, e t al. Atypical rounded
pleural fibrosis and impaired exercise physiology. Chest. atelectasis: diagnosis and management based on results of
1 994; 1 0 5 : 1 370- 1 376. F-1 8 FDG positron emission tomography. Clin Nucl Med.
58. Hillerdal G. Non-malignant asbestos pleural disease. Thorax. 2005;30 :734-73 5 .
1 9 8 1 ;36: 669-675 . 8 0 . Orriols R , Munoz X , Sunyer J, e t al. Radiologically recognized
5 9 . Wright P H , Hanson A , Keel L , e t al. Respiratory function pleural changes in nonpneumoconiotic silica-exposed coal
changes after asbestos pleurisy. Thorax. 1 9 80;3 5 : 3 1 -36. miners. Scand] Work Environ Health. 2005;3 1 : 1 1 5- 1 2 1 .
60. Miller A, Teirstein AS, Selikoff I. Ventilatory failure due to 8 1 . Arakawa H, Honma K , Saito Y, e t al. Pleural disease i n silico
asbestos pleurisy. Am J Med. 1 9 83;75 : 9 1 1 -9 1 9 . sis: pleural thickening, effusion, and invagination. Radiology.
6 1 . Miles SE, Sandrini A, Johnson AR, e t al. Clinical consequences 2005;236:68 5-693.
of asbestos-related diffuse pleural thickening: a review. J Occup 82. Lockey JE, LeMasters GK, Levin L, et al. A longitudinal study
Med Toxicol. 2008;3:20. of chest radiographic changes of workers in the refractory
62. Shih J-F, Hunninghake GW, Goeken NE, et al. The relation ceramic fiber industry. Chest. 2002; 1 2 1 :2044-205 1 .
ship between HLA-A, B, DQ, and DR antigens and asbestos 83. Mazziotti S , Costa C , Ascenti G , et al. Unusual pleural involve
induced lung disease. Chest. 1 993; I 04:26-3 1 . ment after exposure to amorphous silicates (Liparitosis) :
63. Hillerdal G , Lee J , Blomkvist A, et al. Pleural disease during report of two cases. Eur Radio/. 2002; 1 2 : 1 058-1 060.
treatment with bromocriptine in patients previously exposed 84. Gamble JF, Fellner W, Dimeo MJ . An epidemiologic study of a
to asbestos. Eur Respir J 1 997; 1 0 :27 1 1 -27 1 5 . group of talc workers. Am Rev Respir Dis. 1 979; 1 1 9 :74 1 -753.
64. Aberle DR, Balmes JR. Computed tomography o f asbestos 8 5 . Agarwal P, Fahn S, Frucht SJ. Diagnosis and management of
related pulmonary parenchymal and pleural diseases. Clin pergolide-induced fi b rosis. Mov Disord. 2004; 1 9 :699-704.
Chest Med. 1 99 1 ; 1 2: 1 1 5- 1 3 1 . 86. Sharma S, Smith R, AJ-Hameed F. Fibrothorax and severe
6 5 . Yates DH, Browne K, Stidolph PN, et al. Asbestos-related lung restriction secondary to lupus pleuritis and its successful
bilateral diffuse pleural thickening: natural history of radio treatment by pleurectomy. Can Respir J 2002;9:33 5-337.
graphic and lung function abnormalities. Am J Respir Crit 87. Buchanan DR, Johnston ID, Kerr IH, et al. Cryptogenic
Care Med. 1 996; 1 53:30 1-306. bilateral fibrosing pleuritis. Br J Dis Chest. 1 9 88;82: 1 8 6- 1 93 .
66. Van Cleemput J, De Raeve H, Verschakelen JA, et al. Surface 8 8 . Azoulay E, Paugam B, Heymann MF, e t al. Familial exten
of localized pleural plaques quantitated by computed tomog sive idiopathic bilateral pleural fibrosis. Eur Respir J
raphy scanning. No relation with cumulative asbestos expo 1 999; 1 4:971-973 .
sure and no effect on lung function. Am ] Respir Crit Care 8 9 . Morton JR, Boushy SF, Guinn GA. Physiological evalua
Med. 200 1 ; 1 63:705-7 1 0 . tion of results of pulmonary decortication. Ann 7horac Surg.
6 7 . Picado C, Laporta D, Grassin o A , e t al. Mechanisms affecting l 970;9:32 1 -326.
exercise performance in subjects with asbestos-related pleural 90. Shapiro DH, Anagnostopoulos CE, Dineen JP. Decortication
fibrosis. Lung. 1 987; 1 65 :45-57. and pleurectomy for the pleuropulmonary complications of
68. Leung AN, Muller NL, Miller RR. CT in differential pancreatitis. Ann 7horac Surg. 1 970;9 :76-80.
diagnosis of diffuse pleural disease. A]R Am J Roentgenol. 9 1 . Brunk JR, Drash EC, Swineford 0 . Rheumatoid pleuritis
1 990; 1 54:487-492. successfully treated with decortication. Report of a case and
69. Fielding DI, McKeon JL, Oliver WA, et al. Pleurectomy for review of the literature. Am ] Med Sci. 1 966;25 1 : 545-5 5 1 .
persistent pain in benign asbestos-related pleural disease. Tho 92. Gilbert L, Ribot S , Frankel H , e t al. Fibrinous uremic pleuri
rax. 1 99 5 ; 5 0 : 1 8 1 - 1 8 3 . tis: a surgical entity. Chest. 1 975;67:53-56.
70. Munoz X , Roger A , Pallisa E, e t a l . Ventilatory insufficiency 93. Lee-Chiong TL Jr, Hilbert ]. Extensive idiopathic benign bilat
due to asbestos-related diffuse pleural fibrosis successfully eral asynchronous pleural fibrosis. Chest. 1 996; 1 09 : 5 64-5 6 5 .
treated with noninvasive home mechanical ventilation. Respi 94. Hughes R , Jensik RJ, Faber L P, e t a l . Evaluation of unilateral
ration. 200 1 ;6 8 : 5 33-536. decortication: a patient successfully treated 44 years after onset
7 1 . Blesovsky A. The folded lung. Br J Dis Chest. 1 966; 60: 1 9-22. of tuberculosis. Ann 7horac Surg. 1 975 ; 1 9 :704-7 1 5 .
72. Batra P, Brown K, Hayashi K, et al. Rounded atelectasis. 9 5 . Gaensler EA . Lung displacement: abdominal enlargement,
] 7horac Imaging. 1 996; 1 1 : 1 87- 1 97. pleural space disorders, deformities of the thoracic cage. In:
73. Dernevik L, Gatzinsky P. Pathogenesis of shrinking pleuri Fenn WO, Rahn H, eds. Handbook of Physiology, Section 3.
tis with atelectasis-"rounded atelectasis." Eur J Respir Dis. Respiration, Vol. 2. Washington, DC: American Physiological
1 987;7 1 :244-249. Society, 1 965 : 1 623- 1 66 1 .
C H A PT E R 27 I OTH E R PLE U RAL D I S EASES 445
96. Reddy TL, Tominaga M, Hansell OM, et al. Pleuroparen 99. Syed S, Zaharopoulos P. Thoracic splenosis diagnosed by
chymal fibroelastosis: a spectrum of histopathological and fine-needle aspiration cytology: a case report. Diagn Cytopa
imaging phenotypes. Eurp Resp ]. 20 1 2;40:377-3 8 5 . thol. 200 1 ;25:32 1 -324.
9 7 . Yousem SA. Thoracic splenosis. Ann Thorac Surg. 1 987;44: 1 00 . Strzelczyk J, Holloway BJ, Pernicano PG, et al. Rolling
4 1 1 -4 1 2. stones in the pleural space: thoracoliths on CT, and a review
9 8 . Hagman TF, Winer-Muram HT, Meyer CA, et al. Intra of the literature. Clin Radio/. 2009;64: 1 00- 1 04 .
thoracic splenosis. Superiority of technetium Tc 99 heat
damaged RBC imaging. Chest. 200 1 ; 1 20:2097-2098.
T h o ra ce ntes i s ( D i a g n o st i c a n d T h e ra p e u t i c)
a n d P l e u ra l B i o psy
Thomsen et al. (1) have created an excellent video on are not needed before thoracentesis in patients with
performing a thoracentesis, and watching this video mild coagulopathy and no clinical evidence of bleed
is recommended before one performs his or her first ing (3) . These authors did note an increased risk of
diagnostic or therapeutic thoracentesis. bleeding if the creatinine level was elevated above
6 mg/dL (3) , presumably because of decreased plate
let function in the setting of uremia.
D IAG N O ST I C T H O RACE NTE S I S
It appears that thoracentesis can be safely per
I n d ications formed on patients who are undergoing mechanical
A diagnostic thoracentesis should be performed on ventilation. In one study from Beth Israel Hospital
almost every patient with a pleural effusion of un in New York, 232 thoracenteses were performed in
known origin. Empirically, I have found it difficult 2 1 1 patients receiving mechanical ventilation under
to obtain fluid with a diagnostic thoracentesis if the ultrasound guidance (5) . In this series, 95 .6% of
thickness of the fluid on the decubitus chest radio the patients were on positive end expiratory pres
graph or the CT scan is less than 1 0 mm, and I usually sure (PEEP) , 63% were receiving vasopressors, and
6 1 % underwent therapeutic thoracenteses (5) . Tho
do not attempt thoracentesis in such patients. If thora
centesis is attempted with small amounts of fluid, the racentesis yielded fluid in 98 .7% of the procedures
proper location can be identified using ultrasound (2) . and the incidence of pneumothorax was only 1 .3%
(5). It should be noted that the ultrasound in this
series was performed by pulmonologists (5). When
Contra i n d ications
thoracentesis has been performed without ultrasound
The main contraindication to a diagnostic thoracen on patients receiving mechanical ventilation, the
tesis is a hemorrhagic diathesis. One should hesitate incidence of pneumothorax has been much higher.
to perform a thoracentesis in a patient who is receiv McCartney et al. ( 6) reported a series of 3 1 patients
ing anticoagulants, particularly thrombolytic agents. who underwent thoracentesis while they were receiv
Depending on the urgency of the situation, however, ing mechanical ventilation; 25 patients were receiving
diagnostic thoracentesis using a small (22-gauge) PEEP between 5 and 20 cm H 2 0 . All thoracenteses
needle can be performed on almost any patient if one were performed with patients in the lateral decubi
is careful. McVay et al. (3) demonstrated that there tus position. Three of the patients ( 1 0%) developed
was no increased risk of bleeding if the prothrombin a pneumothorax, and all were managed with a chest
time or the partial thromboplastin time was not more tube (6) . In a second series, 2 of 32 patients (6%)
than two times the normal value. Likewise, there was developed a pneumothorax after undergoing a tho
no increased risk of bleeding with low platelet counts racentesis while on mechanical ventilation (7) . A
( < 2 5 , 000/mm3) . A more recent study reported thoracentesis should not be attempted through an
similar results (4) . Accordingly, these authors recom area affected by a local cutaneous condition such as
mend that prophylactic blood product transfusions pyoderma or herpes zoster infection.
446
C HAPT E R 28 I T H O RAC E NTESIS (D IAG N OST I C A N D T H E RA P E UTIC) A N D P L E U RAL B I O PSY 447
A B
c D
The materials required to perform diagnostic thora One 3 -way sto pcock
One ste r i l e c o n ta i n e r for p l e u ra l fl u i d
centesis are listed in Table 2 8 . 1 , and these materials
O n e 5 0 - m l syri n g e (a d d it i o n a l )
should be assembled before the procedure is initiated.
Ad d it i o n a l m a te r i a l s for p l e u ra l b i o psy
What is done more commonly, however, is to use a
P l e u ra l b i o psy n e e d l e
thoracentesis kit in which the materials have been Sca l pe l
preassembled. There are several thoracentesis kits F o rm a l i n
available commercially including the Arrow-Clark™
Pleura-Seal" Thoracentesis Kit distributed by Ar
row, and the Argyle™ Turkel'" Safety Thoracentesis by aspirating directly through the side port on the
System distributed by Kendall. When a thoracente catheter. Another nice feature of this set is that one
sis kit is selected, it is important to select one with may easily withdraw large amounts of fluid either
a needle-catheter system, particularly if a therapeutic with a syringe or through vacuum bottles. If a syringe
thoracentesis is going to be performed. If one per is used, aspiration is performed through a Y connec
forms a therapeutic thoracentesis with a sharp needle, tor that has one-way valves so that no stopcocks need
the needle is likely to lacerate the lung as it reex be turned with each aspiration. If vacuum bottles are
pands with the removal of pleural fluid and cause a used, vacuum bottle tubing is included, which at
pneumothorax. taches directly to the sideport of the catheter. There
One excellent thoracentesis kit is the Arrow is also a roller clamp to control the flow of fluid into
Clark'" Pleura-Seal" Thoracentesis Kit manufactured the vacuum bottles. The cost of this thoracentesis kit
by Arrow International, Reading, Pennsylvania (www. is approximately US $3 l .
arrowintl.com, 800-523-8446) . The basic thoracen Another excellent set is the Argyle'" Turkel'" Safety
tesis apparatus in this kit is an 8-F gauge catheter Thoracentesis System set manufactured by Kendall
over an 1 8-gauge needle with a three-way stopcock Company (St. Louis, MO, USA, www. kendallhq.
and self-sealing valve. With this apparatus, one con com, 800-962-9888) . This system incorporates a
stantly aspirates as the catheter is advanced through blunt, multiside fenestrated, spring-loaded inner
the chest wall. Then, when a free flow of fluid is en cannula coaxially housed within a 1 6-gauge conven
countered, the catheter is advanced approximately tional sharp-beveled hollow needle. The advantage
1 cm and the needle is withdrawn completely. One of this system is that as the needle and blunt can
nice feature of this set is that there is a self-sealing nula penetrate the chest wall, the blunt cannula is
valve so that air does not leak into the pleural space forced into the shaft of the needle. Then when the
when the needle is withdrawn; however, the needle tip of the needle encounters low resistance, such as an
cannot be reinserted through the catheter. With this area of pleural effusion within the pleural space, the
system, small amounts of fluid can be withdrawn spring-loaded cannula automatically extends beyond
450 PLE U RAL D I S EASES
the bevel, thereby helping to protect the underlying The next step is to obtain local anesthesia. It is
tissue from further, inadvertent penetration. Another necessary to anesthetize the skin, the periosteum
advantage of this system is an indicator in the needle of the rib, and the parietal pleura. The skin is anes
housing that identifies the position of the blunt can thetized using a short 25-gauge needle by injecting
nula; if resistance is being met such that the sharp end enough lidocaine, approximately 0 . 5 mL, to raise a
of the needle is exposed, then the indicator is red. In small wheal (Fig. 28.2A) . The small needle is then re
contrast, if no resistance is being met, then the indi placed by a 1 . 5-in.-long 22-gauge needle. This needle
cator is green. Therefore, when the pleural space is is inserted to the periosteum of the underlying rib
entered, the indicator turns green. If a diagnostic tho and is moved up and over the rib with frequent injec
racentesis is being performed, the pleural fluid can be tion of small amounts (0 . 1 to 0.2 mL) of lidocaine
withdrawn through the needle. If a therapeutic tho (Fig. 28 .2B) . Once this needle is superior to the rib,
racentesis is being performed, the catheter assembly is it is slowly advanced toward the pleural space with
advanced and then the needle assembly is withdrawn aspiration, followed by the injection of 0 . 1 to 0.2
completely. There is a one-way valve such that there mL of lidocaine every 1 to 2 mm (Fig. 28 .2C) . This
is no possibility of air leaking into the pleural space frequent aspiration and the injection of lidocaine
when the needle is withdrawn. There is a side port for guarantee anesthesia of the parietal pleura. As soon
fluid removal. The cost of these kits is approximately as pleural fluid is aspirated through this needle into
US$24. the syringe containing lidocaine, the needle should be
When comparing the two kits described ear withdrawn from the pleural space and reattached to
lier, the Argyle Turkel kit has the advantage of the a 50- to 60-mL syringe containing 1 mL of heparin.
spring-loaded inner cannula, which should decrease Heparin is added into the syringe to prevent clotting
the incidence of lung laceration. The advantage of the of the pleural fluid as it is difficult to obtain differen
Arrow-Clark system is the ease with which a thera tial white blood cell counts or pH determinations if
peutic thoracentesis can be performed either with a the pleural fluid is clotted. The same needle is rein
syringe or with vacuum bottles. troduced along the same tract slowly with constant
aspiration until pleural fluid is obtained. Aspiration is
then continued until the syringe is filled. The needle
Tech n i q u e
is then withdrawn, and the procedure is finished. The
Th e procedure should b e carefully explained t o the commercially available kits can be used to perform
patient, and a signed consent form should be ob a diagnostic thoracentesis. The special needles that
tained. Routine administration of atropine to prevent come with these kits, however, have few advantages
vasovagal reactions is not recommended as such reac over a syringe and a needle for a diagnostic thoracen
tions are very uncommon during thoracentesis ( 1 3) . tesis. They do have significant advantages for thera
Atropine should b e available, however, and 1 . 0 mg peutic thoracentesis, and they should be used in this
should be administered subcutaneously or intramus situation.
cularly at the first sign of such a reaction. Similarly, I At times, no pleural fluid is obtained when the l . 5 -
do not administer an analgesic, a sedative, or a tran i n . No. 22 needle i s inserted all the way t o its hub.
quilizer routinely before the procedure unless the In such a situation, the needle should be slowly
patient shows excessive anxiety. If there is excessive withdrawn with constant aspiration. The rim of the
anxiety, I administer intravenous midazolam (Versed) pleural fluid is sometimes thin and may be missed as
j ust before the procedure. the needle is inserted. If no pleural fluid is obtained
Once the site for thoracentesis is identified, it is either as the needle is inserted or withdrawn, one of
marked by exerting pressure using the end of a ball four possibilities exists: (a) the needle was too short;
point pen with the tip retracted. This leaves a small (b) placement of the needle was too far superior; (c)
indentation that will not be removed by subsequent placement of the needle was too far inferior; or (d)
cleansing of the area. Then the skin surrounding the no pleural fluid is present. If the patient is markedly
site is cleansed thoroughly with an antiseptic solu muscular or obese and if no air is obtained on the
tion over an area extending at least 4 in. in all di initial attempt, the 1 . 5-in. needle should be replaced
rections from the proposed thoracentesis site. The with a longer needle such as a spinal needle, which
sterile drape with the center hole is then taped to the is used for performing lumbar puncture, and the at
patient's back, and another sterile drape is placed on tempt should be repeated. If no fluid is aspirated, but
the bed. air bubbles are obtained on the initial attempt with
C HAPT E R 28 I T H O RAC E NTESIS (D IAG N OST I C A N D T H E RA P E UTIC) A N D P L E U RAL B I O PSY 45 1
the local anesthetic, the lung parenchyma has been although it is not necessary to pack it in ice as long
penetrated and the needle was inserted too far supe as the pH determination is performed within an hour
riorly (Fig. 28 .2D) . Therefore, the procedure should ( 1 7, 1 8) . For the bacterial cultures, it is best to innoc
be repeated one interspace inferiorly. Penetration of ulate the pleural fluid directly into blood culture bot
the lung with a small needle is not a catastrophe, tles at the bedside ( 1 9) . Interpretation of the results
and only occasionally does a pneumothorax result. of the various tests obtained in Table 28.2 is discussed
If no fluid or air bubbles are obtained on the initial in Chapters 7 and 8. If there is a good chance that the
attempt, the needle was inserted too far inferiorly patient has a transudative pleural effusion, the most
(Fig. 28 .2E) , and the procedure should be repeated cost-effective approach is to measure only the lactate
one interspace superiorly. Pleural fluid is almost never dehydrogenase (LDH) and protein in the pleural
too viscous to be aspirated through a No. 20 or a No. fluid. If these measurements demonstrate that the
22 needle. If no fluid is obtained after two or three patient does not have a transudative pleural effusion
attempts, ultrasound guidance for the thoracentesis (20), the remaining studies should be performed.
is recommended.
