Nonunions

Animesh Agarwal
Editor
Nonunions
Diagnosis,
Evaluation and
Management
123
Nonunions
Animesh Agarwal, MD
Editor
Nonunions
Diagnosis, Evaluation
and Management
123
Editor
Animesh Agarwal, MD
Division of Orthopedic Traumatology,
Department of Orthopedics
University of Texas Health Science
Center
San Antonio, TX
USA
ISBN 978-1-4939-7176-3 ISBN 978-1-4939-7178-7 (eBook)

DOI 10.1007/978-1-4939-7178-7
Library of Congress Control Number: 2017940364
© Springer Science+Business Media LLC 2018

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or
part of the material is concerned, specifically the rights of translation, reprinting, reuse of
illustrations, recitation, broadcasting, reproduction on microfilms or in any other physical way,
and transmission or information storage and retrieval, electronic adaptation, computer software,
or by similar or dissimilar methodology now known or hereafter developed.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are
exempt from the relevant protective laws and regulations and therefore free for general use.
The publisher, the authors and the editors are safe to assume that the advice and information in
this book are believed to be true and accurate at the date of publication. Neither the publisher nor
the authors or the editors give a warranty, express or implied, with respect to the material
contained herein or for any errors or omissions that may have been made. The publisher remains
neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Printed on acid-free paper
This Springer imprint is published by Springer Nature

The registered company is Springer Science+Business Media LLC
The registered company address is: 233 Spring Street, New York, NY 10013, U.S.A.
We must find time to stop and thank the people who make a
difference in our lives.
—John F. Kennedy
First and foremost, I would like to thank my parents, Jagdish
and Kusum, for being extraordinary role models, as well for
their hard work and dedication to parenthood. They
have supported me throughout my life, both personally and
professionally. I am forever grateful. They continue to amaze me
now as grandparents.
To the most delightful, caring, and resilient children I could
have ever had. Thank you Priya, Deven, and Trevor for being so
understanding when I am on call or working late. You remind
me what is most important in life every time I look at you. The
three of you have been and continue to be my inspiration.
I want to thank those who have supported me professionally.
I am lucky to have trained at an institution with many “giants”
of orthopedics, and even more fortunate to have joined them as
a faculty member. They have always treated me as an equal,
despite having been my attendings. So thank you to Drs. James
Heckman, Charles Rockwood, and Fred Corley, who all
believed in me and gave me a chance. Thank you to those
that trained me during my orthopedic trauma fellowship, Drs.
Attila Poka, Robert Ostrum, and Brian Davison for making that
year of training educational, fun, and unforgettable.
Last but not least, I would be remiss if I did not thank my two
research mentors, Drs. Kyriacos Athanasiou and Mauli
Agrawal. They were instrumental in fueling my curiosity,
teaching me to write scientifically, and steering me towards
orthopedics. I wouldn’t be where I am today without their
encouragement and guidance so early in my career.
Animesh Agarwal, MD
Boerne, TX, USA
February 2017
Foreword
It is an honour to be invited to write the foreword to this important contri-

bution to the orthopedic surgical literature. This textbook is an extension
of the robust fracture information and knowledge base originally created in
the Department of Orthopedics at the Health Science Center in San Antonio
by Rockwood and Green in the 1970s. Despite many substantial advances in
primary fracture care over the last several decades, our ability to enhance
fracture healing has not improved very much; thus, delayed unions and
nonunions remain a common clinical challenge. Furthermore, as patients
develop greater expectations for very good or even perfect results after
injury, they are less willing to accept incomplete fracture healing as the final
outcome. Today’s patient wants to be able to return to high-level perfor-
mance following injury and not be sidelined because of a failure of fracture
healing. Therefore, when fracture healing does not occur, it is incumbent
upon the surgeon to identify the cause and develop an effective solution to
restore function in a timely manner. The textbook is a very useful resource to
that end.
Dr. Agarwal has compiled a very comprehensive overview of the causes
of and the effective treatments for virtually all of the common nonunion
patterns. His contributing authors are authoritative and recognized experts in
their respective fields, and the treatments for specific nonunions are presented
in a very comprehensive, well-illustrated, and user-friendly way.