Com p l icati o n s
Process i n g of P l e u ra l F l u i d
Three factors are important in reducing the risk o f
Th e main purpose o f a diagnostic thoracentesis i s to thoracentesis: ultrasonography, simulation and super
examine the pleural fluid. The recommended distri vision (2 1 -23) . The most common complication of
bution of the pleural fluid to various laboratories is thoracentesis is pneumothorax. At the Mayo Clinic,
outlined in Table 28.2. For the cell count and dif the risk of pneumothorax following thoracentesis
ferential, the fluid should be placed in ethylenedia decreased from 8.6% to 1 . 1 % after a program was
minetetraacetic acid (EDTA)-treated tubes (purple instituted to restrict the number of physicians autho
top tubes) if the syringe was not initially heparin rized to perform a thoracentesis to a subset of physi
ized. If the syringe was not initially heparinized and cians who had specific training, used ultrasound, and
if fluid is placed in the tubes without anticoagulants performed the procedure on a regular basis (22) . The
that come with the thoracentesis trays, the cells are incidence of pneumothorax following thoracentesis is
likely to clump or the fluid is likely to clot, giving reduced if experienced individuals such as pulmonary
inaccurate cell counts and differentials ( 1 5) . For the fellows or pulmonologists perform the procedure
determination of pleural fluid pH, the sample should (24) . The incidence of iatrogenic pneumothorax is
be maintained anaerobically and the determination lower if the procedure is performed under ultrasound
should be made with a blood gas machine ( 1 6, 1 7) , guidance (2 1 ,2 5 ) . In the largest study, the incidence
of pneumothorax was only 2 . 5 % in 941 procedures a chest tube. Gervais et al. (28) reported that the
performed under ultrasound guidance and only 0.8% incidence of iatrogenic pneumothorax was approxi
of the patients received a chest tube ( 1 3) . Raptopoulos mately 1 % in nonintubated patients undergoing
et al. (26) reported that the incidence of pneumotho ultrasound-guided thoracentesis and concluded that
rax was 1 8% for 1 54 thoracenteses performed with routine postprocedure chest radiographs are not in
conventional techniques, whereas it was only 3% for dicated in spontaneously breathing patients who un
1 8 8 performed with ultrasound guidance. Patel et al . dergo thoracentesis. Doyle et al. (29) reviewed their
(25) reported that the incidence of pneumothorax in experience with 1 74 thoracentesis and concluded
8 ,824 patients who underwent thoracentesis with ul that postprocedure chest radiographs were indicated
trasound (3.9%) was significantly lower than that in only when a pneumothorax is suspected (29) . Jones
1 0, 5 1 0 patients who underwent the procedure with et al. reported that only 3 of907 patients (0 .3%) who
out ultrasound (4.6%) in 2008. It is noteworthy that were asymptomatic after a thoracentesis procedure
less than 50% of the patients had their thoracentesis developed a pneumothorax and required a chest tube
with ultrasound. ( 1 3) . In view of the series mentioned in the preceding
It appears that the likelihood of the development text, I recommend postprocedure radiographs only
of a pneumothorax may be higher in patients with when air is obtained during the thoracentesis, the
chronic obstructive pulmonary disease (COPD) . patient develops symptoms, or when tactile fremitus
Brandstetter et al. (27) performed thoracentesis in is lost over the superior part of the aspirated hemi
1 06 patients, of whom 36 had COPD. The incidence thorax. The treatment of iatrogenic pneumothorax is
of pneumothorax was significantly higher (4 1 . 7%) discussed in Chapter 24.
in those patients with COPD than in those without Other common complications of thoracentesis are
COPD ( 1 8 . 5%) (27) . Nine of the 1 06 patients were cough and chest pain (24) . Cough most frequently
treated with chest tubes and 7 of them had COPD complicates thoracentesis when it is performed for
(27) . The explanation of the very high incidence of therapeutic reasons and usually occurs toward the
pneumothorax in this series is not clear. In contrast, end of the thoracentesis (24) . Indeed if coughing
Raptopoulos et al. (26) were unable to find a relation occurs during a thoracentesis, it should serve as an
ship between the occurrence of a pneumothorax and indication to stop the procedure. The chest pain that
the presence of underlying lung disease. complicates thoracentesis is of three types. First, the
There are two different reasons that patients de patient may experience sharp pain when the skin is
velop pneumothorax after a thoracentesis. First, air anesthetized or when the parietal pleura is pierced.
may flow from the atmosphere into the pleural space This pain should not be persistent if the parietal
if the pleural space (with its negative pressure) com pleura is adequately anesthetized. Second, the patient
municates freely with the atmosphere. This most may experience chest tightness or dull pain as fluid is
commonly happens when a syringe is removed from removed during a therapeutic thoracentesis. This type
a needle or catheter and the air then flows from the of pain usually indicates that the patient's lung is not
atmosphere into the pleural space and produces a expanding rapidly and should serve as an indication
pneumothorax. This problem can be prevented if spe to stop the procedure. Third, the patient may develop
cial needles with one-way valves (such as the Arrow pleuritic chest pain after the procedure, which is usu
ClarkT" Pharma-Seal" or Argyle™ Turkel™) are used ally because of the roughened pleural surfaces rub
during thoracentesis. It can also be prevented if the bing on each other after some of the fluid has been
patient hums (producing positive pleural pressure) withdrawn. On auscultation, a pleural rub is often
while the needle is being changed. Second, the needle heard in these cases.
for thoracentesis may lacerate the lung and permit air At times, a diagnostic thoracentesis provokes a
to enter the pleural space from the alveoli. This can be vasovagal reflex characterized by bradycardia, a de
prevented if catheters, rather than sharp needles, are creased stroke volume, and a resultant fall in cardiac
used to perform therapeutic thoracenteses. output and blood pressure. This reaction is blocked
Should chest radiographs be obtained routinely by the intramuscular administration of 1 mg of atro
after diagnostic thoracentesis? It appears that rou pine. A similar syndrome may be provoked by various
tine chest radiographs are not indicated. Aleman et noxious, emotional, and physical stimuli such as ap
al. (24) reported that only 5 of 488 patients without prehension, pain, or the sight of blood and is char
symptoms after thoracentesis developed a pneumo acterized by the sudden loss of peripheral vascular
thorax and that only 1 of these 5 patients required resistance without significant bradycardia. The patient
C HAPT E R 28 I T H O RAC E NTESIS (D IAG N OST I C A N D T H E RA P E UTIC) A N D P L E U RAL B I O PSY 453
develops hypotension, pallor, cold and clammy skin, effusions or empyema (see Chapter 1 2) , therapeutic
and faintness. This syndrome is not blocked by atro thoracentesis provides symptomatic relief, but it is
pine. The recommended treatment is termination of not the definitive therapy. Pleural fluid collects be
the procedure and the immediate placement of the cause the rate of pleural fluid formation has exceeded
patient in a reverse Trendelenberg position. the rate of pleural fluid absorption. If nothing is done
Another complication of thoracentesis is infec to change these rwo factors, then pleural fluid will
tion of the pleural space. Approximately 2% of all reaccumulate after the therapeutic thoracentesis. The
pleural infections are due to contamination of the thoracentesis itself does not alter the basic condition
pleural space at the time of thoracentesis. For this rea that produced the pleural effusion. However, a thera
son, sterile technique must be strictly followed dur peutic thoracentesis should always be performed in
ing thoracentesis, and the skin must be thoroughly an acutely dyspneic patient with a large pleural effu
cleansed before the procedure is started. The treat sion to alleviate the dyspnea.
ment of pleural infections is discussed in Chapter 1 2 . Serial therapeutic thoracenteses can be performed
Diagnostic thoracentesis can also produce a he in patients who are dyspneic from malignant pleural
mothorax if an intercostal artery is lacerated. This effusions with mediastinal shift toward the contralat
complication can usually be avoided if thoracente eral side in which a pleurodesis cannot be successfully
sis is performed j ust superior to a rib, as previously performed. It is recommended, however, that such
described. In older patients, however, the intercostal patients have an indwelling catheter or a pleuroperito
arteries may be tortuous, and a hemothorax can re neal shunt placed. A therapeutic thoracentesis is also
sult even with proper technique (30) . The treatment indicated in a patient with a malignant pleural effu
of iatrogenic hemothorax is described in Chapter 2 5 . sion and dyspnea to determine whether the dyspnea
Other rare complications o f diagnostic thoracentesis can be relieved by the thoracentesis. This procedure
include splenic or hepatic laceration, soft tissue in should be performed before a chest tube is inserted
fection secondary to seeding of the needle tract with and pleurodesis is attempted (see Chapter 1 0) . The
bacteria, seeding of the needle tract with tumor cells, contraindications for therapeutic thoracentesis are
and adverse reactions to the local anesthetic. the same as for diagnostic thoracentesis.
Another uncommon complication of which one
should be aware is human immunodeficiency virus
Tech n i q u e
(HIV) infection with seroconversion. Oksenhendler
et al. (3 1 ) reported an instance in which a nurse re Th e positioning o f the patient and the selection o f the
ceived a superficial self-inflicted needle stick injury to site for the thoracentesis are the same as for a diag
the finger while recapping a needle contaminated by nostic thoracentesis. The most important difference
the bloody pleural fluid of a patient with persistent berween a therapeutic and a diagnostic thoracentesis
generalized lymphadenopathy, pleural effusion, and is that one must not use a sharp needle for the thera
seropositivity for HIV and hepatitis B surface anti peutic thoracentesis. As the fluid is removed, the lung
gen. Anicteric hepatitis developed 53 days later, and expands and can easily be lacerated if a sharp needle is
serum samples became HIV antibody positive by day present in the pleural space. Therefore, either a plastic
6 8 . This case emphasizes the need for strict precau catheter or a blunt pleural biopsy needle should be
tions regarding the handling of needles and body flu used for therapeutic thoracentesis. If a pleural biopsy
ids from patients infected with HIV. is also indicated, the therapeutic thoracentesis can be
performed through the pleural biopsy needle once
the biopsy specimens have been obtained.
TH E RA P E U T I C T H O RACE NTE S I S The additional materials required for the proce
dure are listed in Table 2 8 . 1 . It is recommended that
I n d ications
the Arrow-Clark™ Pleura-Sear or the Argyle'" Turkel'M
The three main indications for therapeutic thoracen thoracentesis kits be used for therapeutic thoracente
tesis are to remove the pleural fluid in patients with sis. Directions for their use come with the kits and
parapneumonic effusions or empyema, to relieve the additional information is available on the Internet.
symptom of dyspnea secondary to a pleural effusion, When these kits are used, it is important to make a
and to remove the pleural fluid so that the status of the large enough incision in the skin to allow easy pas
lung underlying a pleural effusion can be evaluated. sage for the needle with its overlying catheter. If the
With the exception of patients with parapneumonic incision is too small, the catheter may be damaged
4 54 PLE U RAL D I S EASES
during the insertion. It is important to use a kit with localized and identified by means of the lidocaine
a catheter over a needle, rather than one that only has filled syringe, as in diagnostic thoracentesis, a stan
a sharp needle. The Arrow-Clark� Pleura-Seal" or the dard 1 6-gauge (Intracath) needle is attached to a
ArgyleT• Turker" thoracentesis kits are recommended plastic syringe. With gentle constant suction on the
because they each contain the catheter and they each syringe, the needle is carefully and evenly advanced
have a device that prevents air from entering the pleu until pleural fluid is obtained. When the pleural fluid
ral space when the needle is withdrawn (see the dis has been obtained, the syringe is disconnected from
cussion on these kits earlier in this chapter) . the needle, and the needle is temporarily occluded
If these kits are not available, the procedure can by a finger to prevent the development of a pneu
be performed with a plastic catheter (Intracath) as mothorax. Then, the 1 6-gauge (Intracath) catheter is
outlined in Figure 2 8 . 3 . When the fluid has been inserted through the needle and is directed inferiorly
A B
+1 0
0 &-�
� -& -�- &-�
-o- - -s., -
6 "'
"8 - --o- - e - � -� ..... '€>
I
-1 0
E
�
I
� I
::i
(/) -20 I
(/) I
� I
c.. I
I
"§ -30 :i:
::i
Q) I
c::: I
I
I
-40 I
I
I
I
x
-50
400 800 1 ,200 1 , 600 2 , 000 2,400 2 , 800 3,200 3 , 600
Pleural fluid withdrawal (ml)
removed more than 4,000 mL pleural fluid in a single thoracentesis of greater than 1 ,000 ml and reported
thoracentesis from each of eight separate patients with that only five patients developed reexpansion pul
no adverse consequences. Villena et al. (34) confirmed monary edema. The reexpansion pulmonary edema
these observations in a series of 57 patients. They re was not related to pleural pressure or pleural pres
moved more than 1 , 500 mL from 29 patients with sure changes. In view of these latter series, it is rec
pleural pressure monitoring and had no instances of ommended that no limit be placed on the amount of
reexpansion pulmonary edema (34) . In their 57 pa pleural fluid withdrawn during a therapeutic thora
tients, the thoracentesis was stopped in 1 6 (28%) due centesis. However, the procedure should be stopped
to excessively negative ( < 20 cm H 2 0) pleural pres
- if the patient develops more than minimal coughing,
sure without any symptoms, in 29 patients (5 1 %) due chest tightness, chest pain, or shortness of breath.
to the development of symptoms (chest pain, cough,
or chest tightness) , in 1 0 patients ( 1 8%) because no
Com p l ications
more fluid could be obtained, and in 2 patients (4%)
because the physician considered that too much fluid Therapeutic thoracentesis is associated with the same
had been evacuated. In another study, Jones et al. ( 1 3) complications as diagnostic thoracentesis, includ
performed 94 1 therapeutic thoracenteses without ing vasovagal reaction, pneumothorax, pleural in
pressure monitoring. The procedure was terminated if fection, and hemothorax. In addition, reexpansion
the patient developed chest pain, more than minimal pulmonary edema and hypovolemia may complicate
coughing, or shortness of breath, or if no more fluid therapeutic thoracentesis, and, as mentioned ear
could be obtained. In this series, reexpansion pulmo lier, these complications may be related to the de
nary edema only occurred in two patients from whom velopment of abnormally negative pleural pressures.
1 ,000 and 1 ,200 mL pleural fluid had been with Pneumothorax is more common with therapeutic
drawn. In this series, more than 1 ,000 mL of fluid was than with diagnostic thoracentesis for two reasons.
removed with 20 1 procedures, more than 1 , 500 mL First, if a sharp needle is used for therapeutic tho
with 1 1 9 procedures, and more than 2,000 mL with racentesis, the lung is likely to be lacerated as it re
53 procedures ( 1 3) . Feller-Kopman et al. (35) mea expands, leading to a bronchopleural fistula and a
sured pleural pressures in 1 8 5 patients undergoing a pneumothorax. Second, because the pleural pressure
C HAPT E R 28 I T H O RAC E NTESIS (D IAG N OST I C A N D T H E RA P E UTIC) A N D P L E U RAL B I O PSY 45 7
a t times becomes quite negative during a therapeutic edema developed if the lung was reexpanded by
thoracentesis, air is more likely to enter the pleural underwater-seal drainage after 3 days of collapse or
space through faulty technique or even through the if negative pressure was used in a lung that had been
thoracentesis tract. collapsed for only 1 hour.
The treatment of iatrogenic pneumothorax is dis In a study in rabbits, Pavlin and Cheney (4 1 )
cussed in Chapter 24. A special situation exists in found that reexpansion pulmonary edema was much
some patients with malignant pleural effusion who more extensive in lungs that had been collapsed
undergo therapeutic thoracentesis. In the series of for 7 days than in those that had been collapsed
Boland et al. (36) , therapeutic thoracentesis was per for 3 days. Reexpansion with - 20 mm Hg pleural
formed on 5 1 2 patients with malignant pleural ef pressure led to no more edema than did reexpan
fusion over a 3-year period. Pneumothoraces were sion with positive airway pressure, but reexpansion
documented in 40 patients (8%), and 29 were treated with - 40 mm Hg or - 1 00 mm Hg increased the
with tube thoracostomy. However, the pneumotho amount of edema. In some of these animals, con
rax in 17 of the 29 (59%) persisted despite the ap tralateral pulmonary edema also developed, but to a
plication of suction and the insertion of larger (28 to lesser extent than in the ipsilateral lung (4 1 ) . Some
36 F) chest tubes. After the drainage catheters were cases of reexpansion pulmonary edema in humans,
removed, the effusion completely reaccumulated in however, have occurred when no negative pressure
all patients. This study demonstrates that a small was applied to the pleural space (42,43) . Almost all
subgroup of patients with malignant pleural effu cases of reexpansion pulmonary edema occur when
sions who undergo therapeutic thoracentesis develop the pneumothorax or pleural effusion has been pres
asymptomatic hydropneumothoraces due to poor ent for at least 3 days .
lung compliance. If a chest tube is inserted in these Reexpansion pulmonary edema appears to be
patients, the lung will not reexpand and there will due to increased permeability of the pulmonary vas
be no evidence of a bronchopleural fistula. In such a culature. In both humans (44) and rabbits (45 ) , the
situation, the chest tube should be removed as soon edema fluid has a high protein content, suggesting
as the situation is recognized. that it is leakiness of the capillaries rather than an in
creased hydrostatic pressure gradient that leads to the
edema. Pavlin et al. (46) have hypothesized that the
R E EX PA N S I O N P U L M O NARY E D E M A
mechanical stresses applied to the lung during reex
Reexpansion pulmonary edema is characterized by pansion damage the capillaries and lead to the devel
the development of unilateral pulmonary edema in a opment of pulmonary edema. There is no evidence
lung that has been rapidly reinflated following a vari that the collapsed lung has increased permeability
able period of collapse secondary to a pleural effusion before reinflation (46) .
or pneumothorax (37) . Unilateral pulmonary edema An alternate hypothesis that has become popular
is associated with a variable degree of hypoxia and is that reexpansion pulmonary edema is due to a re
hypotension, sometimes requiring intubation and perfusion injury (47,48) . With atelectasis, hypoxia of
mechanical ventilation, and occasionally leading to the atelectatic lung may be severe because oxygen de
death (38,39) . livery to the lung is reduced by absent ventilation and
hypoperfusion. Then when the hypoxic areas are re
perfused, oxygen-free radical formation is promoted
Path ophysiolog i c Featu res
and lung injury can result. If the lung is only partially
The exact mechanisms responsible for reexpansion collapsed, the reexpansion edema sometimes occurs
pulmonary edema are not known. In the experimen only in the part of the lung that has been atelectatic
tal animal, reexpansion pulmonary edema occurs (49) . Mechanical stress is probably not the sole fac
only if the lung has been collapsed for several days tor responsible for reexpansion pulmonary edema
and if negative pressure is applied to the pleural space. because the edema is associated with neutrophil in
Miller et al. (40) studied monkeys in which a pneu flux into the lung in both animals (50) and humans
mothorax had been present for 1 hour or 3 days. (5 1 ) , and the edema fluid contains interleukin-8 and
These researchers found that reexpansion pulmonary leukotriene B 4 • Neutrophils are not responsible for
edema occurred only when the pneumothorax had the reexpansion edema, however, because neutrophil
been present for 3 days and the lung was reexpanded depletion in the animal model does not prevent its
with - 1 0 mm Hg pleural pressure. No pulmonary occurrence (50) . The reperfusion injury hypothesis is
45 8 PLE U RAL D I S EASES
supported by the observation that the administration pneumothorax, the incidence of reexpansion pulmo
of an increased Fio 2 (40%) for the duration of the nary edema was 1 % (57) .
pneumothorax prevents edema when lungs are re
expanded (52) . Supplemental oxygen eliminates the
Preve ntion
systemic hypoxemia although the lung is collapsed.
Additional support for this hypothesis is provided The possibility of reexpansion pulmonary edema
by the observation that the administration of anti should be considered in patients with large pleural
oxidants before reexpansion minimizes both the per effusions or pneumothoraces of more than a few days'
meability edema and the degree of inflammation in duration who are undergoing tube thoracostomy or
rabbits (53). thoracentesis. When tube thoracostomy is performed
for spontaneous pneumothorax, the tubes should be
connected to an underwater-seal drainage apparatus
C l i n ical M a n ifestati o n s rather than to a negative pressure apparatus in view of
Patients who develop reexpansion pulmonary edema the animal studies of Miller et al. (40) and Pavlin and
typically develop pernicious coughing or chest tight Cheney (4 1 ) . If underwater-seal drainage does not ef
ness during or immediately following thoracentesis fect reexpansion of the underlying lung within 24 to
or chest tube placement. The cough sometimes is 48 hours, then negative pressure can be applied to the
productive of copious amounts of frothy pink spu pleural space.
tum. Other symptoms include dyspnea, tachypnea, When a therapeutic thoracentesis is performed,
tachycardia, fever, hypotension, nausea, vomiting, the procedure should be stopped if the patient devel
and cyanosis. The symptoms progress for 24 to 48 ops chest tightness, chest pain, shortness of breath, or
hours, and the chest radiograph reveals pulmonary more than minimal coughing. However, as discussed
edema throughout the ipsilateral lung. Pulmonary in the preceding text, no firm limit needs to be set on
edema may also develop in the contralateral lung the amount of fluid that is withdrawn.
(42,54) . If the patient does not die within the first
48 hours, recovery is usually complete. The serious Treatment
ness of the syndrome is emphasized by reports that it
has been responsible for the death of healthy, young The treatment of reexpansion pulmonary edema is pri
individuals. In one review of the subject, the outcome marily supportive with intravenous fluids, oxygen, and
was fatal in 1 1 of 53 reported cases (20%) (42) . The morphine. Diuresis may be detrimental and should be
overall mortality rate is probably much less than 20% avoided (57) . Suggested escalating levels of treatment
because fatal cases are more likely to be reported than include no treatment for an abnormality on radiog
are nonfatal cases. raphy alone; nasal supplemental oxygen for mild hy
The incidence of reexpansion pulmonary edema poxemia; continuous positive airway pressure through
is not known, but it is thought to be uncommon. face mask for moderate hypoxemia (58), and intu
Until 1 9 8 8 , a total of only 5 3 cases had been reported bation, mechanical ventilation, and PEEP for severe
(53). With therapeutic thoracentesis, the incidence is hypoxemia; and volume replacement and inotropic
very low. Jones et al. ( 1 3) reported that reexpansion agents for hypotension with low cardiac output (37) .
pulmonary edema occurred after only 2 of 94 1 thera
peutic thoracenteses. In this series, no pleural pres N E E D LE B I O PSY O F TH E P LE U RA
sure monitoring was performed and 1 1 9 patients had
Tra i n i n g Req u i re m e n ts
more than 1 , 500 mL pleural fluid withdrawn. Mila
nez de Campos et al. (55) reported that reexpansion The guidelines from the American College of Chest
pulmonary edema occurred in 2% of 500 patients Physicians recommend that physicians who perform
who underwent thoracoscopy and talc insuffiation needle biopsy of the pleura should be competent
for treatment of recurrent pleural effusion. In the in thoracentesis, familiar with the mechanisms and
Veterans Administration (VA) cooperative study on technique of the biopsy needle being used and com
spontaneous pneumothorax, there were no cases of petent to recognize and treat the common complica
reexpansion pulmonary edema among the 229 study tions. Trainees should perform at least five procedures
subjects despite the use of suction in more than 80% in a supervised setting to establish basic competency.
of the cases (56) . In another study of 320 patients Then to maintain competency, dedicated operators
treated with tube thoracostomy for spontaneous should perform at least five procedures per year (59) .