This textbook clearly advances our understanding of the evaluation and
treatment of fracture nonunions, and it will be a very useful guide to the
surgeon who must find an effective means to restore function when the
normal fracture healing process has failed.
Manchester, VT, USA James D. Heckman, MD

March 2017
vii
Preface
An investment in knowledge pays the best interest.
—Benjamin Franklin
Nonunions have been a challenging aspect for orthopedic trauma surgeons.

The literature is scarce with recommendations for many of these problems.
The principles of nonunion management have been traditionally based on the
type of nonunion and its aetiology, whether it is a mechanical or biological
one. As orthopedic care has advanced with newer implants—such as locked
plating—old problems have been alleviated, but new issues have arisen as
well. Although patient factors and injury factors contribute significantly to
the development of nonunions, iatrogenic causes are not infrequent. Ortho-
paedic trauma care still requires attention to detail and the basic principles of
fracture management to help prevent the surgeon’s contribution to nonunion
development.
Over the last twenty years of my practice, nonunions have always been a
difficult problem to assess. New patient issues such as vitamin D deficiency
have come to light. Infection all too often accompanies nonunions and must be
evaluated. This text was designed to provide a single reference for the basic
principles of nonunion diagnosis, evaluation, and management. Although not
every single treatment option can be covered for every single anatomical area
and type of nonunion, we hope that this text is useful, not only for managing
these difficult problems but also in preventing nonunions from occurring by
avoiding surgical causes and mitigating patient risk factors.
The contributors to this text were selected based on their interest and
expertise in this subject. Nonunion management is unique in that it is rarely
taught; rather, surgeons have learned over the years through their own per-
sonal experience, oftentimes by trial and error. Learning a new subject,
technique, or gaining new knowledge requires the desire to be more
well-informed. To quote the great Woody Allen, “Eighty percent of success is
showing up”. The contributors have certainly shown up. If you are reading
this preface, we hope that you have taken a step towards “showing up”, and
thus being successful in the management of these difficult and challenging
cases.
Boerne, TX, USA Animesh Agarwal, MD
ix
Contents
1 Principles of Nonunions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
Animesh Agarwal
2 Fracture Healing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45
Saam Morshed and Anthony Ding
3 Clavicle Nonunions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75
Laura A. Schemitsch, Emil H. Schemitsch
and Michael D. McKee
4 Proximal Humerus Nonunions . . . . . . . . . . . . . . . . . . . . . . . . . 95
Ethan S. Lea and Philip R. Wolinsky
5 Supracondylar Humeral Nonunions . . . . . . . . . . . . . . . . . . . . . 115
Joseph Borrelli, Jr.
6 Nonunions of the Forearm . . . . . . . . . . . . . . . . . . . . . . . . . . . . 125
Fred G. Corley and Ben S. Francisco
7 Nonunions of the Wrist and Hand . . . . . . . . . . . . . . . . . . . . . . 143
Matthew Lyons, Ahmad Fashandi and Aaron M. Freilich
8 Acetabular and Pelvic Nonunions . . . . . . . . . . . . . . . . . . . . . . . 183
Kyle F. Dickson
9 Proximal Femur Nonunions . . . . . . . . . . . . . . . . . . . . . . . . . . . 207
Roman A. Hayda
10 Femoral Shaft Nonunions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 227
Gil R. Ortega and Brian P. Cunningham
11 Distal Femoral Nonunions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 243
Animesh Agarwal
12 Nonunions of the Tibial Plateau and Proximal
Tibial Metaphysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 275
Thomas F. Higgins
13 Tibial Nonunions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 287
Samuel E. Galle and David P. Zamorano
14 Distal Tibia and Ankle Nonunions . . . . . . . . . . . . . . . . . . . . . . 309
Kevin J. Pugh
xi
xii Contents
15 Special Techniques for Nonunions Associated

with Traumatic Bone Loss . . . . . . . . . . . . . . . . . . . . . . . . . . . . 323
Mark A. Lee and Michael P. Leslie
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 339
Contributors
Animesh Agarwal MD Department of Orthopedics, Division of Orthopedic

Traumatology, UTHSCSA, San Antonio, TX, USA
Joseph Borrelli, Jr. MD, MBA Orthopedic Surgery, BayCare Medical
Group, St. Joseph’s Hospital-North, Lutz, FL, USA
Fred G. Corley MD Department of Orthopedics, University of Texas Health
Science Center at San Antonio, San Antonio, TX, USA
Brian P. Cunningham MD Department of Orthopedic Surgery, University
of Minnesota, Regions Hospital, St. Paul, MN, USA
Kyle F. Dickson MD, MBA Southwest Orthopedic Group, Baylor Univer-
sity, Houston, TX, USA
Anthony Ding MD Department of Orthopedic Surgery, San Francisco
General Hospital, Orthopedic Trauma Institute, University of California San
Francisco, San Francisco, CA, USA
Ahmad Fashandi MD Department of Orthopedics, University of Virginia,
Charlottesville, VA, USA
Ben S. Francisco MD Department of Orthopedics, University of Texas
Health Science Center at San Antonio, San Antonio, TX, USA
Aaron M. Freilich MD Department of Orthopedics, University of Virginia,
Samuel E. Galle MD Orthopedic Trauma Service, Orthopedic Trauma &
Fracture Clinic, Saint Alphonsus—Boise Campus, Boise, ID, USA
Roman A. Hayda MD Rhode Island Hospital, Brown University Warren
Alpert School of Medicine, Providence, RI, USA
Thomas F. Higgins MD Department of Orthopedics, University Orthopedic
Center, University of Utah Health Sciences, Salt Lake City, UT, USA
Ethan S. Lea MD, MSc Department of Orthopedic Surgery, University of
California at Davis Medical Center, Sacramento, CA, USA
Mark A. Lee MD Department of Orthopedic Surgery, University of Cali-
fornia, Sacramento, CA, USA
xiii
xiv Contributors
Michael P. Leslie DO Department of Orthopedics and Rehabilitation, Yale

School of Medicine, New Haven, CT, USA
Matthew Lyons MD Department of Orthopedics, University of Virginia,
Michael D. McKee MD, FRCS(C) Division of Orthopedic Surgery,
Department of Surgery, University of Toronto, Toronto, Canada
Saam Morshed MD, PhD, MPH Department of Orthopedic Surgery, San
Francisco General Hospital, Orthopedic Trauma Institute, University of
California San Francisco, San Francisco, CA, USA
Gil R. Ortega MD, MPH Department of Orthopedic Surgery, Mayo Clinic
Arizona Orthopedic Residency Program, Mayo Clinic Arizona, HonorHealth
Scottsdale Osborn Trauma Center, Scottsdale, AZ, USA
Kevin J. Pugh MD Orthopedic Trauma and Reconstructive Surgery, Grant
Medical Center, Columbus, OH, USA
Emil H. Schemitsch MD, FRCS(C) Division of Orthopedic Surgery,
Department of Surgery, University of Toronto, Toronto, Canada
Laura A. Schemitsch MA Division of Orthopedic Surgery, Department of
Surgery, University of Toronto, Toronto, Canada
Philip R. Wolinsky MD Department of Orthopedic Surgery, University of
California at Davis Medical Center, Sacramento, CA, USA
David P. Zamorano MD Orthopedic Trauma Service, Orthopedic Trauma &
Fracture Clinic, Saint Alphonsus—Boise Campus, Boise, ID, USA
Principles of Nonunions
1
Animesh Agarwal, MD
There are however many factors to take into

1.1 Introduction
consideration such as the particular bone
involved, the specific anatomic regions of the
Fracture healing is a very unique process in the
particular bone, the fracture pattern, as well as
human body. Bone is a unique tissue in that it can
the method of treatment. There are certainly
regenerate itself during the process of healing. This
specific areas within the skeleton that already
requires a very complex process which is regulated
have a predisposition to impaired healing due to
by various metabolic and hormonal factors to
both biologic and mechanical factors such as the
include various growth factors. These biological
subtrochanteric femoral region. Additionally, the
processes occur at the cellular level requiring
treatment method may contribute to a nonunion
recruitment proliferation and differentiation of
due to the inadequate mechanical environment
many cells including endothelial cells, osteopro-
provided by the choice of fixation. Often times
genitor cells, platelets, macrophages, mesenchy-
the diagnosis is more retrospective in nature then
mal stem cells (MSCs), and monocytes. These
prospective. Nonoperative interventions such as
cells secrete various biologically active molecules
various noninvasive stimulation devices or
at the site of injury to facilitate fracture repair. The
medications can potentially augment the slow
bone morphogenetic proteins (BMPs) are
fracture healing process.