C HAPT E R 28 I T H O RAC E NTESIS (D IAG N OST I C A N D T H E RA P E UTIC) A N D P L E U RAL B I O PSY 459
I n d ications Tech n i q u e
With a needle biopsy of the pleura, a small piece of Th e materials necessary fo r pleural biopsy are listed
the parietal pleura is obtained for microscopic or in Table 2 8 . 1 . Most frequently, one uses a thoracen
microbiologic evaluation. The main diagnoses estab tesis kit plus the pleural biopsy needle. When there
lished with a needle biopsy of the pleura are tuber is a moderate or larger pleural effusion, the biopsy is
culous pleuritis and malignancy of the pleura. The usually done with no imaging. If the effusion is small
biopsy can either be done blindly or it can be image or loculated, then either ultrasound or computed to
guided. Currently, blind needle biopsy of the pleura mography (CT) can accurately identify the location
is used less than in the past because the diagnosis of of the fluid. Ultrasound is the preferred technique for
tuberculous pleuritis can be made by measuring the guiding biopsy because it offers the advantage of a
adenosine deaminase (ADA) or interferon-gamma real-time approach to the biopsy and has the added
level in the pleural fluid, the diagnosis of pleural ma advantages of the absence of ionizing radiation, por
lignancy is usually established by pleural fluid cytol tability, ready availability, and low expense. Because
ogy or thoracoscopy (60) and image guided biopsies the patient can be imaged in the erect position, the
provide higher yields (6 1 ,62) . At the University of depth of the fluid is maximized, thereby minimizing
Texas, Galveston, 17 4 biopsies of the pleura were complications (64) .
performed between 1 996 and 2006 including 1 03 The patient is positioned, and the site is selected
(59%) blind, 3 8 (22%) image guided, and 33 ( 1 9%) as for diagnostic thoracentesis (described earlier in
thoracoscopically or at thoracotomy (6 1 ) . The diag this chapter) . The skin is cleaned, and the local anes
nostic yield with the three procedures was 42%, 79%, thetic is administered as for diagnostic thoracentesis.
and 93%, respectively (6 1 ) . The proportion of blind Liberal amounts of lidocaine should be injected once
pleural biopsies declined from 78% in 1 996 to 27% the rib is passed to ensure adequate anesthesia of the
in 2006 (6 1 ) . parietal pleura. In general, if no fluid is obtained with
A needle biopsy o f the pleura i s currently recom the local anesthetic, biopsy should not be attempted.
mended when tuberculous pleuritis is suspected and When pleural fluid has been obtained with the lido
the pleural fluid ADA or interferon-gamma levels caine syringe and needle, a pleural biopsy can be per
are not definitive. A needle biopsy of the pleura is formed with an Abram's or a Cope needle. A biopsy
also recommended when malignancy is suspected is sometimes attempted without free pleural fluid. If
but the pleural fluid cytology is negative. It is rec there is no fluid, the procedure should be performed
ommended that most needle biopsies be image with ultrasonic or CT guidance (64) .
guided.
Abra m 's Needle
incision should be made along the lines of cleavage with the other hand to cut off a small piece of parietal
to minimize postoperative scarring. The inner cutting pleura (Fig. 28 .7C) . Usually, mild resistance is met
cannula (Fig. 28 .6B) fits tightly in the outer trocar immediately before the needle is completely closed,
(Fig. 2 8 . 6A) and can be locked in one of two posi and this resistance is because the inner cannula is sev
tions: (a) a closed position in which the inner can ering the entrapped pleura for the biopsy specimen.
nula obstructs the notch on the outer trocar to make Once the initial biopsy specimen is obtained, the
the needle airtight and (b) an open position in which needle can either be withdrawn from the pleural space
the inner cannula is slightly withdrawn so that the in the closed position or reinserted into the pleural
notch on the outer trocar is not occluded. An indi space. If the needle is withdrawn from the chest,
cator knob in the hexagonal grip of the larger outer the pleural biopsy specimen is found in the tip of the
trocar indicates the position of the notch in the distal needle and can be closely examined, but the needle
end of the trocar. then has to be reinserted. Reinsertions of the needle
To insert the Abram's pleural biopsy needle, the are through the same tract, however, and are easier
styler is placed in the inner cannula, which, in turn, than the original insertion. If the needle is reinserted
is placed in the outer trocar. The inner cannula into the pleural space without a complete withdrawal,
(Fig. 28 .6B) is twisted clockwise to close the distal the tissue specimen can be aspirated through the sy
notch of the outer trocar. The needle is pushed into ringe. The biopsy procedure can be repeated without
the pleural space by exerting firm pressure on the
styler. Because the needle has a large diameter and
is blunt, a substantial amount of pressure is needed.
Usually, a pop is heard as the needle enters the pleural
space. The inability to pass the needle into the pleural
space is usually because of an insufficiently large skin
incision. At times, the ribs are too close together to
allow the needle to pass. In such situations, rotation
of the patient's arm and shoulder over his or her head A
frequently separates the ribs sufficiently.
Once the tip of the needle is thought to be in the
pleural space, the inner styler (Fig. 2 8 .6C) is removed,
and with the inner cannula in the closed position,
a syringe is attached to the connection on the inner I/ji
cannula. Then, the inner cannula is rotated counter
clockwise in the outer trocar so that the distal notch 1
I
A B
FIGURE 28.9 • A: The pa rieta l p l e u ra is hooked with the h o l l ow, b l u nt-ti pped b i o psy troca r. B: The b i o psy
spec i m e n is obta i n ed by a dva n c i n g the outer ca n n u l a with a rota ry motion (arrows) to sever the e n g a g ed
p i ece of p l e u r a .
large outer cannula toward the pleural space using a larger biopsy specimen, and it is safer for concomi
rotary motion to sever the engaged piece of pleura tant therapeutic thoracentesis because the end of the
(Fig. 28.9B). Then, the hooked biopsy trocar contain outer cannula is blunt.
ing the tissue specimen is removed while the patient
holds his breath, and the hooked trocar is replaced
I m age-G u i ded Needle B i o psy
with the beveled trocar and obturator before the pro
cedure is repeated for an additional biopsy specimen. The use of image-guided needle biopsy of the pleura
Once the required biopsy specimens have been ob has become much more widespread in the past de
tained, a therapeutic thoracentesis may be performed cade. The reason for this is that it is more effective at
by attaching a large syringe and a three-way stopcock establishing a diagnosis than is blind needle biopsy.
to the outer cannula. The biopsy can be either CT (62,69) or ultrasound
The biopsy site and the biopsy specimens are han guided (70) and either an Abram's needle (69,70)
dled identically with both Abram's and Cope pleural or a cutting-needle (62,70) can be used. Maskell et
biopsy needles. al. (62) randomly assigned 50 patients with unilat
eral pleural effusion and a clinical suspicion of ma
Abra m 's versus Cope Needles
lignancy to Abram's needle biopsy of the pleura or
The rate of success in obtaining a pleural biopsy CT-guided cutting-needle biopsy. The CT-guided bi
specimen depends more on the skill of the operator opsy was made where the pleura was the most thick
than on the choice of instruments (67) . Morrone et ened, and, at most, two biopsies were performed.
al. (68) obtained pleural biopsies simultaneously with The Abram's needle biopsy was not necessarily aimed
the Abram's and Cope needles, and they reported where the pleural thickening was greatest. The diag
that the diagnostic yields were virtually identical. The nosis of malignancy was made in 1 3 of 1 5 patients
Abram's needle provided slightly larger specimens (87%) with the CT-guided biopsy but in only 8 of 1 7
and was slightly superior in detecting mesothelial patients (47%) with the Abram's needle (62) . These
cells. The Cope needle provided larger specimens of results look very promising but it should be noted
intercostal muscle. The Abram's needle is generally that 20 of the 27 patients with malignancy in this
preferred to the Cope needle, however, because it is study had mesothelioma that is notoriously difficult
easier to use, it is a closed system and hence the likeli to diagnose with Abram's needle biopsy of the pleura.
hood of a pneumothorax is decreased, it provides a Metintas et al. (7 1 ) earlier reported that they were
C HAPT E R 28 I T H O RAC E NTESIS (D IAG N OST I C A N D T H E RA P E UTIC) A N D P L E U RAL B I O PSY 463
able to establish the diagnosis of malignant mesothe effects. Penetration of the spleen frequently requires
lioma in 25 of 30 patients (83.3%) with CT-guided splenectomy (75 ) , however, and one should therefore
closed pleural needle biopsy using an Abram's pleu be careful not to perform pleural biopsy or thoracen
ral biopsy needle. Metintas et al. (69) subsequently tesis too far inferiorly on the left side.
randomized 1 24 patients suspected of having pleural
malignancy to CT guided biopsy with an Abram's
needle or thoracoscopy. They reported that the di REFERENCES
agnostic sensitivity was 87.5% in the needle biopsy 1 . Thomsen TW, DeLaPena J, Setnik GS. Videos in clinical med
group and 94. 1 % in the thoracoscopy group which icine. Thoracentesis. N Engl] Med. 2006; 3 5 5 : e l 6 .
did not differ significantly (69) . In the needle biopsy 2 . Adams FV, Galati V. M-mode ultrasonic localization o f pleu
group the diagnostic sensitivity was 80% with ma ral effusion. JAMA. 1 978;239: 1 76 1 - 1 764.
3 . McVay PA, Toy PT. Lack of increased bleeding after paracen
lignant mesothelioma ( 1 2/ 1 5 ) , 93% with malignant
tesis and thoracentesis in patients with mild coagulation ab
effusion due to lung cancer ( 1 4/ 1 5 ) , 8 8 % with ma normalities. Transfusion. 1 9 9 1 ;3 1 : 1 64- 1 7 1 .
lignancy from other organs (7/8), and 90% in tuber 4. Bass J , White DA. Thoracentesis i n patients with hematologic
culous pleuritis (9/ 1 0) (69) . Koegelenberg et al. (70) malignancy: yield and safery. Chest. 2005 ; 1 27:2 1 0 1 -2 1 0 5 .
using ultrasound guidance performed four Abram's 5 . Mayo P H , Goltz HR, Tafreshi M, e t al. Safery o f ultrasound
guided thoracentesis in patients receiving mechanical ventila
needle biopsies and four Tru-cut needle biopsies on tion. Chest. 2004; 1 2 5 : 1 059-1 062.
89 patients suspected of having tuberculous pleuritis. 6. McCartney JP, Adams JW, Hazard PB. Safery of tho
They reported that biopsies with the Abram's needle racentesis in mechanically ventilated patients. Chest.
were more likely to contain pleural tissue (9 1 %) than 1 993; 1 03: 1 920-1 9 2 1 .
7. Godwin JE, Sahn SA. Thoracentesis: a safe procedure
were the biopsies obtained with the Tru-cut needle
in mechanically ventilated patients. Ann Intern Med.
(79%) and were more likely to establish the diagnosis 1 990; 1 1 3 : 800-802.
of tuberculous pleuritis (82% vs. 65%) (70) . 8 . Weingardt JP, Guico RR, Nemcek AA Jr, et al. Ultrasound
findings following failed, clinically directed thoracenteses. J
Clin Ultrasound. 1 994;22:4 1 9-426.
Co m p l icat i o n s 9 . Lipscomb DJ, Flower CDR, Hadfield JW. Ultrasound of
the pleura: an assessment of its clinical value. Clin Radio!.
Pleural biopsy has the same complications as diagnos 1 9 8 1 ;32:289-290.
tic thoracentesis. One might expect pneumothorax 10. Diacon AH, Brutsche MH, Soler M . Accuracy of pleural
to be more common with pleural biopsy than with puncture sites: a prospective comparison of clinical examina
thoracentesis for two reasons. First, the atmosphere tion with ultrasound. Chest. 2003; 1 23:436-44 1 .
1 1 . Kohan JM, Poe RH, Israel RH, et al. Value o f chest ultraso
has much more opportunity to be in communication nography versus decubirus roentgenography for thoracentesis.
with the pleural space with the biopsy, particularly Am Rev Respir Dis. 1 986; 1 3 3 : 1 1 24- 1 1 26.
when the Cope needle is used. Second, when the bi 12. Grogan DR, Irwin RS, Channick R, et al. Complications
opsy specimen is obtained, the visceral pleura may be associated with thoracentesis. A prospective, randomized
study comparing three different methods. Arch Intern Med.
inadvertently incised, leaving a small bronchopleural
1 990; 1 5 0 : 873-877.
fistula that can lead to a large pneumothorax. How 1 3 . Jones PW, Moyers JP, Rogers JT, et al. Ultrasound-guided tho
ever, the incidence of pneumothorax and the require racentesis: is it a safer method? Chest. 2003 ; 1 23 : 4 1 8-423.
ment for tube thoracostomy are comparable after 14. Havelock T, Teoh R, Laws D , et al. BTS Pleural Disease
thoracentesis and pleural biopsy (72) . This is prob Guideline Group. Pleural procedures and thoracic ultrasound:
British Thoracic Sociery Pleural Disease Guideline 20 1 0 . Tho
ably because more experienced individuals usually
rax. 2 0 1 O; 65 suppl 2:ii6 l -76.
perform the pleural biopsy. 1 5 . Conner BD, Lee YCG, Branca P, et al. Variations in pleural fluid
The second major complication of pleural biopsy WBC count and differential counts with different sample con
is bleeding. If an intercostal artery or vein is inad tainers and different methods. Chest. 2003; 1 2 3 : 1 1 8 1 - 1 1 87.
vertently biopsied, a hemothorax can result (63,73) . 16. Cheng D-S, Rodriguez RM , Rogers J, et al. Comparison of
pleural fluid pH values obtained using blood gas machine, pH
There is one case report of an arteriovenous fistula
meter, and pH indicator strip. Chest. 1 998; 1 1 4: 1 368-1 372.
from an intercostal artery to an intercostal vein de 17. Rahman NM, Mishra EK, Davies HE, et al. Clinically im
veloping after pleural biopsy (74) . The fistula subse portant factors influencing the diagnostic measurement of
quently ruptured causing a hemothorax. The pleural pleural fluid pH and glucose. Am J Respir Grit Care Med.
biopsy needle can also be mistakenly inserted into 2008; 1 78:483-490.
18. Sarodia BD, Goldstein LS, Laskowski DM, et al. Does
the liver, spleen, or kidney. Although hepatic or kid pleural fluid pH change significantly at room tempera
ney tissue may be demonstrated in the biopsy speci ture during the first hour following thoracentesis? Chest.
men, the patient usually suffers no significant adverse 2000; 1 1 7 : 1 043- 1 04 8 .
464 PLE U RAL D I S EASES
19. Menzies SM, Rahman NM, Wrighrson JM, et al. Blood cul 4 1 . Pavlin J, Cheney FW Jr. Unilateral pulmonary edema in
ture bottle culture of pleural fluid in pleural infection. 7horax. rabbits after re-expansion of collapsed lung. J Appl Physiol.
20 1 1 ;66:65 8-662. 1 979;46:3 1 -3 5 .
20. Seneff MG, Corwin RW, Gold LH, et al. Complications as 4 2 . Mahfood S, Hix WR, Aaron Bl, e t al. Re-expansion pulmo
sociated with thoracocentesis. Chest. 1 9 86;90:97- 1 00. nary edema. Ann 7horac Surg. 1 988;45 :340-345.
2 1 . Wrightson JM, Fysh E, Maskell NA, et al. Risk reduction in 43. Olcott EW. Fatal reexpansion pulmonary edema fol
pleural procedures: sonography, simulation and supervision. lowing pleural catheter placement. J Vase Interv Radio/.
Curr Op in Pulm Med. 20 I O ; 1 6:340-3 5 0 . 1 994;5: 1 76- 1 7 8 .
22. Duncan D R , Morgenthaler T I , Ry u JH, et a l . Reducing iatro 4 4 . Waqaruddin M, Bernstein A. Re-expansion pulmonary
genic risk in thoracentesis. Chest. 2009; 1 3 5 : 1 3 1 5- 1 320. edema. 7horax. I 975;30: 54-60.
23. Daniels CE, Ryu JH. Improving the safety of thoracentesis. 4 5 . Sprung CL, Loewenherz JW, Baier H, et al. Evidence for in
Curr Opin Pulm Med. 20 1 1 ; 1 7:232-236. creased permeability in re-expansion pulmonary edema. Am
24. Aleman C, Alegre J, Armadans L, et al. The value of chest j Med. 1 98 1 ;7 1 :497-5 00.
roentgenography in the diagnosis of pneumothorax after tho 46. Pavlin DJ, Nessly ML, Cheney FW Increased pulmonary vas
racentesis. Am J Med. 1 999; I 07:340-343. cular permeability as a cause of re-expansion edema in rabbits.
25. Patel PA, Ernst FR, Gunnarsson CL. Ulrrasonography guid Am Rev Respir Dis. 1 9 8 1 ; 1 24:422-427.
ance reduces complications and costs associated with thora 47. Pavlin DJ . Lung re-expansion: for better or worse. Chest.
centesis procedures. ] Clin Ultrasound. 20 1 2;40: 1 3 5-1 4 1 . 1 986;89 :2-3 .
26. Raptopoulos V, Davis LM, Lee G , e t al. Factors affecting the 4 8 . Sivrikoz MC, Tuncozgur B, Cekmen M, et al. The role of tis
development of pneumothorax associated with thoracentesis. sue reperfusion in the reexpansion injury of the lungs. Eur
AJR Am j Roentgenol. 1 99 1 ; 1 56:9 1 7-920. J Cardiothorac Surg. 2002;22: 72 1 -727.
27. Brandstetter RD, Karerzky M, Rastogi R, et al. Pneumorhorax 49. Woodring JH. Focal reexpansion pulmonary edema after
after rhoracentesis in chronic obstructive pulmonary disease. drainage of large pleural effusions: clinical evidence suggesting
Heart Lung. 1 9 94;23:67-70. hypoxic injury to the lung as the cause of edema. South Medj.
28. Gervais DA, Petersein A, Lee MJ , et al. US-guided thoracen 1 997;90: 1 1 76- 1 1 82.
tesis: requirement for postprocedure chest radiography in pa 5 0 . Jackson RM, Veal CF, Alexander CB, et al. Neutrophils in
tients who receive mechanical ventilation versus patients who reexpansion pulmonary edema. J Appl Physiol. 1 9 88;65:
breathe spontaneously. Radiology. 1 9 97;204 : 503-5 06. 228-234.
29. Doyle ]], Hnatiuk OW, Torrington KG, et al. Necessity of 5 1 . Nakamura H, Ishizaka A, Sawafuji M, et al. Elevated levels of
routine chest roentgenography after rhoracentesis. Ann Intern interleukin-8 and leukorriene B 4 in pulmonary edema fluid of
Med. 1 996; 1 24 : 8 1 6-820. a patient with reexpansion pulmonary edema. Am J Respir Crit
30. Carney M, Ravin CE. Intercostal artery laceration dur Care Med. 1 994; 1 4 9 : 1 037- 1 040.
ing thoracentesis. Increased risk in elderly patients. Chest. 52. Pavlin DJ, Nessly ML, Cheney FW. Hemodynamic effects of
1 979;75 : 5 20-52 1 . rapidly evacuating prolonged pneumothorax in rabbits. J Appl
3 1 . Oksenhendler E , Harzic M , Le Roux JM, et al. HIV infection Physiol. 1 9 87;62:477-484.
with seroconversion after a superficial needlestick injury to the 53. Jackson RM , Veal CF, Alexander CB, et al. Re-expansion pul
finger. N Engl] Med. 1 9 86;3 1 5 : 58 2. monary edema: a potential role for free radicals in its patho
32. Light RW, Jenkinson SG, Minh V, e t al. Observations on pleu genesis. Am Rev Respir Dis. 1 9 8 8; 1 37: 1 1 65-1 1 7 1 .
ral pressures as fluid is withdrawn during thoracentesis. Am 54. Heller BJ, Grathwohl MK. Contralateral reexpansion pulmo
Rev Respir Dis. 1 9 80; 1 2 1 :799-804. nary edema. South Medj. 2000;93: 828-83 1 .
33. Doelken P, Huggins JT, Pastis NJ, et al. Pleural ma 5 5 . de Campos JRM, Vargas FS, Werebe EC, et al. Tho
nometry: technique and clinical implications. Chest. racoscopy talc poudrage. A 1 5-year experience. Chest.
2004; 1 26: 1 764- 1 769. 200 1 ; 1 1 9 : 8 0 1 -806.
34. Villena V, Lopez-Encuentra A, Pozo F, et al. Measurement of 56. Light RW, O'Hara VS, Moritz TE, et al. Intrapleural tetracy
pleural pressure during therapeutic thoracentesis. Am J Respir cline for the prevention of recurrent spontaneous pneumotho
Crit Care Med. 2000; 1 62 : 1 534- 1 538. rax. JAMA. I 990;264:2224-2230.
3 5 . Feller-Kopman D , Berkowitz D, Boiselle P, et al. Large 57. Rozenman J, Yellin A, Simansky DA, et al. Re-expansion pul
volume rhoracentesis and the risk of reexpansion pulmonary monary edema following spontaneous pneumothorax. Respir
edema. Ann 1horac Surg. 2007;84: 1 65 6- 1 66 1 . Med. 1 996;90:23 5-23 8 .