osteoinductive agents which promote the prolifer-
A delayed union may eventually heal or
ation and differentiation of undifferentiated cells to
eventually may become a nonunion. Often times
become either osteoprogenitor or chondroprogen-
it is difficult to diagnose a nonunion in real time,
itor cells. Although our bodies have the inherent
and much of the time the diagnosis is made ret-
capability to repair the fracture, the fracture healing
rospectively. If the process stops altogether, a
process can be impaired for numerous reasons.
nonunion has developed which may require
When the fracture healing cascade stalls, a
intervention. The US Food and Drug Adminis-
delayed union may develop, but the process may
tration (FDA) has defined a nonunion as a frac-
altogether cease. In a delayed union, both clinical
ture that is at least nine months old and has not
evidence and radiographic evidence of healing
shown any signs of healing progression for at
do progress, but it lags behind what the normal
least three consecutive months [1]. From a clin-
healing time should be for a particular bone.
ical perspective, we define a nonunion as one in
which the normal fracture healing process has
ceased, to the extent that, without further treat-
A. Agarwal (&) ment, healing will not progress. Thus, the
Division of Orthopedic Traumatology, Department
nine-month rule should not be applied to all
of Orthopedics, University of Texas Health Science
Center, San Antonio, TX, USA fractures and be based more upon the clinical
e-mail: Agarwal@uthscsa.edu presentation and the individual patient [2].
© Springer Science+Business Media LLC 2018 1

A. Agarwal (ed.), Nonunions, DOI 10.1007/978-1-4939-7178-7_1
2 A. Agarwal, MD
In addition to a lack of clear-cut “time” guideli- scores. Long-bone nonunions had a utility score of
nes for a nonunion, there is difficulty in assessing 0.68 that was well below that of type-1 diabetes
a fracture for a nonunion based upon radiological (0.88), stroke (0.81), and HIV (0.79). Those with
findings and a wide disparity exists in orthopedic forearm nonunions had the worst quality of life.
surgeons’ perceptions of nonunion criteria and Unfortunately, even with successful treatment of
time points for nonunions [3]. Additionally, it is the nonunion, it has been shown that, at least in
well known that there are certain bones that are at respect to tibial nonunions, there is a long-term
a greater risk to go on to a nonunion. This may negative impact on one’s quality of life [9]. The
be due to the location on a certain bone due to indirect burden to society remains unanswered.
vascularity issues or the whole bone itself, e.g., It has been estimated that between 5 and 10%
scaphoid. In certain situations, the associated all patients will have some difficulty in healing
bone loss that occurs clearly exceeds any critical their fracture [6, 7]. It has also been reported that
size defect and will not heal with fixation alone, 1 out of 6 fractures that have delayed healing will
and thus, a nonunion is the expected result. It go onto a nonunion [10]. Additionally, the inci-
would be inappropriate to delay intervention in dence is also variable depending upon the ana-
these patients until 9 months per the FDA defi- tomic area in question. Unfortunately, the overall
nition. One can clearly see that the details of each incidence of delayed union and nonunion fol-
case must be taken into consideration when lowing fractures has been thought to be increas-
deeming it a nonunion. ing due to various factors including an aging
There has been considerable discussion population, increased obesity, diabetes, smoking,
regarding the costly burden of nonunions finan- vitamin D deficiency, as well as improved sur-
cially, but the affects on functional outcome and vival rates of patients with multiple injuries.
the quality of life can be devastating. In a study of These aforementioned factors certainly affect the
tibia nonunions, the authors found that these biological aspect of fracture healing; however,
patients had high per patient costs overall with the mechanical aspects of fracture healing can
increased healthcare resource usage [4]. In a study also be problematic. The mechanical factors are
by Kanakaris and Giannoudis [5], the increased often dependent upon the type of treatment
costs were also associated with humeral and method chosen by the surgeon in discussion with
femoral nonunions in addition to tibia nonunions. the patient. The mechanical stability that can be
Not only are there direct costs associated with the achieved at the fracture site is dependent upon
treatment, but also significant indirect costs the type of stabilization method used whether it
associated with losses in productivity [6]. Earlier be nonoperative or operative means. Cast stabi-
treatment based on earlier diagnosis could result in lization of the fracture has the least amount of
significant financial savings to the healthcare stability, but can be effective in many fractures
system and society. In addition to the additional that are amenable to nonoperative management.
cost, there are significant impacts to the quality of Methods of surgical fixation include open
life and functional outcome of these patients. In a reduction and internal fixation, external fixation,
study evaluating patients that have tibial shaft and intramedullary nailing. This multitude of
nonunions with functional outcome scores, Brin- options can lead to a vast spectrum of stability.
ker et al. [7] found that the SF-12 scores (physical This affects the type of fracture healing that can
and mental) indicated an extremely disabling occur, either primary or secondary fracture
effect on physical and mental health. The impact healing, in which callous formation occurs in the
on physical health was comparable to that of latter type. The interplay of biologic factors,
end-stage hip arthrosis and worse than congestive including osteogenic cells and the extracellular
heart failure. In a follow-up study, Schottel et al. matrix, which acts as a natural scaffold, and
[8] found that all longbone nonunions had a very growth factors inherent to fracture hematoma
low health-related quality of life based upon Time along with the mechanical environment forms
Trade-off direct measures to determine utility the basis of the diamond concept of fracture
1 Principles of Nonunions 3
healing introduced by Giannoudis et al. [11]. All the risk of complications. A thorough and com-
of these factors should be taken into considera- plete physical examination should be performed
tion in the management of nonunions as well. on all patients presenting with a nonunion. The
Neglect of one of these key cornerstones of physical examination should include a general
fracture healing can doom the treatment of the physical which may point to other underlying
nonunion. disorders that may have been overlooked.
Many people have tried to elucidate factors, Detailed examination of the extremity involved
biological markers, or other aspects of the frac- should be performed to include an evaluation of
ture or treatment that could contribute to a non- the neurovascular status, looking for open
union allowing one to potentially predict which wounds (draining sinuses), healed lacerations
fractures or which patients may progress on to a (indicative of perhaps an open injury), healed
nonunion [12–30]. The establishment of a non- incisions, clinical alignment, joint motion, and
union on radiographs does not necessarily imply examination of the presumed nonunion site for
the need for operative intervention. Nonunions motion. Any open wound or draining sinus in
maybe asymptomatic, and therefore, both clinical proximity to the fracture should lead one to
and radiological findings as well as the patient’s suspect a septic nonunion and is so until proven
current function and wishes are necessary to otherwise. Such open wounds must be taken into
determine the best course of action in the man- consideration, and a soft tissue reconstruction
agement of a nonunion. Surgical intervention of plan will need to be integral to the overall bony
the original fracture can often times make the reconstruction. Previous incisions may limit
diagnosis of a nonunion difficult especially in the options and may dictate how previous hardware
absence of associated hardware failure. Thus, the is removed. Alternative approaches may need to
evaluation, diagnosis, and the treatment of a be employed if the existing soft tissues are
nonunion can be very complicated [10, 31]. It scarred in or suboptimal for further surgical
requires a thorough understanding of the original intervention. If there is a deformity, correction of
injury and treatment, subsequent treatments as the malalignment has to be taken into consider-
well as patient comorbidities, which may have ation as well. This includes any leg length dis-
contributed to the development of the nonunion. crepancy that may need to be addressed. Joint
motion may be limited from arthrofibrosis or a
result from a false joint at the nonunion site, or
1.2 History patients may have developed contractures. Any
surgical plan must take into consideration the
Evaluation of a nonunion should begin, first and need for lysis of adhesions, soft tissue releases,
foremost, with an evaluation of the patient and etc., to insure the best possible overall outcome.
their medical history. A thorough evaluation and In short, preoperative planning taking all these
review of the patient’s past medical and surgical factors into consideration before going down the
history including medications are very important reconstructive pathway is paramount.