36. Boland GW, Gazelle GS, Girard MJ, e t al. Asymptom 5 8 . Iqbal M, Multz AS, Rossoff LJ, e t a l . Reexpansion pulmonary
atic hydropneumothorax after therapeutic thoracente edema after VATS successfully created with continuous posi
sis for malignant pleural effusions. AJR Am J Roentgenol. tive airway pressure. Ann 1horac Surg. 2000;70: 669-67 1 .
1 99 8 ; 1 70: 943-946. 5 9 . Ernst A, Silvestri GA, Johnstone D , et al. American College
37. Tarver RD, Broderick LS, Conces DJ Jr. Reexpansion pulmo of Chest Physicians. Interventional pulmonary procedures:
nary edema. J 1horac Imaging. 1 996; 1 1 : 1 9 8-20 8 . guidelines from the American College of Chest Physicians.
3 8 . Trapnell DH, Thurston JGB. Unilateral pulmonary edema Chest. 2003; 1 23 : 1 693-1 7 1 7.
after pleural aspiration. Lancet. 1 970; 1 : 1 367- 1 369. 60. Light RW Closed needle biopsy of the pleura is a valuable
3 9 . Peatfield RC, Edwards PR, Johnson NM. Two unexpected diagnostic procedure. Con closed needle biopsy. J Bronchol.
deaths from pneumothorax. Lancet. 1 979; 1 :356-3 5 8 . 1 998;5 :332-336.
40. Miller WC, Toon R , Palat H, e t al. Experimental pulmonary 6 1 . Gopal M, Romero AB, Baillargeon J, et al. Trends in pleural
edema following re-expansion of pneumothorax. Am Rev biopsies berween 1 996 and 2006 at a tertiary medical center.
Respir Dis. 1 973; 1 08 : 664-666. Am j Med Sci. 2 0 1 0;339:345-349.
C HAPT E R 28 I T H O RAC E NTESIS (D IAG N OST I C A N D T H E RA P E UTIC) A N D P L E U RAL B I O PSY 465
62. Maskell NA, Gleeson FV, Davies RJ. Standard pleural biopsy needle biopsy for diagnosis of patients with pleural effusions:
versus CT-guided cutting-needle biopsy for diagnosis of ma a randomized controlled trial. Chest. 20 1 0 ; 1 37: 1 362- 1 368.
lignant disease in pleural effusions: a randomised controlled 70. Koegelenberg CF, Bolliger CT, Theron J, et al. A direct com
trial. Lancet. 2003;36 1 : 1 326- 1 330. parison of the diagnostic yield of ultrasound-assisted Abrams
63. Levine H, Cugell DW Blunt-end needle biopsy of pleura and and Tru-cut needle biopsies for pleural tuberculosis. Thorax.
rib. Arch Intern Med. 1 97 1 ; 1 09 : 5 1 6-52 5 . 2 0 1 0;65 : 8 57-862.
6 4 . Screaton NJ, Flower CD. Percutaneous needle biopsy of the 71. Metintas M, Ozdemir N, Isiksoy S, et al. CT-guided pleural
pleura. Radio! Clin North Am. 2000;38 :293-30 1 . needle biopsy in the diagnosis of malignant mesothelioma.
6 5 . Ogirala RG, Agarwal V, Aldrich TK. Raja pleural biopsy nee J Comput Assist Tomogr. 1 99 5 ; 1 9 :370-374.
dle. A comparison with the Abrams needle in experimental 72. Poe RH, Israel RH, Utell MJ, et al. Sensitivity, specificity, and
pleural effusion. Am Rev Respir Dis. 1 989; 1 39 : 9 84-987. predictive values of closed pleural biopsy. Arch Intern Med.
66. Ogirala RG, Agarwal V, Vizioli LD, et al. Comparison of the 1 984; 1 44:325-328.
Raja and the Abrams pleural biopsy needles in patients with 73. Ali J, Summer WR. Hemothorax and hyperkalemia after pleu
pleural effusion. Am Rev Respir Dis. 1 993; 1 47: 1 2 9 1 - 1 294. ral biopsy in a 43-year-old woman on hemodialysis. Chest.
67. Walsh LJ, Macfarlane JT, Manhire AR, et al. Audit of pleural 1 994; 1 06: 1 235-1 236.
biopsies: an argument for a pleural biopsy service. Respir Med. 74. Lai JH, Yan HC, Kao SJ, et al. lntercostal arteriovenous fistula
1 994; 8 8 : 503-5 0 5 . due to pleural biopsy. Thorax. 1 990;4 5 : 976-978.
68. Morrone N, Algranti E , Barrero E. Pleural biopsy with Cope 75. Mearns AJ. Iatrogenic rupture of the spleen. Br Med J
and Abram's needles. Chest. 1 987;92 : 1 050- 1 052. 1 973; 1 :395-396.
69. Metintas M, Ak G, Dundar E, et al. Medical thoracos
copy versus computed tomography guided Abram's pleural
C h est Tu b e s
Chest tubes are frequently used in the practice of pul series (6) to date, 1 ,092 patients had ( 1 2 F) catheters
monary medicine. Indeed, in 1 99 5 , it was estimated inserted using guidewires. The percentages of suc
that 1 ,330,000 chest tubes were placed in the United cessful cases were 9 3 . 8 % of 324 malignant effusions,
States ( 1 ) . However, it has been my impression that 93% of 399 pneumothoraces, 92.3% of 272 non
many medical personnel do not understand how the malignant effusions, and 74.2% of 97 empyemas (6) .
drainage system for chest tubes functions and how
to troubleshoot problems with chest tubes. In this
C H E ST TU B E I N S E RT I O N
chapter, the various methods of inserting chest tubes
are discussed, followed by a more in-depth discussion I n general, chest tubes are inserted into the pleural
of the different drainage systems used with chest space by four methods: tube thoracostomy with a
tubes, plus recommendations for troubleshooting guidewire and dilators, tube thoracostomy with a tro
related problems. The indications for chest tube inser car, operative tube thoracostomy, and tube thoracos
tion with pneumothorax, hemothorax, empyema, and copy through a single-port thoracoscope. There are
malignant pleural effusion are discussed in the respec no controlled studies comparing the efficacy of the
tive chapters on these entities. different methods of placing chest tubes (2) . It has
been recommended that most chest tubes be placed
in the "triangle of safety" which is the area bordered
C H E ST TU B E S I Z E
by the lateral edge of the latissimus dorsi and the lat
I n general, there has been a tendency over the past eral border of the pectoralis major and superior to the
decade to use smaller ( 1 0- 1 4 F) chest tubes (2-4) . The horizontal level of the fifth intercostal space (4) . This
advantages of the small-bore catheters are that they position minimizes the risk to underlying structures
are easier to insert and there is less pain with their and avoids damage to muscle and breast tissue.
insertion and while they are in place. Small-bore
chest tubes are recommended when pleurodesis is
Tra i n i n g Req u i re m e n ts
performed. Small-bore chest tubes are also recom
mended for the treatment of pneumothorax. However, The guidelines from the American College of Chest
patients on mechanical ventilation with barotrauma Physicians recommend that trainees should perform
induced pneumothoraces are best managed with at least 1 0 procedures in a supervised setting to estab
large-bore chest tubes (2) . Small-bore catheters are lish basic competency. Then to maintain competency,
also recommended for complicated parapneumonic dedicated operators should perform at least five
effusions (5). When a small-bore chest tube is used procedures per year (7) .
to treat a complicated parapneumonic effusion,
it should be Bushed with 20 to 30 ml saline every
G u idewire Tu be Thoracosto my
6 hours via a three-way stopcock (5) . Patients with
hemothorax are best managed with large-bore This is probably the easiest way to insert a chest tube.
(> 20 F) chest tubes because of blood clots and In many hospitals, chest tubes are inserted by a radiol
the high volume of pleural fluid (2) . In the largest ogist, who uses this technique with either ultrasound
466
C H A PT E R 29 I C H EST TU B E S 467
or computed tomography (CT) guidance (8,9) . Com The wire should always project out beyond the end
mercial kits are available for guidewire tube thoracos of the dilator or inserter. The first dilator is removed,
tomy. This procedure uses the Seldinger technique leaving the wire in place. Then the next size dilator is
with guidewires and dilators ( 1 0) . In general, after the advanced over the guidewire into the pleural space and
skin, periosteum, and parietal pleura are anesthetized removed. Finally, the chest tube containing the inserter
as is done for pleural biopsy (see Chapter 28), an inci is threaded over the guidewire (Fig. 29. l E) . The tube
sion is made in the skin that is ample to permit passage should pass readily, following the path made by the
of the chest tube of desired size (Fig. 29. lA) . Then, an dilators and guided by the wire ( 1 0) . It is important
1 8-gauge needle attached to a syringe is introduced to make certain that all of the side holes in the tube are
into the pleural space. Fluid or air is aspirated to con in the pleural space.
firm the intrapleural position (Fig. 29. l B) . The syringe Once the tube is in place, the inserter and the
is removed, and the "J" wire is threaded through guidewire are withdrawn (Fig. 29 . 1 F) . The tube is
the needle in the desired direction into the pleural then clamped until it is attached to the chest drain
space (Fig. 29. l C) . The needle is then removed, and age system. The tube is anchored in place by means
more local anesthetic is injected into the intercostal of a long suture through the skin and around the
muscles surrounding the wire. The smallest dilator is tube. The incision is sutured without tension to avoid
inserted, and with a rotating movement, it is advanced necrosis of the skin next to the tube. The operative area
into the pleural space over the guidewire (Fig. 29 . 1 D) . is cleaned and is covered with plain 4 X 4 gauze pads.
I�
RIB
"'
11
Ol
c
::::i
....J
A
Pleural
space
B
\\J1 , \j\1 , c
� /
D E F
The gauze is then covered with tape, and additional (Fig. 29 .2B) . The trocar is removed by sliding it back
fixation of the tube is obtained by the tape. over the tube. When the trocar has been removed
There are many different types of chest tubes that from the chest, the chest tube is clamped between
can be inserted with the technique mentioned in the trocar and the chest wall so that the clamp on
the preceding text. Kits are made so that chest tubes the distal end of the chest tube can be removed. The
from 8 . 0 to 36.0 F can be inserted. The kits with chest tube must remain clamped until it is attached
the smaller chest tubes ( < 1 2 .0 F) have no dilators. to an underwater seal to prevent air from entering the
Some catheters have different characteristics. For pleural space.
example, with the Wayne Pneumothorax Set (Cook An alternate trocar method uses a chest tube with
Medical Incorporated, Bloomington, IN) , a 1 0 .2 a trocar positioned inside the tube. The procedure
or 1 4 . 0 F size catheter is inserted over a 1 9-gauge with this apparatus is similar to that already detailed.
needle; the catheter with this set is curved at the end Once the pleural cavity is entered, the inner trocar
like a pig's tail, and hence the name pigtail catheter. is gradually removed from the chest tube. When the
The Thal-Quick Chest Tube Sets come with chest proximal end of the trocar clears the chest wall, a
tube sizes from 8 . 0 to 36.0 F (Cook Medical Incor clamp is placed between the trocar and the chest wall
porated, Bloomington, IN, 800-457-4500, www. until the trocar can be completely withdrawn and the
cookcriticalcare.com) . tube attached to a water-seal drainage system.
This method is similar to the guidewire tube thora Dev et al. ( 1 1 ) have created an excellent video on per
costomy, except that there is no guidewire or dilators. forming operative tube thoracostomy, and watching
In general, it is not recommended because its use this video is recommended before one performs their
appears to be associated with more complications ( 1 ) . first operative tube thoracostomy. With this method,
This method initially requires a 2 - t o 4-cm incision an incision is made in the chest wall, and then after
parallel to the superior border of the rib through the blunt dissection with a hemostat, the operator places
skin and subcutaneous tissues after local anesthesia his finger into the pleural space to break adhesions
is obtained. The trocar can then be inserted between between the lung and chest wall and to ascertain the
the ribs into the pleural cavity, with the flat edge position of the chest tube. It is a more extensive pro
of the stylet tip cephalad to prevent damage to the cedure than guidewire tube thoracostomy or trocar
intercostal vessels (Fig. 29.2A) . Because significant tube thoracostomy, but it is probably safer. The most
force is often required to insert the trocar, the hand serious complications of tube thoracostomy are inser
not applying the force should be placed next to the tion of the tube ectopically, namely, into the lung,
patient's chest wall to control the depth of penetra stomach, spleen, liver, or heart. These complications
tion. Once the trocar is in the pleural space, the stylet are more likely when a trocar chest tube is used.
is removed. When the stylet is removed, the operator With the operative method, digital exploration of the
should immediately cover the trocar with the thumb insertion site delineates whether the tract leads into
to prevent a pneumothorax. Then, when the operator's the pleural space and whether any tissue or organ is
thumb is removed, the chest tube, with its distal end adherent to the parietal pleura at the planned site of
clamped, is quickly inserted into the pleural space tube insertion ( 1 2) .
Lung
A B
FIGURE 29.2 • Trocar tube tho racosto my. A: I nsert i o n of troca r i nto t h e p l e u ra l space. N ote the position
of the h a n ds, the position of the troca r re l ative to the ri bs, and the cepha l a d posit i o n of the flat edge of the
troca r. B: I nsertion of ch est t u b e through the troca r.
C H A PT E R 29 I C H EST TU B E S 469
Lung
A 8
FIGURE 29.3 • O p e rative t u b e thoracosto my. A: The physici a n 's i n dex f i n g e r is used to e n l a rg e the open i n g
a n d t o exp l o re the p l e u r a l space. B : P l acement o f ch est t u b e i ntra p l e u ra l ly u s i n g a l a rg e h e m ostat.
470 PLE U RAL D I S EASES
To patient
Vent
Water-seal Collection
bottle bottle
and therefore the pressure will have to be higher and an Emerson pump. In many facilities, however, suc
higher in the rigid straw to allow additional air or tion is provided by wall suction or other pumps in
fluid to exit from the pleural space. Another disad which the level of suction is not easily controlled.
vantage of this system is that if the bottle is inadver Because uncontrolled high levels of suction are con
tently placed above the level of the patient's chest, sidered dangerous, it is necessary to have some means
fluid can flow back into the pleural cavity. of controlling the amount of suction.
Controlled amounts of suction can be readily
applied to the system if a third bottle, the suction
Two- Bottle Co l lecti o n Syste m
control bottle, is added to the system, as illustrated
This system (Fig. 29.7) is preferred to the one-bottle in Figure 2 9 . 8 . A vent on the suction-control bottle
collection system when substantial amounts of liquid is connected to a vent on the water-seal bottle. The
are draining from the pleural space. With this system, suction-control bottle has a rigid straw similar to that
the bottle adjacent to the patient acts as a collection of the water-seal bottle. The suction is connected
bottle for the drainage, and the second bottle pro to a second vent on the suction-control bottle. When
vides the water seal and the air vent. Therefore, the suction is applied to the suction-control bottle, air
degree of water seal does not increase as the drainage enters this bottle through its rigid straw if the pres
accumulates. The water-seal bottle functions identi sure in the bottle is more negative than the depth
cally in both the one- and two-bottle systems. to which the straw is submerged. Therefore, the
amount of negative pressure in the system is equal
to the depth to which the rigid straw in the suction
S u ction a n d Th ree-Bottle
control bottle is submerged below the surface as long
Col l ecti o n Syste m s
as bubbles are entering the suction-control bottle
A t times, i t i s desirable t o apply negative pressure through its rigid straw. In the example in Figure 29.8,
to the pleural space to facilitate reexpansion of the air enters the suction-control bottle from the atmo
underlying lung or to expedite the removal of air or sphere while its rigid straw is submerged at 20 cm
fluid from the pleural space. Suction at a fixed level, H20 . Therefore, the pressure in the suction-control
usually 1 5 to - 20 cm H20, can be applied to the
- bottle is - 20 cm H20. The same pressure exists in
vent on a one- or two-bottle collection system with the water-seal bottle because these two bottles are in
To suction To patient
FIGURE 29.8 • Th ree-bott l e co l l ecti o n syste m . The arro ws d escr i b e the pathway for a i r to l eave the
pleural space.
C H A PT E R 29 I C H EST TU B E S 473
direct communication. The pressure in the drainage applied through the suction (D) . Of course, if there is
collection bottle is less negative than that in the other no liquid in the tube, the actual applied pressure will
bottles, however, on account of the intervening water be the suction pressure minus the depth of the water
seal. In this case, the depth of the water seal is 2 cm, seal. In the clinical situation, the presence of fluid in a
so the pressure in the drainage collection bottle and dependent loop will result in an increased pressure at
the pleural space (if no liquid is present in the chest the connection between the chest tube and the drain
tube) is - 1 8 cm H 2 0 . age system and will result in a decrease in the hourly
The amount o f negative pressure i n the system drainage (23) .
can be changed by adjusting the position of the rigid
straw in the suction-control bottle or by changing
CO M M E R C I A L LY AVA I LA B L E
the depth of the water in the suction-control bottle.
D RA I NAG E SYSTE M S
Bubbles must come continuously from the bottom
of the suction-control straw if one is to obtain the As can be appreciated from Figure 2 9 . 8 , three-bottle
expected degree of negative pressure. The bubbling systems are unwieldy to set up and are cumber
does not need to be vigorous, j ust continuous; some to move if the patient needs to be transported.
vigorous bubbling only creates more noise and hastens However, a number of more compact and conven
evaporation of the solution in the control bottle. ient chest drainage units are commercially available.
The main disadvantage of these units is that they
are more expensive than the older systems. The average
I NTRI N S I C N E G ATIVE P R E S S U R E I N
cost is approximately US$7 5 . Several different com
C H E ST TU B E S
panies manufacture drainage systems including the
The presence ofliquid i n the chest tubes can markedly Pleur-Evac systems (Teleflex Medical, www. teleRex
influence the negative pressure that is applied to the medical.com) , the Atrium systems (Atrium Medical
pleural space, as illustrated in Figure 29.9. The liquid Corporation, www. atriummed.com, 603 880- 1 433) ,
in the tube that runs from the patient to the Boor pro and the Argyle systems (Sherwood Medical, Tullamore,
duces the effects of a siphon (2 1 ,22) . If the distance Ireland) .
from the patient's chest to the top of the collection There are several types of drainage systems avail
apparatus is 50 cm and the tube is filled with liquid, able. The suction control can be wet or dry, the water
there will be a negative pressure of 50 cm H 2 0 in seal can be wet or dry, the amount of maximal suction
the pleural space if no suction is applied. The actual varies, and the capacity of the collection chamber
negative pressure applied to the pleural space from varies. In addition, some systems have a graduated
the entire system is the net vertical distance that the leak monitor and some have an autotransfusion con
liquid occupies in the tube (A- B) minus the level of nection. The various models available are compared
fluid in the water seal (C) plus the negative pressure nicely in an article by Manzanet et al. (24) .
FIGURE 29.9 • The p resence of l i q u i d i n the t u b e ca n affect the a m o u n t of neg ative pressu re a p p l ied to the
pleural space. The act u a l negative p ress u re i n the chest = A - B - C + D.
474 PLE U RAL D I S EASES
An acceptable drainage system should have the The maximal amount of air that can be removed
following characteristics: (a) the water seal should varies from one drainage system to another (24,25 ) .
be easily visualized, so one can determine whether the Baumann e t al. (25) measured the air flow a t different
chest tube is patent and whether an air leak is present. levels of suction in eight commercially available pleu
Some systems have a one-way valve that does not con ral drainage units and reported that the maximal
tain water, but one can (and should, if dealing with a amount of air flow varied from 1 0. 8 to 40.7 L/minute.
pneumothorax) fill the chamber with water to view However, the most important factor limiting air flow
the bubbling; (b) the system should be functional through chest tubes is probably the pleural drainage
when no suction is applied; (c) the volume of the col catheter. At 20 cm H 2 0 suction, the air flow through
lection chamber should be adequate and the markings a 8 . 0 F catheter varies from 2.6 to 6 . 5 L/minute
should be such that the drainage is easily quantitated; whereas the air flow through a 1 4 .0-F catheter varies
(d) there should be a pop-off valve to provide a safety from 1 2.3 to 1 6. 8 L/minute (25 ) .
factor if pressure builds up in the system. Neither of
the single-chamber systems mentioned in the article
P l e u r- Evac U n it
by Manzanet et al. (24) have this capability and are
not recommended. If the patient has a large amount A typical drainage system (The Pleur-Evac system,
of blood in the pleural space, consideration should be Teleflex Medical, www.teleflexmedical.com) is depicted
given to using a unit with autotransfusion capabilities. in Figure 29. 1 0 . The drainage system is a disposable,
Suction
To
patient
( 1 600 700
600
b
20
2300 1 400 500
6
15 400
6
2 1 00 1 200 300
10
200
-
1 900 1 000 1 50
5
1 00
1 700 50
0 800
B u b b l i n g th ro u g h Wate r-Sea l C h a m ber it will not have participated in gas exchange in the
lung, and the Pco2 should be less than 1 0 mm Hg.
If air bubbles are escaping through the water seal,
In such patients, additional sutures should be placed
it means that air is entering the chest tube between
to make the chest tube insertion site airtight.
the pleural space and the water seal. If the patient
Bubbling through the water-seal chamber should
is receiving water-seal drainage without suction, the
not be confused with bubbling through the suction
presence of bubbling in the water seal usually indi
control chamber. If suction is working properly,
cates a persistent air leak from the lung into the
bubbling through the suction-control chamber or
pleural space. If no air bubbles are seen on the ini
the suction-control bottle will always be present.
tial inspection of the water seal, the patient should be
When suction is being applied, the suction-control
asked to cough, and the water seal should be observed
chamber should be checked to make certain that
for bubbling. The coughing maneuver increases the
there is a continuous stream of small bubbles and that
patient's pleural pressure and should demonstrate
the liquid is at the desired height.
small air leaks into the pleural space. Cerfolio's (27)
classification of air leaks is discussed in Chapter 24.
Is the Ch est Tu be F u n ction i n g ?
If the patient is receiving suction, disconnection
or partial disconnection anywhere between the water Each time a patient with a chest tube i s evaluated,
seal and the patient will lead to bubbling through the the functional status of the chest tube itself should
water seal. For example, if the cap on the collection be evaluated. If the patient is not receiving suction,
bottle in Figure 29.8 is not airtight, air will be pulled one should observe the level of the liquid in the
into the collection bottle by the negative pressure and water seal. If the chest tube is patent and in the pleu
will exit through the water-seal bottle producing bub ral space, the level of the liquid should move higher
bling. Leaks in the system may be detected by clamp on inspiration in the limb of the water seal proximal
ing the chest tube at the point where it exits from to the patient, indicating a more negative pleural
the chest. If bubbling through the water seal persists, pressure. Of course, if the patient is receiving
the drainage system itself is responsible for the leak, mechanical ventilation, the level of liquid in the
and it should be examined thoroughly for leaks. If the proximal limb will go down on inspiration because
bubbling stops when the chest tube is clamped, then the pleural pressure becomes more positive. When no
the air is coming from the pleural space. fluctuations are observed synchronous with respira
The presence of bubbling through the water tory movements, the patient should be asked to make
seal does not necessarily indicate a communication a maximal inspiratory effort, and if still no move
between the lung and the pleural space. If the chest ment is observed, it indicates that the chest tube is
tube is not inserted far enough into the pleural space, not functioning.
one or more of the holes in the chest tube may lie When the patient is receiving suction, it may be
outside the pleural space. Obviously, in such a situa more difficult to ascertain chest tube function. If
tion, air enters the chest tube directly from the atmo large bubbles of air are entering the suction-control
sphere if negative pressure is applied to the chest tube. chamber, the level of liquid in the water-seal cham
The possibility is evaluated by inspecting the chest ber will fluctuate, depending on the number and size
tube. At times, particularly in debilitated patients with of bubbles in the chamber. These fluctuations should
poor tissue turgor, the negative pleural pressure will not be mistaken for evidence that the tube is func
cause air to enter the pleural space around the chest tional. When the patient is receiving suction, the neg
tube at the insertion site. At times, it may be difficult ative pressure in the suction-control chamber should
to tell whether the air is leaking around the chest tube be transmitted to the pleural space continuously to
or whether it is due to a bronchopleural fistula. One keep the pleural pressure constant.
may make this differentiation by measuring the level Therefore, to detect changes in pleural pressures
of Pco2 in the air coming from the chest tube. The air with respiration, the suction must be temporarily dis
from the chest tube is collected in a syringe and ana continued. When suction is discontinued, the volume
lyzed with the regular blood gas analyzer. Usually, a of air and liquid between the water seal and the pleu
simple modification must be made on the analyzer so ral space should not change, and therefore, the level
that it can analyze gas rather than liquid. If the air of liquid in the water-seal chamber should rise to be
came from the lung through a bronchopleural fistula, equivalent to the suction previously applied. This ini
then the Pco2 should be greater than 20 mm Hg. tial rise occurs whether or not the chest tube is patent.
Alternatively, if the air leaked around the chest tube, After this initial rise, the level of the fluid in the water
C H A PT E R 29 I C H EST TU B E S 477
seal should be observed for fluctuations with the based on the quantity of the drainage. The character
respiratory cycle to verify patency of the chest tube. of the drainage is best described by quantitating the
If a chest tube is not functioning, its functional percentage of solid drainage material. This quantita
status should be restored, or it should be removed. tion is easily done by marking the level of the sediment
Chest tubes can become obstructed with tissue in the collection chamber each day. If the increase in
around the holes or by clots within the tube. The sim volume of the entire collection system is known and
plest method for restoring patency is to flush the tube if the increase in volume of the solid sediment is
with 50 mL of saline. This frequently clears the tube known, it is simple to calculate what percentage of
by pushing the clot out of the tube or by pushing the daily drainage is solid.
the tissue away from the holes in the tube. The patency
of a chest tube obstructed by clots in the extratho CO M P L I CAT I O N S O F TU B E
racic portion of the tube can frequently be restored T H O RACO STO M Y
by "stripping" the tube. The usual technique is to grip
There are numerous complications o f tube thoracos
and stabilize the tubing adjacent to the chest with the
tomy and some are serious or even fatal. Lamont et al.
thumb and index finger of one hand and then to slide
(30) reviewed safety reports from the National Patient
the other hand toward the drainage unit to compress
Safety Agency in England and Wales between January
a section of the tubing. Then, the first hand is repo
2005 and March 2008 and reported that there were
sitioned adjacent to the second hand, and the proce
1 2 deaths and 1 5 cases of severe harm from chest tube
dure is repeated until the entire length of the tubing is
insertion. They felt that the true rates were likely to
cleared. A special chest tube roller is sometimes used.
be much higher (30) . They attributed the complica
Stripping may relieve the obstruction in the tube.
tions to (a) poor selection of site for drain insertion
An alternate strategy is to instill a fibrinolytic into
and the lack of use of ultrasonography; (b) inad
the chest tube. Although there are no studies to my
equate supervision of trainee doctors; (c) equipment
knowledge assessing the efficacy of fibrinolytics for
problems including lack of familiarity with different
obstructed chest tubes, the utility of alteplase (tissue
kits and excessive length of available dilators; and
plasminogen inhibitor) in clearing obstructed perito
(d) lack of awareness of national clinical guidelines
neal catheters has been documented. Zorzanello et al.
(30) . Harris et al. (3 1 ) subsequently sent question
(28) reported that alteplase was effective in relieving
naires regarding harm from chest tube insertion
the obstruction in 24 of 29 instances of catheter
between 2003 and 2008 to 1 4 8 acute hospitals in the
obstruction. These researchers used 8 mg of alteplase
UK from which they received 1 0 1 responses. Thirty
rather than the 2 mg that are commonly used for
one cases of chest tube misplacement were reported
obstructed intravascular catheters (28) .
with seven deaths. The chest tubes were placed in
Chest tubes that are no longer patent and are no
the liver ( 1 0) , the peritoneal space (6) , the heart
longer draining fluid should be removed because they
(5), the spleen (5), subclavian vessels (2) , colon ( 1 ) ,
serve as conduits for bacterial infection of the pleural
esophagus ( 1 ) , and inferior vena cava ( 1 ) . Forty-seven
space. Chest tubes frequently become colonized with
cases of serious lung or chest wall inj uries with eight
bacteria. In one study (29) , bacterial cultures of the
deaths were reported. Six chest tubes were placed on
fluid from the chest tubes were positive in 83% of
the wrong side with two deaths (3 1 ) .
36 patients who had undergone lung resections.
I n order t o minimize the risks o f inserting chest
When a chest tube becomes obstructed with a clot,
tubes, Harris et al. (3 1 ) recommend the following
the clot may contain bacteria which can lead to pleu
with which I agree. (a) Chest tubes are only inserted
ral infection. At times, a chest tube does not fluctuate
by staff with relevant competencies and adequate
but continues to drain fluid. In such a situation, the
supervision. (b) Ultrasound guidance is recommended
chest tube need not be removed.
when inserting a chest tube for fluid. (c) Clinical
guidelines are followed and staff are made aware of
A m o u n t a n d Type of D ra i nage
risks. (d) Written informed consent is obtained when
The amount and the character of the drainage from a chest tube is inserted.
the chest tube should be recorded for each 24-hour One of the most common complications is mis
period. The amount of drainage is most easily quan placement of the chest tube. The incidence of this
titated by marking the level of the liquid in the col complication has varied markedly. In one study, the
lection chamber each day. This record-keeping is incidence of malpositioned chest tubes was 26% with
important because many therapeutic decisions are the emergency insertion of 77 tubes (32) . In contrast,
478 PLE U RAL D I S EASES
the incidence of chest tube malposition was only 1 % emphysema, which usually presents as soft tissue
in one series of 447 patients in whom the chest tubes crepitus around the drain site but may rapidly spread
were inserted through blunt dissection (33) . A PA and to virtually any place in the body. The presence of
lateral chest radiograph should always be obtained subcutaneous emphysema in patients with tube tho
after a chest tube is inserted. It should be noted, how racostomies indicates one of three possibilities (39) :
ever, that frequently the malposition is not diagnosed (a) a side-hole on the chest tube is lying outside the
by these routine films. Accordingly, if a patient has an pleural space within the chest wall, allowing air to
air or Buid collection that is not being drained ade enter the tissue planes, (b) the chest tube is blocked,
quately, a chest CT scan should be obtained to assess or (c) the drainage system cannot cope with the air
the position of the chest tube (and the presence of leak. The latter situation is unusual and may be
loculi) . In one series, malpositioned chest tubes were related to a chest tube that is too small or a massive
diagnosed by supine chest radiographs in only 7 of air leak. Cerfolio et al. (40) reported that 6.3% of
20 patients (35%) whereas the CT scan was necessary 4,023 patients undergoing lung resection for cancer
for diagnosis in the remaining 1 3 patients (32) . developed subcutaneous emphysema. These workers
Another complication of chest tubes is that they (40) reported that they performed thoracoscopy on
fall out. In one study (34) from Oxford in which 1 00 64 of the patients and found that the lung was
1 2-F chest tubes were inserted by the Seldinger tech adhered to the part of the chest that had previously
nique, 2 1 % of the chest tubes fell out (34) . The drain been opened. When the lung was removed from the
became blocked in an additional nine patients (34) . opening and another chest tube was place, the subcu
There researchers recommended regular chest tube taneous emphysema resolved within 24 hours in all
Bushes to prevent blockage of the chest tube since but one patient (40) .
only 1 of 58 chest tubes became obstructed when The insertion of a chest tube creates inflammation
the tubes were Bushed. in the pleural space. Carvalho et al. (4 1 ) studied the
Pleural infection is another complication of tube pleural Buid characteristics of sheep with an experi
thoracostomy. The administration of antibiotics to mental pleural effusion and an Argyle 32-Fr tube in
patients who have chest tubes for thoracic trauma the pleural space. The white blood cell count in the
may decrease the prevalence of empyema. Brunner pleural effusion increased from 1 2 5 to more than
et al. (3 5) randomly allocated 90 such patients to 6,000/mm3 within 6 hours. In this model, the pleural
receive cefazolin or nothing immediately before and Buid protein level increased from 0 to 3 .7 g/dL within
then every 6 hours until tube removal. They reported 48 hours, and the pleural Buid lactate dehydrogenase
that there were six empyemas and three cases of level increased from 44 to 638 IU/L within 24 hours.
pneumonia in the control group but only one case of There is one report of three patients (all with quad
pneumonia and no empyema in the antibiotic group riplegia) who developed lung entrapment as a result
(3 5 ) . Two subsequent studies, which also evaluated of prolonged chest tube drainage. They could only
trauma patients, reported similar results (36,37) . be weaned off the ventilator after a decortication was
A recently published meta-analysis (38) that included performed (42) .
nearly 2,500 patients concluded that prophylactic
antibiotics in patients receiving chest tubes for blunt A U TOTRAN S F U S I O N
or penetrating thoracic injuries had significantly less
Autotransfusion involves the collection, filtration,
infections and empyemas than did patients who did
and reinfusion of the patient's shed blood for reple
not receive the antibiotics. In view of these studies,
tion of intravascular volume and the diminution of
prophylactic antibiotics are recommended for all
transfusion requirements. Postoperative patients with
trauma patients who receive a chest tube. The antibi
chest tubes and patients with hemothorax should be
otic chosen should have activity against Staphylococcus
considered as candidates for autotransfusion. Several
aureus because this is the organism that causes the most
autotransfusion systems are commercially available.
infections (37) . The utility of prophylactic antibiotics
Atrium Medical Corporation manufactures a system
in other situations such as in postoperative patients,
capable of continuous infusion.
patients with spontaneous pneumothorax, and those
with malignant effusions undergoing pleurodesis, is
C H E ST TU B E R E M OVALS
yet to be evaluated.
Another occasional complication of tube tho The indications for the discontinuation of the tube
racostomy is the development of subcutaneous thoracostomy for various conditions are discussed in
C H A PT E R 29 I C H EST TU B E S 479
the respective chapters on these conditions. In general, 8 . Moulton JS. Image-guided management of complicated pleu
ral fluid collections. Radio/ Clin North Am. 2000;38:345-374.
chest tubes for pneumothorax are removed when the
9 . Cantin L, Chartrand-Lefebvre C, Lepanto L, et al. Chest tube
lung has reexpanded and no air leaks are present. drainage under radiological guidance for pleural effusion and
Chest tubes for hemothorax and empyema are pneumothorax in a tertiary care university teaching hospital:
removed when pleural drainage of blood or pus, review of 5 1 cases. Can Respir]. 2005; 1 2 :29-33.
respectively, has become minimal. 1 0 . Thal AP, Quick KL. A guided chest tube for safe thoracostomy.
Surg Gynecol Obstet. 1 9 88; 1 67: 5 1 7.
Before chest tube removal is attempted, the pro
1 1 . Dev SP, Nascimiento B Jr, Simone C, et al. Videos in clinical
cedure should be explained to the patient. In addi medicine. Chest-tube insertion. N Engl] Med. 2007;3 57:e l 5 .
tion, petrolatum-impregnated gauze and an occlusive 1 2 . Symbas PN. Chest drainage tubes. Surg Clin North Am.
bandage should be prepared for use in a sterile field. 1 9 89;69:4 1 -46.
Then, the dressing covering the thoracostomy site is 13. Luketich JD, Kiss M, Hershey J, et al. Chest tube insertion:
a prospective evaluation of pain management. Clin J Pain.
removed, and the suture restraining the chest tube is
1 998; 1 4 : 1 52-1 54.
cut. Petrolatum-impregnated gauze is placed around 1 4 . Zgoda MA, Lunn W, Ashiku S, et al. Direct visual guidance
the tube on the patient's chest wall so that it can be for chest tube placement through a single-port thoracoscopy:
moved to cover the wound when the tube is removed. a novel technique. Chest. 2005 ; 1 27 : 1 805-1 807.
It is important that the pleural pressure be positive 1 5 . Gilbert TB, McGrath BJ, Soberman M . Chest tubes: indica
tions, placement, management and complications. J Intensive
when the chest tube is removed so that air will not
Care Med. 1 993;8:73-86.
enter the pleural space. This can be accomplished 1 6 . Gayer G, Rozenman ], Hoffmann C, et al. CT diagnosis of
by having the patient hum or perform a Valsalva malpositioned chest tubes. BrJ Radio/. 2000;73 :786-790.
maneuver. The chest tube is quickly pulled out of 1 7 . Curtin JJ, Goodman LR, Quebbeman EJ, et al. Thoracostomy
tubes after acute chest injury: relationship between location
the chest, and the wound is covered immediately
in a pleural fissure and function. AJR Am J Roentgenol. 1 994;
with the gauze. The wound usually closes sufficiently 1 63 : 1 339- 1 342.
without using sutures. The procedure is completed 1 8 . Cerfolio RJ, Bass CS, Pask AH, et al. Predictors and treatment
by placing an occlusive dressing over the gauze. of persistent air leaks. Ann Thorac Surg. 2002;73 : 1 727- 1 730.
An ultrasound (43) or a chest x-ray should be 19. Waller DA, Edwards JG, Rajesh PB. A physiological compari
son of flutter valve drainage bags and underwater seal systems
obtained after removal of the chest tube to assess
for postoperative air leaks. Thorax. 1 999;54 :442-443.
for residual air. Ultrasound is probably more cost 20. Mainini SE, Johnson FE. Tension pneumothorax complicating
effective (43) . Recurrent pneumothoraces following small-caliber chest tube insertion. Chest. 1 990;97:759-760.
chest tube removal result more frequently from air 2 1 . Enerson OM, Mcintyre J. A comparative study of the physiol
entering the chest from the atmosphere during removal ogy and physics of pleural drainage systems. J Thorac Cardio
vasc Surg. l 966;52:40-46 .
than from a leak in the lung parenchyma (39) .
22. Kam AC, O'Brien M, Kam PC. Pleural drainage systems.
Anaesthesia. 1 993;48 : 1 54- 1 6 1 .
23. Schmelz JO, Johnson 0 , Norton JM, e t al. Effects o f position
REFERENCES of chest drainage tube on volume drained and pressure. Am J
Crit Care. l 999;8:3 1 9-323.
I. Munnell ER. Thoracic drainage. Ann Thorac Surg. 1 997;63: 24. Manzanet G, Vela A, Corell R, et al. A hydrodynamic study of
1 497- 1 502. pleural drainage systems: some practical consequences. Chest.
2 . Light RW. Pleural controversy: optimal chest tube size for 2005 ; 1 27:22 1 1 -222 1 .
drainage. Respirology. 20 1 1 ; 1 6:244-248. 2 5 . Baumann MH, Patel PB, Roney CW, et al. Comparison of
3 . Fysh ET, Smith NA, Lee YC. Oprimal chest drain size: the rise function of commercially available pleural drainage units and
of the small-bore pleural catheter. Semin Respir Crit Care Med. catheters. Chest. 2003; 1 2 3 : 1 878- 1 886.
20 1 0; 3 1 :760-768 . 26. Bar-El Y, Ross A, Kablawi A, et al. Potentially dangerous
4. Havelock T, Teoh R, Laws D, et al. Pleural procedures and negative intrapleural pressures generated by ordinary pleural
thoracic ultrasound: British Thoracic Society Pleural Disease drainage systems. Chest. 200 1 ; 1 1 9 : 5 1 1 -5 1 4 .
Guideline 20 1 0 . Thorax. 2 0 1 0;6 5 (suppl. 2) :ii6 1-ii76. 2 7 . Cerfolio RJ. Recent advances i n the treatment o f air leaks.
5. Davies HE, Davies RJ, Davies CW, et al. Management of Curr Opin Pulm Med. 2005 ; 1 1 :3 1 9-323.
pleural infection in adults: British Thoracic Society Pleural 28. Zorzanello MM, Fleming WJ, Prowant BE. Use of tissue plas
Disease Guideline 20 1 0. Thorax. 2 0 1 0;65 (suppl. 2) :ii4 1 -ii53. minogen activator in peritoneal dialysis catheters: a literature
6. Cafarotti S, Dall'armi V, Cusumano G, et al. Small-bore wire review and one center's experience. Nephrol Nurs ]. 2004;
guided chest drains: safety, tolerability and effectiveness in 3 1 : 534-5 37.
pneumothorax, malignant effusions, and pleural empyema. 29. Korona-Glowniak I, Los R, Rybojad P, et al. Bacterial coloniza
J Thorac Cardiovasc Surg. 20 1 1 ; 1 4 1 :683-687. tion of pleural drains in patients with lung cancer: An analysis
7. Ernst A, Silvestri GA, Johnstone 0 , et al. American College of risk factors. Med Sci Monit. 20 1 0; 1 6: CR84-CR9 1 .
of Chest Physicians. Interventional pulmonary procedures: 30. Lamont T, Surkitt-Parr M , Scarpello J , et al. Insertion of chest
guidelines from the American College of Chest Physicians. drains: summary of a safety report from the National Patient
Chest. 2003; 1 23: 1 693- 1 7 1 7. Safety Agency. BM]. 2009;339 :b4923.
480 PLE U RAL D I S EASES
3 1 . Harris A, O'Driscoll BR, Turkington PM. Survey of maj or 38. Bosman A, de Jong MB, Debeij J, et al. Systematic review and
complications of intercostal chest drain insertion in the UK. meta-analysis of antibiotic prophylaxis to prevent infections
Postgrad Med]. 2 0 1 0;86:68-72. from chest drains in blunt and penetrating thoracic injuries.
32. Baldt MM, Bankier AA, Germann PS, et al. Complications Brit] Surg. 2 0 1 2;99:506-5 1 3
after emergency tube thoracostomy:assessment with CT. 3 9 . Tang AT, Velissaris TJ, Weeden D F. An evidence-based
Radiology. 1 99 5 ; 1 9 5 : 539-543. approach to drainage of the pleural cavity: evaluation of best
33. Millikan JS, Moore EE, Steiner E, et al. Complications of tube practice. J Eva! Clin Pract. 2002;8:333-340.
thoracostomy for acute trauma. Am] Surg. 1 980; 1 40:73 8-74 1 . 40. Cerfolio RJ, Bryant AS, Maniscalco LM. Management of
34. Davies HE, Merchant S, McGown A. A study of the com subcutaneous emphysema after pulmonary resection. Ann
plications of small bore "Seldinger" intercostal chest drains. 1horac Surg. 2008 ; 8 5 : 1 759-1 763.
Respirology. 2008 ; 1 3 :603-607. 4 1 . Carvalho P, Kirk W, Butler J, et al. Effects of tube thoracos
3 5 . Brunner RG, Vinsant GO, Alexander RH, et al. The role of tomy on pleural fluid characteristics in sheep. J Appl Physiol.
antibiotic therapy in the prevention of empyema in patients 1 993;74:2782-2787.
with an isolated chest injury (ISS 9- 1 0) : a prospective study. 42. Peterson WP, Whiteneck GG, Gerhart KA. Chest tubes,
] Trauma. 1 990;30: 1 1 48-1 1 5 3 . lung entrapment, and failure to wean from the ventilator.