in helping to elucidate the etiology of the non- It is extremely important to obtain an accurate
union. It is important to take a medical history history of the original injury mechanism as well
and assess for vascular disease, malnutrition, as other fracture characteristics. It is important to
diabetes, social history, and metabolic bone dis- determine whether or not the fracture was from a
ease such as osteoporosis, endocrine disorders, high-energy or low-energy injury. The extent of
vitamin D deficiency, hepatic and renal disor- the initial soft tissue injury as well as the amount
ders, steroid use, and rheumatologic disorders. of periosteal stripping that may have been
Many of these comorbidities will be discussed encountered at the time of surgery or because of
below under “etiology.” Social issues such as the surgery may shed light on the potential cause
smoking or illicit drug use are important to note of the development of the nonunion. It has been
as these things may prevent healing or increase recently suggested that compartment syndrome
4 A. Agarwal, MD
and associated fasciotomy may be a risk factor 1.3 Risk Factors for Nonunion
for the development of nonunion in tibia frac-
tures [12]. Open fractures obviously have much Biological factors and mechanical factors can
more soft tissue damage, and the potential for an contribute to the development of a nonunion.
occult infection and septic nonunion must also be These can be related to the patient or the inter-
taken into consideration. vention performed by the surgeon. If the patient
A careful evaluation of all previous surgeries has been referred in, as mentioned previously, it
is critical, especially the index operation. Review is helpful to obtain previous injury radiographs,
of the operative reports and/or injury radiographs computed tomography (CT) scans, and other
along with the immediate postoperative films can imaging studies as well as operative reports to
be crucial to understanding the underlying cause. understand what was done and why it was done.
Subsequent interventions should also be evalu- If you are the index surgeon, it is important to
ated in a similar manner, taking into considera- critically asses your own surgical intervention to
tion the pre- and post-op radiographs and the determine whether things that were done may
details of the surgical procedure. If bone grafting have contributed to the nonunion. Decision
or biologic adjuncts had been done or used at any errors can always occur, and what is successful in
time, the type of bone graft or adjunct, the one patient may not be so in another patient. In
location of harvest of the autogenous bone graft, any event, risk factors for the development of a
should be noted. Previous sites of harvest may nonunion can be classified as patient dependent
limit future options. Inadequate fixation or or independent [10, 25]. Many of the indepen-
extensive surgical exposures can be large deter- dent factors are more surgeon-dependent factors
minants in the development of a nonunion. In or injury characteristics.
fractures treated with intramedullary nails, ex- The injury characteristics unique to a specific
ternal fixation, cast stabilization, or bridge plat- fracture location will be discussed in each specific
ing, a relatively stable construct has been created anatomic section, but some generalities can be made.
allowing for callous formation. In cases of open Areas that are known to have tenuous blood supplies
reduction and internal fixation (ORIF), an envi- have been shown to be at risk of nonunion [10, 28,
ronment with absolute stability often is created 32]. Such areas include the femoral neck, sub-
allowing for primary bone healing without cal- trochanteric region of the femur, the scaphoid, the
lous formation. The surgical assault obviously talus, the metadiaphyseal region of the fifth meta-
affects the amount of soft tissue stripping which tarsal, and tarsal navicular body. Open fractures with
can affect the amount of blood supply to the their significant soft tissue stripping clearly have
fracture site. Additionally, past surgical inter- increased risks of nonunion as well as infection [23,
ventions and hardware that is present can cer- 25, 26, 28, 29]. The associated soft tissue injury and
tainly affect future treatment options for the muscle loss in severe open injuries can result in loss
management of the nonunion. of the blood supply to the bone resulting in a detri-
A thorough evaluation of prior complications mental effect on the healing process and increasing
should be performed. Any history of infection the risk of infection. Lin showed that functional
should increase one’s suspicions for continued outcomes in patients with open tibia fractures were
infection even in the absence of clinical signs or worse than those with closed fractures [33]. West-
symptoms. Nerve injuries should be assessed as geest et al. [29] found that fractures which were
this may limit the overall outcome of any non- classified as open grade IIIA injuries were associated
union reconstruction and may lean one toward a with delayed healing and nonunion. Additionally, in
more definitive intervention such as amputation. this prospective cohort of 736 subjects, all with open
Previous vascular injuries may require further long bone fractures, deep infection was associated
assessment in terms of viability of the previous with delayed healing and nonunions. In a retro-
repair and a thorough assessment of the vascular spective study of long-bone fractures treated with
status of the limb. intramedullary nailing, Malik et al. [23] found that
open fractures had a significant association with the stabilization affects gene expression involved in
development of deep infection which also was fracture healing. Relative stability constructs
associated with the development of a nonunion. In such as intramedullary nailing, cast immobiliza-
the same study, they alluded that opening of a closed tion, and external fixation allow the fracture to
fracture also was a significant contributor to the heal by callus formation; however, excessive
development of a nonunion, and therefore, opening motion could lead to a hypertrophic nonunion.