36. Nichols RL, Smith J W, Muzik AC, e t al. Preventive antibi Report of three patients with quadriplegia. Chest. 1 994; 1 05 :
otic usage in traumatic thoracic injuries requiring closed tube 1 292- 1 294.
thoracostomy. Chest. 1 9 94; 1 06: 1 493- 1 4 9 8 . 43. Saucier S, Motyka C, Killu K. Ultrasonography versus chest
3 7. Gonzalez RP, Holevar M R . Role of prophylactic antibiotics for radiography after chest tube removal for the detection of
tube thoracostomy in chest trauma. Am Surg. 1 998;64: 6 1 7-620. pneumothorax. MCNAdv Grit Care. 20 1 0; 2 1 :34-3 8 .
T h o ra cosco py
Although thoracoscopy has been a part o f thoracic can be performed in an endoscopy suite, it does not
surgical practice for many years, the advent of video involve general anesthesia, and is less expensive than
assisted techniques has greatly expanded the indi VATS (2) . For diagnostic purposes, either VATS or
cations and the uses of this procedure. In the past, medical thoracoscopy is appropriate and the choice
thoracoscopy was performed mainly for diagnostic of procedure depends primarily on its availability at
purposes. Presently, video-assisted thoracic surgery one's institution. The number of centers performing
(VATS) has assumed a major role in the therapy of medical thoracoscopy in the United States and the
chest pathology. Indeed, in some institutions it is now United Kingdom is gradually increasing. For exam
the most commonly used operative approach in the ple, between 1 999 and 2004 the number of centers
general thoracic surgical practices ( 1 ) . The primary in the United Kingdom where medical thoracoscopy
advantage of VATS is that it produces less morbidity was performed increased from 1 1 to 1 7 (3) .
and mortality, and shorter duration of hospitaliza
tion than does thoracotomy. Presently, VATS is used
H I STO RY
for many surgical procedures in the chest other than
those related to pleural disease including pulmonary Thoracoscopy was developed by Jacobaeus in the
nodule removal, lobectomy, lung biopsy, exploration early 1 900s because a method was needed to break
of the mediastinum, myotomy for achalasia, sympa down adhesions in patients with pulmonary tuber
thectomy, esophagectomy, and pericardia! window culosis so that an artificial pneumothorax could be
creation. In this chapter, we discuss only those proce produced (4) . Thoracoscopy was used extensively for
dures that deal with pleural disease. this purpose until 1 94 5 , at which time streptomycin
It is important to understand that there are two was introduced for the treatment of tuberculosis (5) .
different techniques for thoracoscopy-VATS and In one report, the results in 1 ,000 patients in whom
medical thoracoscopy. VATS is performed in an oper thoracoscopy was used to break down adhesions were
ating room under general anesthesia with the patient detailed (6) . Jacobaeus also published an early report
selectively intubated to allow for single lung ventila on the use of thoracoscopy to localize and diagnose
tion. Multiple puncture sites are made in the chest benign and malignant lesions of the pleura and pul
wall through which the thoracoscope and surgical monary parenchyma (7) .
instruments are introduced. Medical thoracoscopy After 1 9 50, thoracoscopy was rarely performed in
differs from VATS in that the patient is usually not the United States, although some physicians in Europe
intubated and breathes spontaneously (2) . The pro continued to perform the procedure (8) . During this
cedure is usually performed with conscious sedation period, thoracoscopy was used primarily to assist in
and local anesthesia. Medical thoracoscopy primarily diagnosing pleural effusions, although pleurodesis was
serves as a diagnostic tool rather than for intervention. sometimes attempted with talc (9, 1 0) or silver nitrate
It is usually performed by pulmonologists, whereas ( 1 1 , 1 2) . A number of instruments were employed
VATS is performed by thoracic surgeons (2) . The pri including rigid bronchoscopes, mediastinoscopes, flexi
mary advantages of medical thoracoscopy are that it ble bronchoscopes, and specialized rigid fiber-optic
48 1
482 PLE U RAL D I S EASES
thoracoscopes ( 1 3) . Two older books ( 1 4, 1 5) provide For cauterization o f adhesions and blebs, o r i n case
state-of-the-art discussions on thoracoscopy before the of bleeding after biopsy, electrocoagulation or laser
advent of VATS, which has become available mainly coagulation should be available.
since 1 990. Medical thoracoscopy is usually performed under
The recent revival of thoracoscopy was made pos local anesthesia, but some premedication (e.g., mid
sible by the tremendous advances in endoscopic tech azolam) should be routinely administered. Thora
nology ( 1 6) . The development of the charged coupling coscopy is usually performed with the patient in the
device and a silicon chip that is light sensitive led to lateral decubitus position with the hemithorax to be
the sufficient miniaturization of a video camera. When studied facing upward.
attached to a fiber-optic telescope, the video camera The site for the introduction of the thoracoscope
produces a well-defined, magnified image on a video depends upon the location of radiographically detected
monitor that allows the operating surgeon to work abnormalities, while avoiding potentially hazardous
with an assistant. Previously, the surgeon had to hold areas such as that of the internal mammary artery,
the thoracoscope, and only he could look into it while the axillary region with the lateral thoracic artery, and
working, which did not allow for the aid of an assistant the infraclavicular region with the subclavian artery.
and, therefore, limited the complexity of the proce It is important not to insert the thoracoscope too low
dures undertaken ( 1 3) . because the diaphragm or spleen may be accidentally
injured ( 1 8) . The usual site for insertion of the tho
racoscope is in the sixth or seventh intercostal space
PROCE D U RE between the mid- and anterior-axillary lines.
Med i ca l Thoracoscopy Before the thoracoscope is introduced, a pneumo
thorax of several hundred cubic centimeters of air is
Training Requiremen ts for Medical
usually induced ( 1 8) . Thoracic ultrasound prior to
Thoracoscopy
thoracoscopy improves pleural access and predicts
The guidelines from the American College of Chest fibrous septation ( 1 9) . Examination of the pleural
Physicians (ACCP) recommend that trainees should space is only possible if the space between the lung
perform at least 20 procedures in a supervised setting and chest wall is sufficiently large to move the instru
to establish basic competency. Then to maintain com ments around easily and to visualize all important
petency, dedicated operators should perform at least areas of the thoracic cavity ( 1 5 ) . After the trocar is
1 0 procedures per year ( 1 7) . introduced, the pleural fluid should be removed as
There are two fundamental techniques by which completely as possible. Then the entire pleural cavity
medical thoracoscopy is performed, namely, single is inspected. Biopsies are obtained from suspicious
puncture and double puncture ( 1 8) . Both techniques areas. Biopsies from the visceral pleural are usually
require a xenon light source, which satisfies the require obtained only when the parietal pleura appears nor
ments for high-quality visual exploration and video mal but the lung surface shows abnormal lesions.
documentation. For the single-puncture technique, Following thoracoscopy, a chest tube should be intro
a rigid thoracoscope with a 9-mm working channel duced and connected to suction.
is used. With the single-puncture technique, various In the last few years, new instrumentation has been
instruments such as the biopsy forceps, needle biopsy, developed which may facilitate medical thoracoscopy.
and suction catheter are used through the working Tassi and Marchetti (20) reported their experience
channel, which also accommodates electrocautery. with a technique called minithoracoscopy in which two
For the double-puncture technique, a smaller 7-mm 3 . 8-mm trocars, one with a 3 .3-mm telescope and the
rigid thoracoscope is used along with a second smaller other with a 3 .0-mm biopsy forceps were used for
5-mm trocar that accommodates biopsy forceps, undiagnosed pleural effusion. They reported that the
brushes, needles, and laser fibers. The single-puncture diagnostic yield with this procedure was 90% and
technique is the easiest method to learn and is com concluded that it is most useful for assessment of
monly used by the chest physician ( 1 8) . small effusions not accessible to conventional medical
Medical thoracoscopy can b e done either under thoracoscopy (20) and for patients with narrow inter
direct visual control through the endoscopic optic costal spaces (2 1 ) .
or indirectly by video transmission that allows a Ernst e t al. (22) described their experience with
magnified view and demonstration to assistants and a newly developed semirigid pleuroscope in the
others, as well as appropriate documentation ( 1 8) . evaluation of the pleural space. This pleuroscope was
C H APTE R 30 / T H O RACOSCO PY 483
developed so that it interfaces with existing processors of staples that are both hemostatic and aerostatic
and light sources that are routinely employed for flexi (Endo-GIA, U.S. Surgical Corp., Norwalk, CT, USA) .
ble bronchoscopy and are available in most endoscopy The operative time for a patient with an undiag
units. An additional benefit is that this pleuroscope nosed pleural effusion is short. Cerfolio et al. (24)
operates much like a bronchoscope, and therefore it reported that when a single-lumen tube was used for
should be easier to use by the pulmonologist who is undiagnosed pleural effusion, the mean operative time
not trained in the use of rigid instruments (22,23) . for pleural evaluation and pleural biopsy in a series of
more than 200 patients was only 17 minutes.
Previous thoracoscopy has been considered a rela
Vid eo-Assi sted Thoracic S u rg e ry
tive contraindication to a second thoracoscopy. How
Most VATS procedures are performed under general ever, Breen et al. (26) reported that they performed
anesthesia because with general anesthesia, endoscopic redo thoracoscopies in 29 patients and were able to
surgical manipulation can be accomplished safely and induce a pneumothorax and complete the procedure
expeditiously ( 1 ) . It is imperative that the anesthesia in all patients although pleural adhesions were more
personnel be experienced in open thoracic procedures. common with the redo procedures. Most of their
In addition, they must be well versed with the prin patients had malignant pleural effusions.
ciples of selective one-lung ventilation. Although most
surgeons perform VATS with a double-lumen endo
tracheal tube, it can be performed with a single-lumen CO NTRAI N D I CATI O N S TO
endotracheal tube if only a pleural effusion is to be T H O RACO S C O PY
drained and a biopsy obtained of the parietal pleura The two primary contraindications to thoracoscopy
(24) . Ventilation for the patient is provided through are the inability to tolerate one-lung ventilation and
the contralateral lung. It is most convenient to work pleural adhesions of sufficient density to preclude
with two video monitors, one on each side, so that entry into the chest ( 1 3 , 1 8) . Of course, the patient
both the operator and the assistant may have an unob must be able to tolerate general anesthesia and must
structed view. VATS has also been performed with not have bleeding abnormalities that would preclude
local anesthesia and sedation (25 ) . other surgical procedures.
The patient i s placed on the operating table, and
the chest is prepared and draped as for a thoracotomy.
After general anesthesia is induced, the thoracoscope I N D I CATI O N S A N D R E S U LTS
is inserted and the ipsilateral lung is collapsed for
U n d i a g n osed Pleural Effu s i o n
unimpaired visibility of the intrathoracic structures.
At this time, the thoracic cavity is systematically On occasion, the etiology of a pleural effusion remains
examined. After the initial thoracoscopic exploration uncertain after the initial diagnostic workup, which
of the pleural cavity is concluded, further intercostal includes a diagnostic thoracentesis with pleural fluid
access for VATS instrumentation is achieved under cytology, a pleural fluid marker for tuberculosis, and
direct thoracoscopic vision. Usually, three incisions are an evaluation for pulmonary embolus. Such patients
made to create a triangular configuration, an arrange are possible candidates for thoracoscopy to establish
ment that facilitates instrument placement and allows the etiology of the pleural effusion. It should be noted,
one to work in coordination with an assistant. The however, that the only two diagnoses that are usually
incisions are placed along a line appropriate for a established with thoracoscopy are malignancy and
thoracotomy incision so that if a subsequent thora tuberculosis. If the patient has a pleural effusion due
cotomy is required, the incisions are simply joined. At to a different etiology, the diagnosis in all probability
the completion of the VATS procedure, a single chest will not be established at thoracoscopy.
tube is placed in the pleural space. Thoracoscopy is an efficient way to establish the
Instrumentation for VATS is slowly improving. diagnosis of malignancy. In the early 1 990s, two sepa
Initially, instruments designed for laparoscopy were rate studies, each with 1 02 patients, were published
used but were less than ideal, particularly for grasping that reported diagnostic yields of 93% (27) and 80%
lung parenchyma, which has a tendency to tear. The (28) . However, when these two studies are examined
most significant advance in instrumentation was the in detail, one finds that the only diagnosis that was
development of an endoscopic linear stapler, which definitely established is malignancy. When the two
simultaneously cuts while laying down parallel rows studies mentioned in the preceding text are combined,
484 PLE U RAL D I S EASES
the diagnosis of malignancy was established in 99 of Our preferred method is pleural abrasion with an alter
the 1 1 7 patients (85%) with malignancy, including native being the instillation of 1 00 mL of 2% iodopo
5 1 of 56 (9 1 %) with mesothelioma. In a more recent vidone (36) . The efficacy of mechanical abrasion was
study from Denmark, medical thoracoscopy estab documented in one study in which 87 patients with
lished the diagnosis in 89 of 1 0 1 patients (88%) with malignant effusions secondary to breast carcinoma
malignancy (29) . Preliminary results in one study were randomized to receive pleurodesis by mechani
suggest that the use of autofluorescence videothora cal abrasion in conjunction with thoracoscopy or by
coscopy may help identify areas of the pleura with talc slurry (37) . Pleurodesis with mechanical abrasion
malignant involvement (30) . Thoracoscopy can also had a higher success rate (89%) than pleurodesis with
establish the diagnosis of tuberculosis (3 1 -3 3 ) . In one talc slurry (74%) (37) . Although talc insuffiation was
study from South Africa, the diagnosis of tuberculosis recommended in the earlier editions of this book,
was established with thoracoscopy in all 42 patients concerns about respiratory failure occurring after talc
with tuberculous pleuritis (33). administration (see Chapter 1 O) have led to this differ
Where is the rightful place of thoracoscopy in ent recommendation. If talc is used in this situation,
the management of the patient with an undiagnosed only graded (large particle size) talc should be used. It
pleural effusion? Thoracoscopic procedures should be should also be noted that the performance of thoracos
used only when the less invasive methods of diagnosis copy without any attempt to create a pleurodesis will
such as pleural aspiration for cytologic, bacteriologic, result in a pleurodesis in more than 50% of patients
and chemical examinations have not yielded a diagno with malignant pleural effusions (38,39) .
sis. In one series of 620 patients with pleural effusions, How sensitive is thoracoscopy i n making the diag
only 48 (8%) remained without a diagnosis and were nosis of malignancy? Davies et al. (40) performed medi
subjected to thoracoscopy (34) . In these 48 patients, a cal thoracoscopy on 1 42 patients with undiagnosed
diagnosis of malignancy was established in 24 (50%), pleural effusion and the diagnosis of malignancy was
and in an additional 16 patients, the diagnosis of established in 89. Forty-four of the patients had non
benign disease was established when the thoraco specific pleuritis at thoracoscopy and 5 of these ( 1 2%)
scopic and clinical findings were considered jointly. were subsequently diagnosed with mesothelioma (40) .
In the remaining eight patients ( 1 6%) , no diagnosis Therefore, the sensitivity of medical thoracoscopy in
was established at thoracoscopy, but six of them were making the diagnosis of malignancy was 94% .
subsequently diagnosed as having malignancy (34) .
Thoracoscopy is recommended for the patient with
M a l i g n a nt P l e u ra l Effu s i o n
an undiagnosed pleural effusion in whom the diagno
sis of malignancy or tuberculosis is suspected, and in If a patient has a known malignancy, should thora
whom at least one pleural Buid cytology and one pleu coscopy be performed to effect a pleurodesis? It appears
ral Buid marker for tuberculosis (adenosine deaminase that the success rates are comparable for pleurodesis
or interferon) have been negative. with thoracoscopy and with tube thoracostomy. In
There are clinical findings that make it more likely one study, Dresler et al. (4 1 ) randomized 482 patients
that malignancy will be diagnosed at the time of with malignant pleural effusions to have pleurodesis
thoracoscopy. Ferrer et al. (3 5) prospectively studied with talc slurry or insuffiated talc and reported that
93 patients referred for thoracoscopy at a tertiary the rate of successful 30-day outcomes did not differ
hospital. They found that the following four variables significantly (78% success with talc insuffiation and
were associated with pleural malignancy in a multi 7 1 % success with talc slurry) . In a smaller study,
variate model: (a) a symptomatic period of more than Yim et al. (42) randomized 5 5 patients to pleurodesis
1 month, (b) absence offever, (c) blood-tinged pleural with thoracoscopy and talc insuffiation or pleurode
Buid, and (d) chest computed tomography (CT) find sis with tube thoracostomy and talc slurry and found
ings suggestive of malignancy (pulmonary or pleural that there was no significant difference in the results
masses, pulmonary atelectasis, or lymphadenopathy) with the two treatment methods (42) . Therefore, if a
(3 5 ) . Twenty-eight patients had all four criteria and patient is known to have a malignant pleural effusion,
all had malignancy. Twenty-one patients had at most it does not seem reasonable to subject him or her to
one criterion and none had malignancy. general anesthesia and the extra expense of thora
When one performs thoracoscopy for diagnostic coscopy when he or she could be managed just as
purposes, it is important to be prepared to perform effectively with tube thoracostomy and a tetracycline
a procedure to create a pleurodesis during surgery. derivative intrapleurally (43) .
C H APT E R 30 I T H O RACOSCO PY 48 5
There are certain situations in which thoracoscopy stay ( 1 8 .4 ± 6.5 days) compared with those who did
is indicated in patients with malignancy. Thoracoscopy not require another procedure ( 1 1 .4 ± 3 . 1 days) (50) .
should be considered if a patient has a malignant pleu Multiloculated empyema can also be treated with
ral effusion that is loculated because the loculations medical thoracoscopy. Brutsche et al. (5 1 ) reported
can be broken down and the pleural space cleared with their experience with 1 27 patients with multiloculated
thoracoscopy. Thoracoscopy should be considered in empyema subjected to medical thoracoscopy at three
patients with ovarian carcinoma and pleural effusion different centers. There were no mortalities and 62 of
because the amount of tumor present in the pleural the patients (49%) received intrapleural fibrinolytic
space will dictate surgical therapy (44) . In the patient therapy for 3 to 5 days after thoracoscopy. The results
with lung cancer and cytology negative pleural effusion, in this retrospective study with medical thoracoscopy
thoracoscopy is indicated to rule out tumor involve were excellent-1 1 5 patients (9 1 %) needed no addi
ment of the pleura (23). tional procedures (5 1 ) . The median duration of chest
tube drainage post medical thoracoscopy was 7 days,
and there was no in-hospital mortality (5 1 ) .
Para p n e u m o n i c P l e u ra l Effu s i o n
When faced with a patient with a complicated
Thoracoscopy should be considered for the patient parapneumonic effusion that is not completely drained,
with a parapneumonic effusion that is not drained there are basically four alternatives: (a) insert additional
with either a therapeutic thoracentesis or tube thora chest tubes, (b) instill fibrinolytics and DNAse intra
costomy. During thoracoscopy, the loculi in the pleu pleurally, (c) perform thoracoscopy, or (d) perform
ral space can be disrupted, the pleural space can be thoracotomy. The insertion of additional chest tubes
completely drained, and the chest tube can be opti is not recommended because it is usually ineffective.
mally placed. In addition, the pleural surfaces can be A recent randomized double-blind study (52) dem
inspected to determine the necessity for further inter onstrated that the intrapleural instillation of 1 0-mg
vention such as decortication. Thoracoscopy should tissue plasminogen activator (tPA) plus 5-mg DNase
not be performed only for pleural thickening as such twice a day for 3 days resulted in more clearing
thickening resolves with time. A CT scan should be of the chest radiograph and a reduction in the need
obtained before thoracoscopy to provide anatomic for additional surgery compared with either agent by
information about the size and extent of the empy themselves or placebo. Therefore, if fibrinolytic ther
ema cavity and the thickness of the peel over the apy is used, this is the preferred regimen. If fibrino
visceral pleura. If the fibropurulent material cannot lytics are used and drainage is incomplete after 3 to
be removed completely and if complete expansion of 5 days, thoracoscopy should be performed (53) . Some
the lung cannot be obtained, the procedure should be have recommended not using the fibrinolytics and
converted to an open thoracotomy (45 ) . directly proceeding to thoracoscopy (54), and there are
Incompletely drained parapneumonic effusions no controlled studies comparing intrapleural therapy
are treated very effectively with thoracoscopy. Most with tPA and DNase with thoracoscopy. Thoracos
reports on thoracoscopy for complicated parapneu copy seems more effective if it is performed before
monic effusion have used VATS. When four series fibrinolytics are administered (54) . At the present time,
(46-49) from the late 1 990s using VATS are com I proceed directly to thoracoscopy if the patient is a
bined, thoracoscopy was the definitive procedure in good operative candidate because I believe that this
1 78 of 232 patients (77%) . The overall mortality was will result in a shorter hospitalization. However, if the
3%. The median time for chest tube drainage after patient is not a good operative candidate, I will try tPA
the procedure ranged from 3.3 to 7. 1 days, and the plus DNase initially. Thoracotomy is reserved for those
median duration of hospital stay after thoracoscopy cases in which thoracoscopy fails or in which thoracos
ranged from 5 . 3 to 1 2.3 days. Luh et al. (50) reviewed copy is indicated but is unavailable.
their experience with VATS in 234 patients with com
plicated parapneumonic effusions in 200 5 . Thora
Postp n e u m o n ecto my E m pye m a
coscopy was the definitive procedure in 1 94 patients
(83%), whereas 40 patients required conversion to Thoracoscopy also appears to be useful in patients
open thoracotomy or reoperation (50). The mortality with postpneumonectomy empyemas. Gossot et al.
rate within 30 days of surgery was 3%. Patients who (55) reported their experience using thoracoscopy
required procedures in addition to thoracoscopy had in 1 1 patients with postpneumonectomy empyema.
a longer mean duration of mean preoperative hospital Thoracoscopy was performed to remove as much
486 PLE U RAL D I S EASES
infected material as possible, to check that there was reported by Margolis et al. (62) who treated 1 56 young
no bronchopleural fistula, and to wash and drain the adults with primary spontaneous pneumothorax via
pleural cavity. Eight of the 1 1 patients were cured of VATS with stapling of blebs and pleural abrasion. In
their postpneumonectomy empyema within a mean this uncontrolled study, there were no postoperative air
drainage time of approximately 1 0 days postthora leaks and the mean duration of hospital stay was only
coscopy ( 5 5 ) . Hollaus et al. (56) reported that five 2.4 days. There were no recurrences during the median
patients with postpneumonectomy empyema without follow-up period of 62 months (62) . Thoracoscopy is
bronchopleural fistula, were cured with thoracoscopy also effective for treating secondary spontaneous pneu
followed by tube thoracostomy until the cultures were mothorax (63,64) .
negative. Most series that have utilized thoracoscopy for
the treatment of spontaneous pneumothorax have
used VATS, but there are some proponents of medi
P n e u m othorax
cal thoracoscopy with the insuffiation of talc (65 ,66) .
Thoracoscopy is effective in the treatment of sponta The effectiveness of talc insuffiation at medical tho
neous pneumothorax and the prevention of recur racoscopy was compared with chest tube drainage
rent pneumothorax. With thoracoscopy, there are two in a prospective randomized multicenter study of
primary objectives: (a) to treat the bullous disease 1 08 patients with primary spontaneous pneumothorax,
responsible for the pneumothorax and (b) to create a most of which were recurrent (65 ) . Patients with bullae
pleurodesis. The most common means by which the greater than 5 cm in diameter were excluded. In this
bullae are treated presently is with an endoscopic sta study, the recurrence rate was 5% in the group that
pling device. The primary disadvantage of the endo received talc and 34% in the group that received
stapler is cost, which is approximately US$ l ,OOO per chest tubes. However, it should be noted that 1 0 of
procedure (57) . Previously, the bullae were treated the 1 6 recurrences in the chest tube group occurred
with electrocoagulation, which was associated with a during the initial hospitalization, whereas only 1 recur
higher recurrence rate ( 5 8 ) . An alternative method of rence occurred in the talc group during hospitalization.
dealing with the apical bullae is to ligate the bullae The recurrence rates after the initial hospitalization
with a Roeder loop (59) . However, Inderbitzi et al. were 5% in the talc group and 1 3% in the chest tube
(59) reported a relatively high recurrence rate after group, although 10 of the 47 patients in the chest tube
use of the loop and recommend that it be abandoned group also got talc during their initial hospitalization.
in favor of wedge resection with the endostapler. The In an uncontrolled study, Lee et al. (66) evaluated the
best way to induce a pleurodesis appears to be with effectiveness of medical thoracoscopy with the insuf
pleural abrasion. It is comparable in effectiveness to flation of talc in the treatment of secondary spontane
talc insuffiation and does not carry the risk of induc ous pneumothorax in patients with advanced chronic
ing acute respiratory failure. It is also comparable in obstructive pulmonary disease (COPD) . They insuf
effectiveness to partial parietal pleurectomy but is flated 3 g of talc in 4 1 patients with a mean age of
easier to perform. 70.7 years and a mean FEV1 of 0 . 8 8 L. The 30-day
The effectiveness ofVATS in conjunction with endo mortality in this group of patients was 1 0% and all
stapling has been demonstrated in several large series. the patients that died had an FEV1 between 0 . 5 and
Cardillo et al. (60) used VATS to treat 432 patients 0 .7 L. The recurrence rate in the survivors was 2 of
with primary spontaneous pneumothorax between 37 (5 .4%) (66) .
1 992 and 1 998. They used subtotal pleurectomy to Which patients with spontaneous pneumothorax
induce a pleurodesis in some patients and talc insuf should be subjected to thoracoscopy? In some centers,
flation in others. In 2.3% of their 432 patients, con all patients with spontaneous pneumothorax are sub
version to an open procedure was necessary, usually jected to thoracoscopy to evaluate the status of the
because of extensive adhesions. The recurrence rate was underlying lung (67) . This approach seems overly
4.4%, with a mean follow-up period of38 months (60) . aggressive because approximately 50% of patients
Yim and Liu (6 1 ) treated 483 patients with primary with their initial pneumothorax will never have a
spontaneous pneumothorax creating a pleurodesis recurrence without any treatment. However, thora
with mechanical abrasion. In this series, the mean coscopy is recommended for patients with primary
postoperative stay was only 3 days and the recurrence pneumothorax in whom aspiration therapy has failed
rate was only 1 .74%, with a mean follow-up period or who have a recurrent pneumothorax. The rea
of 20 months (6 1 ) . The best results with VATS were son for this recommendation is that the duration of
C H APTE R 30 / T H O RACOSCO PY 487
hospital stays are comparable with thoracoscopy and (72) reviewed the literature in 1 999 and reported
tube thoracostomy, but the recurrence rates are much that VATS was effective in controlling the bleeding
less after thoracoscopy. Because the recurrence of a in 33 of 40 (82%) of the patients on whom it was
pneumothorax is more life threatening in patients attempted. Thoracoscopy was effective in controlling
with secondary spontaneous pneumothorax, it is the bleeding when it arose from intercostal vessels or
recommended that thoracoscopy be performed in most from lung lacerations.
patients with secondary spontaneous pneumothorax Approximately 1 0% of traumatic hemothoraces
after they are initially managed with tube thoracos are complicated by retained clotted blood (73) . If
tomy (see Chapter 24) . In general, it seems reasonable more than 30% of the hemithorax is occupied by
to hypothesize that VATS with the stapling of blebs clotted blood after tube thoracostomy, removal of
and pleural abrasion would be superior to medical the blood is usually recommended. Traditionally, the
thoracoscopy with the insufHation of talc but there are clotted blood has been removed with a thoracotomy.
no controlled studies comparing the two approaches. However, thoracoscopy now appears to be the optimal
The ACCP and the British Thoracic Society (BTS) method for the removal of this clotted blood (72-75) .
have both published guidelines for the management Carrillo and Richardson (7 4) reviewed 25 patients
of spontaneous pneumothorax in the last few years. with retained thoracic collections who underwent
The ACCP guidelines (68) stated that thoracoscopy 26 VATS procedures for this problem. They reported
was the preferred intervention for primary sponta that the procedure was successful in 1 9 of the 25 patients.
neous pneumothorax and that it be performed after If the procedure was performed within 7 days of the ini
an ipsilateral recurrence. They recommended that tial inj ury, it was more likely to be successful. Advan
patients with apical bullae should undergo intra tages of thoracoscopy over thoracotomy are a lower
operative bullectomy. They also recommended that incidence of wound and pulmonary complications,
parietal pleural abrasion should be performed in a lower need for narcotics, a sooner return to normal
most patients to induce a pleurodesis. The ACCP activities, and a higher rate of return to a normal life
guidelines for patients with secondary spontaneous style (75 ) . Oguzkaya et al. (73) showed that VATS was
pneumothorax recommended an intervention to pre more effective in removing clotting blood than was
vent pneumothorax recurrence after the first occur the intrapleural administration of streptokinase.
rence because of the potential lethality of secondary
pneumothoraces (68) . Otherwise, the recommenda
Chylothorax
tions for primary and secondary pneumothorax were
very similar. The BTS (69) concluded that chemical One of the options for the treatment of a chylothorax
pleurodesis can prevent recurrent pneumothorax, is ligation of the thoracic duct. It is possible to ligate
but it should be performed only if the patient is the thoracic duct through thoracoscopy. Although
unwilling or unable to undergo surgery. The BTS gave thoracoscopy has not been widely employed for the
the following indications for operative intervention: control of chylothorax, there are now more than
(a) second ipsilateral pneumothorax, (b) first contra 1 4 reports of chylothorax being managed successfully
lateral pneumothorax, (c) bilateral spontaneous pneu with VATS in conjunction with either ligation or
mothorax, (d) persistent air leak (> 5-7 days of tube clipping of the thoracic duct (76) . This intervention
drainage; air leak or failure to completely reexpand) , was successful in all the reported cases (76) . It remains
(e) professions at risk (e.g. , pilots, divers) . The BTS to be seen whether the endoscopic closure of chylous
did not make a definitive statement about which sur leaks is more successful or better tolerated than current
gical procedure is preferred for either primary or sec open techniques (77) .
ondary spontaneous pneumothorax (69) .
H e patic Hyd rothorax
H e m oth orax
The management of hepatic hydrothorax is a diffi
Thoracoscopy may replace thoracotomy in some cult problem. The best treatment for this problem is
patients with traumatic hemothorax who otherwise insertion of a transjugular intrahepatic portal systemic
would have been subjected to thoracotomy. How (TIPS) shunt or liver transplantation. If neither of these
ever, thoracotomy rather than thoracoscopy should is feasible, the best alternative treatment is probably
be performed if there is exsanguinating hemorrhage VATS with closure of the diaphragmatic defects and
through the chest tubes (70,7 1 ) . Villavicencio et al. pleurodesis. In one report, 1 8 patients were subjected
488 PLE U RAL D I S EASES
to 21 thoracoscopies with talc insuffiation (3 patients The complications of medical thoracoscopy are
were subjected to a second procedure after the first fewer than those of VATS (85). The mortality rate is
failed) (78) . Diaphragmatic defects were detected and less than 0.5%, and most deaths are thought to be
closed in 5 of the 1 8 patients (28%) . The procedure was unrelated to the procedure (85). The same type of
effective in 1 0 of 2 1 patients (48%) . The median dura complications occur with medical thoracoscopy as with
tion of hospital stay was 1 5 days. The precarious medi VATS and include intrapleural bleeding, prolonged air
cal condition of patients with hepatothorax is reflected leak, and empyema ( 1 8) .
in the 30% mortality in the 3 months following the
surgery (78). Mouroux et al. (79) performed VATS
in eight patients. These workers found and closed dia REFERENCES
phragmatic defects in six of the patients, and none had
I. Landreneau RJ, Mack MJ, Hazelrigg SR, et al. The role of
a recurrent pleural effusion. No defects were found in
thoracoscopy in the management of intrathoracic neoplastic
the remaining two patients, but after talc insuffiation, processes. Semin Thorac Cardiovasc Surg. 1 993;5 : 2 1 9-228 .
the effusions occupied only the lower one third of the 2 . Loddenkemper R . Thoracoscopy-state o f the art. Eur Respir].
hemithorax. Ferrante et al. (80) could find no diaphrag 1 998; 1 1 :2 1 3-22 1 .
matic defects in 1 5 patients with hepatic hydrothorax 3 . Burrows NJ, Ali NJ, Cox GM. The use and development of
medical thoracoscopy in the United Kingdom over the past
subjected to thoracoscopy and talc insuffiation. In
5 years. Respir Med. 2006; 1 00 : 1 234- 1 23 8 .
this series, the hepatic hydrothorax was controlled by 4. Jacobaeus HC. Ueber d i e moglichkeit die zystoskopie bei
the talc insuffiation in 8 of the 1 5 (53%) patients (80) . untersuchung seroser hohlungen anzuwenden. Munch Med
I do not recommend talc insuffiation for the reasons Wochenschr. 1 9 1 0; 57:2090-2092.
5 . Braimbridge MV. The history of thoracoscopic surgery. Ann
discussed in Chapter 1 0, but instead would recom
Thorac Surg. 1 993;56:6 1 0-6 1 4 .
mend pleural abrasion. 6. Day J C , Chapman PT, O'Brien EJ. Closed intrapleural pneu
monolysis: an analysis of 1 ,000 consecutive operations. ] Thorac
Surg. 1 948; 1 7: 5 37-5 54.
CO M P L I CAT I O N S O F T H O RACO S C O PY 7. Jacobaeus HC. The practical importance of thoracoscopy in
surgery of the chest. Surg Gynecol Obstet. I 922;34:289-296.
Thoracoscopy has a relatively low rate of complica 8. Tassi GF, Tschopp JM. The centenary of medical thoracos
tions. Data collected on 1 ,820 patients from the VATS copy. Eur Respir]. 2 0 1 0;36: 1 229- 1 23 1 .
Study Group registry (8 1 ) in 1 993 revealed that the 9 . Adler RH, Rappole BW. Recurrent malignant pleural effu
overall mortality was 2%. Prolonged air leak was the sions and talc powder aerosol treatment. Surgery. 1 967 ;62:
1 000- 1 006.
most frequent complication and occurred in 3.2%;
10. Weissberg D, Ben-Zeev I. Talc pleurodesis. Experience with
significant bleeding resulting in transfusion occurred 360 patients. J Thorac Cardiovasc Surg. 1 993; I 06:689-69 5 .
in only 1 %. Pneumonia and empyema occurred in 1 1 . Andersen I, Poulsen T. Surgical treatment o f spontaneous
1 . 1 % and 0.6%, respectively (8 1 ) . Other researchers pneumothorax. Acta Chir Scand. 1 95 9; 1 1 8 : 1 0 5-1 1 2 .
have reported comparable complication rates (82) . 1 2 . Wied U , Andersen K , Schultz A , e t al. Silver nitrate pleurodesis
in spontaneous pneumothorax. Scand ] Thorac Cardiovasc
Thoracoscopy is tolerated relatively well by the patient
Surg. 1 9 8 1 ; 1 5 :305-307.
with COPD. In the study mentioned in the preceding 1 3 . Kaiser LR. Video-assisted thoracic surgery. Current state of
text, there were 59 patients who had an FEV1 below the art. Ann Surg. 1 994;220:720-734.
1 L who underwent thoracoscopy. There was one 14. Brandt H-J, Loddenkemper R, Mai J. Atlas ofDiagnostic Tho
death ( 1 .7%) , and the average postprocedure hospital racoscopy. New York, NY: Thieme Medical Publishers; 1 9 8 5 .
1 5 . Boutin C, Viallat J R , Aelony Y. Practical Thoracoscopy. Berlin,
stay was only 5 .4 days (8 1 ) . Jancovici et al. (83)
Germany: Springer-Verlag; 1 99 1 .
reviewed 937 VATS procedures at four surgical insti 1 6 . Loddenkemper R, Boutin C . Thoracoscopy: present diagnostic
tutions from June 1 99 1 through May 1 99 5 . In this and therapeutic indications. Eur Respir]. 1 993;6: 1 544- 1 5 5 5 .
series, approximately half the procedures were per 1 7 . Ernst A , Silvestri GA, Johnstone D . American College of
formed for pleural problems. They reported that Chest Physicians. Interventional pulmonary procedures:
guidelines from the American College of Chest Physicians.
the in-hospital mortality rate was 0 .5 % , and death Chest. 2003; 1 23: 1 693- 1 7 1 7.
occurred principally in patients operated on for malig 1 8 . Loddenkemper R. Medical thoracoscopy. In: Light RW,
nant pleural effusions. In their series, the overall inci Lee YCG, eds. Textbook ofPleural Diseases, 2nd edition. London,
dence of postoperative complications was 1 0 .9%; the England: Hodder Arnold Publishers; 2008 : 5 83-5 9 8 .
1 9 . Medford AR, Agrawal S, Bennett JA, e t al. Thoracic ultra
most common complications were prolonged air leak
sound prior to medical thoracoscopy improves pleural access
(6.7%) and pleural effusion (0.7%) . There have also and predicts fibrous septation. Respirology. 20 1 0; 1 5 : 804-80 8 .
been at least 2 1 instances of tumor seeding of VATS 20. Tassi G, Marchetti G. Minithoracoscopy: a less invasive
incisions reported by the VATS study group (84) . approach to thoracoscopy. Chest. 2003; 1 24: 1 975-1 977.
C H APTE R 30 / T H O RACOSCO PY 489
2 1 . Tassi GF, Marchetti GP, Pinelli V. Minithoracoscopy: a com 42. Yim A P, Chan AT, Lee TW, e t al. Thoracoscopic talc insuf
plementary technique for medical thoracoscopy. Respiration. llation versus talc slurry for symptomatic malignant pleural
20 1 1 ;82:204-206. effusion. Ann Thorac Surg. 1 996;62: 1 65 5 - 1 6 5 8 .
22. Ernst A, Hersh CP, Herth F, et al. A novel instrument for the 43. Light RW. Should thoracoscopic talc pleurodesis be the first
evaluation of the pleural space: an experience in 34 patients. choice management for malignant pleural effusion? No. Chest.
Chest. 2002; 1 22: 1 530- 1 534. 20 1 2 ; 1 42: 1 7- 1 9 .
23. Lee P, Colt H G . Rigid and semirigid pleuroscopy: the future is 4 4 . Chi DS, Abu-Rustum NR, Sonoda Y, e t al. Th e benefit of
bright. Respirology. 200 5 ; 1 0: 4 1 8-425 . video-assisted thoracoscopic surgery before planned abdomi
2 4 . Cerfolio RJ, Bryant AS , Sheils TM, e t al. Video-assisted nal exploration in patients with suspected advanced ovarian
thoracoscopic surgery using single-lumen endotracheal tube cancer and moderate to large pleural effusions. Gynecol Oncol.
anesthesia. Chest. 2004; 1 26:28 1 -2 8 5 . 2004;94:307-3 1 1 .
2 5 . Migliore M, Giuliano R , Aziz T, e t al. Four-step local anesthesia 45. Wurnig PN, Wittmer V, Pridun NS, et al. Video-assisted thoracic
and sedation for thoracoscopic diagnosis and management of surgery for pleural empyema. Ann Thorac Surg. 2006;8 l :309-3 1 3 .
pleural diseases. Chest. 2002; 1 2 1 :2032-2035. 4 6 . Landreneau RJ, Keenan RJ, Hazelrigg S R , e t al. Thoracoscopy
26. Breen D, Fraticelli A, Greillier L, et al. Redo medical thora for empyema and hemothorax. Chest. 1 99 5 ; 1 09 : 1 8-24.
coscopy is feasible in patients with pleural diseases-a series. 47. Cassina PC, Hauser M, Hillejan L, et al. Video-assisted tho
Interact Cardiovasc Thorac Surg. 2009;8:330-333. racoscopy in the treatment of pleural empyema: stage-based
27. Hucker J, Bhatnagar NK, al-Jilaihawi AN, et al. Thoracoscopy management and outcome. J Thorac Cardiovasc Surg. 1 999;
in the diagnosis and management of recurrent pleural effusions. 1 1 7:234-238.
Ann Thorac Surg. 1 99 1 ;52: 1 1 45-1 1 47. 48. Lawrence DR, Ohri SK, Moxon RE, et al. Thoracoscopic
2 8 . Menzies R, Charbonneau M . Thoracoscopy for the diagnosis debridement of empyema thoracis. Ann Thorac Surg. 1 997;
of pleural disease. Ann Intern Med. 1 99 1 ; 1 1 4:27 1 -276. 64: 1 448- 1 4 5 0 .
29. Hansen M, Faurschou P, Clementsen P. Medical thoracos 49. Striffeler H, Gugger M, Im Hof V, et al. Video-assisted tho
copy, results and complications in 1 46 patients: a retrospective racoscopic surgery for fibrinopurulent pleural empyema in
study. Respir Med. 1 998;92:228-232. 67 patients. Ann Thorac Surg. 1 998;65 :3 1 9-323.
30. Chrysanthidis MG, Janssen JP. Autofluorescence videothora 50. Luh SP, Chou MC, Wang LS, et al. Video-assisted thoraco
coscopy in exudative pleural effusions: preliminary results. Eur scopic surgery in the treatment of complicated parapneumonic
Respir]. 2005;26:989-989. effusions or empyemas: outcome of 234 patients. Chest. 2005;
3 1 . de Groot M , Walther G . Thoracoscopy in undiagnosed pleural 1 27: 1 427- 1 432.
effusions. S Afr Med]. 1 99 8 ; 8 8 :706-7 1 1 . 5 1 . Brutsche MH, Tassi GF, Gyorik S, et al. Treatment of sono
32. Emad A, Rezaian GR. Diagnostic value o f closed percuta graphically stratified multiloculated thoracic empyema by
neous pleural biopsy vs pleuroscopy in suspected malignant medical thoracoscopy. Chest. 2005 ; 1 28 :3303-3309.
pleural effusion or tuberculous pleurisy in a region with a high 52. Rahman NM, Maskell NA, West A, et al. lntrapleural use of
incidence of tuberculosis: a comparative, age-dependent study. tissue plasminogen activator and DNase in pleural infection.
Respir Med. 1 99 8;92:48 8-492. N Engl] Med. 2 0 1 1 ;365 : 5 1 8-526.
33. Diaco n AH, Van de Wal BW, Wyser C, et a!. Diagnostic tools in 53. Silen ML, Naunheim KS . Thoracoscopic approach to the
tuberculous pleurisy: a direct comparative study. Eur Respir]. management of empyema thoracis. Indications and results.
2003;22 : 5 8 9-5 9 1 . Chest Surg Clin N Am. 1 996;6:49 1 -499.
34. Kendall S W, Bryan AJ, Large SR, e t al. Pleural effusions: is 54. Petrakis IE, Kogerakis NE, Drositis IE, et al. Video-assisted
thoracoscopy a reliable investigation? A retrospective review. thoracoscopic surgery for thoracic empyema: primarily, or after
Respir Med. 1 992;86:437-440. fibrinolytic therapy failure? Am ] Surg. 2004; 1 87:47 1-474.
3 5 . Ferrer ], Roldan J, Teixidor J, et al. Predictors of pleural malig 5 5 . Gossot D, Stern JB, Galetta D, et al. Thoracoscopic manage
nancy in patients with pleural effusion undergoing thoracos ment of postpneumonectomy empyema. Ann Thorac Surg.
copy. Chest. 2005 ; 1 27: 1 0 1 7- 1 022. 2004;78 :273-276.
36. Olivares-Torres CA, Laniado-Laborin R, Chavez-Garcia C, 56. Hollaus PH, Lax F, Wurnig PN, et al. Videothoracoscopic
et al. Iodopovidone pleurodesis for recurrent pleural effusion. treatment of postpneumonectomy empyema. J Thorac Cardio
Chest. 2002; 1 22 : 5 8 1-583. vasc Surg. 1 999; 1 1 7:397-398.
37. Crnjac A, Sok M, Kamenik M . Impact of pleural effusion 57. Hazelrigg SR, Landreneau RJ, Mack M, et al. Thoracoscopic
pH on the efficacy of thoracoscopic mechanical pleurodesis stapled resection for spontaneous pneumothorax. J Thorac
in patients with breast carcinoma. Eur J Cardiothorac Surg. Cardiovasc Surg. 1 993; 1 05 : 3 89-393.
2004;26:432-436. 5 8 . Takeno Y. Thoracoscopic treatment of spontaneous pneumo
3 8 . Groth G, Gatzemeier U, Haubingen K, et al. lntrapleural pal thorax. Ann Thorac Surg. 1 993;56:688-690.
liative treatment of MPEs with mitoxantrone versus placebo 5 9 . Inderbitzi RGC, Leiser A, Furrer M, et al. Three years' experi
(pleural tube alone) . Ann Oncol. 1 9 9 1 ;2:2 1 3-2 1 5 . ence in video-assisted thoracic surgery (VATS) for spontaneous
3 9 . Sorensen PG, Svendsen TL, Enk B . Treatment o f MPE with pneumothorax.J Thorac Cardiovasc Surg. 1 994; 1 07: 1 4 1 0- 1 4 1 5 .
drainage, with and without instillation of talc. Eur J Respir 6 0 . Cardillo G, Facciolo F, Giunti R , e t al. Videothoracoscopic
Dis. 1 9 84;65 : 1 3 1 - 1 3 5 . treatment of primary spontaneous pneumothorax: a 6-year
4 0 . Davies H E , Nicholson J E , Rahman N M , e t al. Outcome of experience. Ann Thorac Surg. 2000;69 :357-36 1 .
patients with nonspecific pleuritis/fibrosis on thoracoscopic 6 1 . Yim AP, Liu H P. Video assisted thoracoscopic management
pleural biopsies. EurJ Cardiothorac Surg. 20 1 0;38 :472-477. of primary spontaneous pneumothorax. Surg Laparosc Endosc.
4 1 . Dresler CM, Olak J, Herndon JE II, et al. Phase III intergroup 1 997;7:236-240.
study of talc poudrage vs talc slurry sclerosis for malignant 62. Margolis M, Gharagozloo F, Tempesta B, et al. Video
pleural effusion. Chest. 2005 ; 1 27:909-9 1 5 . assisted thoracic surgical treatment of initial spontaneous
490 PLE U RAL D I S EASES
pneumothorax in young patients. Ann Thorac Surg. 2003;76: 74. Carrillo EH, Richardson J D . Thoracoscopy in the management
1 66 1 - 1 663. of hemothorax and retained blood after trauma. Curr Opin
63. Wait MA. AIDS-related pneumothorax. Ann Thorac Surg. Pulm Med. 1 998 ;4:243-246.
1 997;64:290-29 1 . 75. Ben-Nun A, Orlovsky M, Best LA. Video-assisted thoraco
64. Waller DA. Video-assisted thoracoscopic surgery fo r sponta scopic surgery in the treatment of chest trauma: long-term
neous pneumothorax-a 7-year learning experience. Ann R benefit. Ann Thorac Surg. 2007; 83:383-3 87.
Coll Surg Engl. 1 999;8 1 :387-392. 76. Kumar S, Kumar A, Pawar DK. Thoracoscopic management of
65. Tschopp JM, Boutin C, Astoul P, et al. ESMEVAT team. thoracic duct injury: is there a place for conservatism? J Post
(European Study on Medical Video-Assisted Thoracoscopy.) grad Med. 2004; 50:57-5 9 .
Talcage by medical thoracoscopy for primary spontaneous 7 7 . Ferguson M K . Thoracoscopy for empyema, bronchopleural
pneumothorax is more cost-effective than drainage: a ran fistula, and chylothorax. Ann Thorac Surg. 1 993;56:644-645 .
domised study. Eur Respir]. 2002;20: 1 003-1 009. 78. Milanez d e Campos JR, Filho LOA, Werebe EC, e t al. Tho
66. Lee P, Yap WS, Pek WY, et al. An audit of medical thora racoscopy and talc poudrage in the management of hepatic
coscopy and talc poudrage for pneumothorax prevention in hydrothorax. Chest. 2000; 1 1 8 : 1 3- 1 7 .
advanced COPD. Chest. 2004; 1 2 5 : 1 3 1 5-1 320. 79. Mouroux J, Perrin C, Venissac N, e t a l . Management o f pleu
67. Janssen JP, van Mourik J, Cuesta Valentin M, et al. Treatment ral effusion of cirrhotic origin. Chest. 1 996; 1 09 : 1 093- 1 096.
of patients with spontaneous pneumothorax during videotho 80. Ferrante D, Arguedas MR, Cerfolio RJ, et al. Video-assisted
racoscopy. Eur Respir]. 1 994;7: 1 2 8 1 - 1 284. thoracoscopic surgery with talc pleurodesis in the manage
68. Baumann MH, Strange C, Heffner JE, et al. Management of ment of symptomatic hepatic hydrothorax. Am J Gastroenterol.
spontaneous pneumothorax: an American College of Chest 2002;97:3 1 72-3 1 7 5 .
Physicians Delphi Consensus Statement. Chest. 200 1 ; 1 1 9 : 8 1 . Hazelrigg SR, Nunchuck SK, LoCicero J. The video assisted
590-602. thoracic surgery study group. Video assisted thoracic surgery
69. Henry M, Arnold T, Harvey J. BTS guidelines for the manage study group data. Ann Thorac Surg. 1 993;56: 1 039-1 044.
ment of spontaneous pneumothorax. Thorax. 2003; (suppl. 2): 82. Imperatori A, Rotolo N, Gatti M , et al. Peri-operative compli
II39-II52. cations of video-assisted thoracoscopic surgery (VATS) . Int ]
70. Smith RS, Fry WR, Tsoi EK, et al. Preliminary report on Surg. 2008;6: (suppl. l ) :S78-S 8 1 .
videothoracoscopy in the evaluation and treatment of thoracic 83. Jancovici R, Lang-Lazdunski L, Pons F, e t al. Complications
injury. Am ] Surg. 1 993; 1 66:690-693. of video-assisted thoracic surgery: a five-year experience. Ann
71. Manlulu AV, Lee Tw, Thung KH, et al. Current indications Thorac Surg. 1 996;6 1 : 533-537.
and results of VATS in the evaluation and management of 84. Downey RJ, McCormack P, LoCicero J III. The video-assisted
hemodynamically stable thoracic injuries. Eur ] Cardiothorac thoracic surgery study group. Dissemination of malignant
Surg. 2004;25 : 1 048- 1 053. tumors after video-assisted thoracic surgery: a report of twenty
72. Villavicencio RT, Aucar JA, Wall MJ Jr. Analysis of thoracos one cases. J Thorac Cardiovasc Surg. 1 996; 1 1 1 :954-960.
copy in trauma. Surg Endosc. 1 99 9 ; 1 3:3-9. 85. Mathur PN, Astoul P, Boutin C. Medical thoracoscopy. Tech
73. Oguzkaya F, Akcali Y, Bilgin M . Videothoracoscopy versus nical details. Clin Chest Med. 1 9 9 5 ; 1 6 :479-486.
intrapleural streptokinase for management of post traumatic
retained haemothorax: a retrospective study of 65 cases. Injury.
2005;36: 526-529.
ERRNVPHGLFRVRUJ
A Amphiboles, 1 8 9
Abdominal surgical procedures, 305 Amphotericin B, and blastomycosis, 264
Abram's needle, 45 9-46 1 Amylase determination, 99
biopsy, 1 60 Amyloidosis, pleural effusions with, 354-3 5 5
Acquired immunodeficiency syndrome (AIDS) : Angioimmunoblastic lymphadenopathy, 1 5 5
and pleural effusions, 277-280 Angiopoietins, 5 9
approach to AIDS patient, 280 Animal models o f pleural disease, 3 1 -5 1
Kaposi's sarcoma, 277-278 asbestos and mesothelioma, 33-35
miscellaneous pleural effusions, 279-280 inhaled asbestos, 34
parapneumonic effusion and empyema, 278-279 intraperitoneal injections of asbestos, 35
pneumocystis jiroveci pneumonia, 279 intrapleural injections of asbestos, 33-34
primary effusion lymphoma, 278 intratracheal injections of asbestos, 34-3 5
tuberculous pleural effusions, 279 empyema, 3 5-39
pneumothorax secondary to, 378-3 80 hypersensitivity reactions, 42
Actinomycosis, 267-268 malignancy, 3 1 -33
clinical manifestations, 267 adenocarcinoma, 3 1 -33
diagnosis, 267-268 fibrosarcoma, 33
treatment, 268 melanoma, 33
Acute eosinophilic pneumonia, pleural effusions with, 3 5 5 pleurodesis, 42--47
Acute pancreatitis, 296-297 tuberculosis, 39--40
exudative pleural effusion accompanying, 296 Ankylosing spondylitis, 337
Acute respiratory distress syndrome (ARDS), 1 70- 1 72 Anthrax, 2 1 5-2 1 6
pleural effusions with, 357-3 5 8 Antigens, tuberculosis, 1 1 2-1 1 3
Adenocarcinoma, 3 1 -33, 1 06, 1 3 5 Antinuclear antibody (ANA), 1 1 3- 1 1 4
i n mice, 3 1 -32 Antisense oligonucleotides, and mesothelioma, 2 0 1
in rabbits, 33 Antivascular endothelial growth factor (anti-VEGF) , 3 5 5
in rats, 32-33 Apical thickening, 7 9
Adenosine deaminase (ADA) , 1 07, 1 33, 2 5 1 -252 Aquaporins (AQPs) , role i n pleural fluid exchange, 1 5
measurement, 109 Argyle drainage systems (Sherwood Medical, Tullamore,
Adenovitus pneumonia, pleural effusion due to, 2 8 1 -282 Ireland), 473
Adult respiratory distress syndrome (ARDS), 26-27 Argyle Turkel Safety Tboracentesis set (Kendall Company,
AgNOR staining, 1 05 St. Louis MO), 368, 449--450, 452, 453
Air leaks, 476 Arrow-Clark� Pleura-Seal• Tboracentesis Kit, 368, 449, 453, 454
classification of, 392 Asbestos, defined, 1 8 9
and mechanical ventilation, 393 Asbestos bodies, 432
prolonged, 393-394 Asbestos exposure:
AIRFIX, 392 and benign exudative pleural effusions, 348-349
Air-fluid level, in pleural space, 76-78 clinical manifestations, 349
Aldan blue stain, 1 06, 1 6 1 diagnosis, 349
All-trans-retinoic acid, and pleural disease, 343 incidence, 348
Allergic pleurisy, model of, in rats, 42 pathogenesis, 348-349
Alveolar-pleural fistulas, 392-3 93 pathologic features, 348-349
air leaks: prognosis, 349
classification of, 392 and pleural calcification, 436
and mechanical ventilation, 393 pleural disease due to, 43 1
postoperative, 393-395 Asbestos-induced thickening, 80, 438
Amebiasis: Asbestos pleural effusion, 3 5 , 348-349
clinical manifestation, 271-272 Ascites:
diagnosis, 27 1 -272 pleural effusion in patients with, 1 44
pathogenesis, 271 pleural effusion secondary to, 145-146
treatment, 272 Aspergillosis, 263-364
American Joint Committee on Cancer (AJCC), 1 96 clinical manifestations, 263
Amiodarone, and pleural disease, 342 diagnosis, 263-264
Amosite, 1 89 treatment, 264
49 1
492 I N D EX
Atrium drainage systems (Atrium Medical Corporation) , 473 and pneumothorax, 38 1-382
Atypical effusion, 67-68 pathogenesis, 3 8 1
Atypical pneumochorax, 82 treatment, 38 1-382
Autopsy studies, and pleural effusion, 1 57 Cellular bridges, 3
Cerebrospinal fluid (CSP), leakage into pleura, 149
B CHARAS oncogene, 108
Bacillus anthracis, 2 1 5 Chemical pleurodesis, prognostic factors of, 1 67-1 68
BACTEC system, 1 1 7 Chemotherapy, 1 57, 1 64
Basic fibroblast growth factor (b-FGF), 58 Chest CT, 74-76
Basophils, 95 Chest radiographs, 1 57-1 58
Behs:ec's syndrome, 337 and diagnostic thoracentesis, 452
Benign exudacive pleural effusions, and asbestos exposure, Chest cubes, 466-480
348-349 autotransfusion, 478
clinical manifestations, 349 care of, 475-477
diagnosis, 349 checking functional status of cube, 476-477
incidence, 348 drainage, 477
pathogenesis, 348 water-seal chamber, 476
pathologic features, 348-349 commercially available drainage systems, 473-475
prognosis, 349 Pleur-Evac system, 474-475
Bevacizumab, 1 64, 355 injection of materials through, 475
Bilateral pleural effusion, 1 38 insertion, 466-470
and superior vena caval thrombosis, 352 guidewire cube thoracostomy, 466-468
Bilateral pleural plaques, 437 minimizing risk of, 477
Bilious pleural effusions, 306-307 operative cube thoracostomy, 468-469
Birr-Hogg-Dube syndrome, 364 single-port thoracostomy, 469
Blastomycosis, and amphotericin B, 264 training requirements, 466
Bleeding, and pleural biopsy, 463 crocar cube choracostomy, 468
Bleomycin, 1 73, 37 1 intrinsic negative pressure in, 473
Blevosky syndrome, 439 large-bore, 466
Blind needle biopsy, 1 60 placement, verification of, 470
and mesotheliomas, 1 94 pleural drainage systems, types of, 470-473
Blood gases: removal of, 478-479
effusions effects on, 26-27 size, 466
pneumothorax effects on, 20-2 1 small-bore, 466
Body cavity-based lymphoma, 203 cube choracostomy, complications of, 477-478
Borderline complicated parapneumonic effusion, 224 Chest x-ray, and chest cube removal, 479
Brachiocephalic vein, obstruction of, 149 Children, empyema in, 233-236
Breast carcinoma, and pleural effusion, 1 54-1 5 5 Chromosomal analysis, 1 08-109
British Thoracic Society (BTS), 128 Chronic obstructive pulmonary disease (COPD) , and
guidelines for management of spontaneous pneumothorax, 452
pneumochorax, 367 Chronic pancreatic pleural effusion, 297-299
Bromocriptine, 44 1 Chrysotile, 33, 1 8 9
and bilateral pleural chickening, 438 Chrysotile asbestos fibers, intrapleural injection of, and pleural
Bronchopleural fisculas, 392-393 effusion, 432
air leaks: Churg-Strauss syndrome, 1 33, 336
classification of, 392 Chyle, 4 1 2
and mechanical ventilation, 393 Chyliform pleural effusions, 425-426
empyema associated with, 236 clinical manifestations, 425-426
postoperative, 393-395 defined, 425
Bronchoscopy, 136 diagnosis, 426
pathogenesis, 425
c treatment, 426
C-reactive protein (CRP) , 1 1 2 Chylothorax, 1 56, 487
Candida albicans, 350 clinical manifestations, 4 1 5
Candidiasis, 267 diagnosis, 4 1 5-4 1 6
Capillary permeability, increased, 1 4 lipophilic dye ingestion, 4 1 6
Carcinoembryonic antigen (CEA), 1 0 5 thoracic duct, imaging of, 4 1 6-4 1 7
Cardiac function, effect o f pneumothorax on, 2 1 -22 triglyceride measurement, 4 1 6
Cardiotoxicity, mitoxantrone, 1 74 diagnosis o f, 1 1 5-1 1 7
Carcageenan-induced pleurisy, 40-4 1 etiologic factors, 4 1 3-4 1 5
Catamenial hemothorax, 4 1 0 ligation o f thoracic duct:
Catamenial pneumochorax: through thoracoscopy, 420-422
diagnosis, 38 1-382 through choracotomy, 420
I N D EX 493
lntrapleural gene transfer, using pleural space for gene therapy, 1 6 adenocarcinoma, 3 1 -33
i n animals, 33-34 fibrosarcoma, 33
lntrapleural injections of asbestos, 33-34, 433 melanoma, 33
lntrasplenic abscess, 302 causes of, 1 54
lntrathoracic splenosis, 442 diagnosis of, in thoracoscopy, 484
diagnosis, 442 tests for diagnosing, 1 04-1 09
lntratracheal injections of asbestos, 34-35 chromosomal analysis, 1 08-1 0 9
Intrinsic negative pressure, in chest tubes, 473 electron microscopic examination, 1 06
Iodopovidone (Betadine) , 44, 1 74-175 flow cyrometry, 1 0 8
lsotretinoin, and pleural disease, 345 histochemical studies, 1 06-1 07
hyaluronic acid, 1 0 8
K immunohisrochemical studies, 1 05- 1 06
Kampmeier foci, 6 lectin binding, 1 0 8
Kaposi's sarcoma, 277-278 nucleolar organizer regions (NOR) , 1 0 5
Karain Rock fiber, 34 oncogenes, 1 07-1 0 8
Karnofsky Performance Scale (KPS) score, 1 62 proteomics, 1 0 9
Kawasaki's disease, 3 3 7 soluble mesothelin related protein (SMRP), 1 07
Klebsiella pneumonia, 2 1 2-2 1 3 tumor markers in pleural fluid, 1 07
Malignant cells:
L size of, 1 04
Lactic acid dehydrogenase (LD H) measurement, vacuolation, 1 04
99-1 0 1 , 1 5 8 Malignant mesotheliomas, 1 8 9-20 1
Lacunas, 1-2 blood markers, 1 95-196
Langerhans cell histiocycosis, pneumothorax due to, 3 8 1 clinical manifestations, 1 9 1
Lassa fever virus, pleural effusion due to, 282 computed tomography (CT) , 193
Lectin binding, 1 0 8 diagnosis, 1 94-1 9 5
Leg veins, duplex ultrasonography o f, 292 etiologic factors, 1 89-1 9 0
Legionnaire's disease, 2 1 6 incidence, 1 9 1
Lemierre syndrome, pleural effusions with, 293 microscopic characteristics o f, 1 9 1
Ligation of superior vena cava, 352 pathologic features, 1 9 1
Ligation of thoracic duct: pleural fluid, 1 93-1 94
through thoracoscopy, 420-422 prognosis, 1 9 6
congenital chylothorax, 421-422 radiographic manifestations, 1 9 1- 1 93
nontraumatic chylothorax, 420-42 1 staging, 1 9 6
through thoracotomy, 420 treatment, 1 97-200
Light index for pneumothorax, 366-367 chemotherapy, 1 99-200
Lipid studies of pleural fluid, 1 1 5-1 1 7, 1 6 1 novel therapies for, 2 0 1
Liver abscess, transdiaphragmatic rupture o f, 272 palliative therapy, 1 97- 1 98
Liver transplant, pleural effusions after, 307 radiotherapy, 200
Living donors for lung transplantation, pleural surgical treatment, 1 98- 1 99
effusion in, 3 5 1 Malignant pleural effusion, 1 53-1 88, 484-48 5
Localized mesothelioma, 2 0 1 Marfan's syndrome, 365
Loculated pleural effusions, 222-223 Massive pleural effusions, 78-79, 1 37-1 3 8
Lung cancer, and pleural effusion, 1 53-1 54 Matrix metalloproteinases, tissue inhibitors o f, 59-60
Lung entrapment, 356 Mediastinal cyst, rupture of, 3 5 5
Lung scans, 290 Mediastinal pleura, 1
Lung sliding, 82 Mediastinal radiation for chylothorax, 1 64-1 6 5
Lupus erychematosus (LE) cells, 1 1 4 Medical thoracoscopy, training requirements, 482-483
Lupus pleuritis, 133 Meigs' syndrome, 1 1 , 1 50, 325-326
Lymphatic clearance, 1 2 clinical manifestations, 325-326
alternative mechanisms, 1 3 diagnosis, 326
through capillaries i n visceral pleura, 1 2- 1 3 etiologic factors, 325
Lymphocytes, 9 5 management of, 326
Lymphomas, and pleural effusion, 1 5 5-1 56 Melanoma, 33
Lysozyme, 1 1 2 Mesalamine, and pleural disease, 345
Mesothelial cells, 95-96
M cells, 4
Macrophages, 96-97 functions of, 3
Magnetic resonance imaging (MRI ) : physicochemical factors, 4
and malignant mesotheliomas, 1 9 3 thickness, 4
and pleural effusion, 76 volume, 4
Malignancy, See also Metastatic malignancies Mesothelioma, 1 06-1 07, 1 6 1
and animal models of pleural disease, 3 1 -33 blood markers of, 1 9 5-196
I N D EX 497