of the fracture, in cases of intramedullary nailing, be The rigidity of the fracture fixation has been
avoided if possible. In the study by Blair et al. [12], shown to improve the process of healing [37].
fasciotomy for compartment syndrome in tibia Reaming of the canal in intramedullary nailing
fractures, which in essence is opening of the fracture, can increase the size of the nail and enhance the
was also associated with significant increase in both mechanical stability. The effect of reaming has
infection and nonunion. In an effort to prevent been looked at extensively [39]. It has been well
infection in open fractures, it is well established that established that reaming enhances fracture heal-
antibiotics be administered as rapidly as possible and ing and that there is a higher incidence of delayed
hopefully within an hour of the fracture presenting union and nonunions in unreamed nails with
[34]. Often times the open fractures are also associ- more secondary procedures to obtain union [23].
ated with significant bone loss and in most cases This is true despite a recent study showing that
such defects cannot heal on their own and are the functional outcomes in tibia fractures were
expected to become nonunions if left alone. These not affected by reaming [33]. Inadequate internal
eventually will require bony reconstruction. The fixation when one is trying to achieve absolute
type of reconstruction, timing of bone graft place- stability to create an environment for primary
ment, and the source of bone graft is highly variable bone healing can also lead to excessive motion
among orthopedic trauma surgeons [35]. Deter- and a subsequent nonunion. Niikura et al. [25]
mining the amount of bone graft for such defects can reviewed 102 nonunions of which almost 80%
be problematic, and some have tried to develop were related to or solely caused by inadequate
quantitative models to determine the amount needed stability or reduction. Conversely, rigidly fixing
[36]. Other fracture characteristics that need to be fracture fragments with gaps or without proper
assessed include the degree of displacement, the internal fixation techniques such as obtaining
extent of comminution, the amount of cortical compression across fracture planes may delay or
apposition at final fixation, and the stability of fixa- even prevent healing [31]. Fixation can be too
tion [24, 25, 28, 32, 37, 38]. rigid leading to a failure in healing. If the patient
Surgeon factors can contribute to either bio- had undergone what was felt to be appropriate
logical reasons for the development of a non- fixation with appropriate surgical technique for
union or a mechanical one [23, 25, 28, 32]. the fracture in question, then it is important to
Contributions to a biological cause include investigate patient-related factors, both biological
excessive stripping of soft tissues, failure to bone and mechanical, that may have contributed to the
graft at the appropriate time, and inadequate development of the nonunion. Brinker et al. [13]
debridement of devitalized/dead bone, which can created an algorithm on when to refer patients for
lead to infection, which then may prevent union. endocrine workups in relation to their nonunion.
Mechanical factors introduced by the surgeon are When evaluating the nonunion, the technical
related to the method of treatment and/or implant aspects of the fracture fixation should be asses-
for the original fracture. Fracture stabilization has sed. If there was no technical error, then it was
significant affects on fracture healing. In a liter- suggested that perhaps there was a metabolic
ature review by Hildebrand et al. [37], the type etiology to the nonunion, and thus, the patient
and timing of fracture stabilization can alter the should be referred to an endocrinologist. If
systemic inflammatory response after trauma technical error was a crucial factor in the
and can affect fracture healing. They also found etiology, referral was not indicated. However, it
that the type and stability of the fracture is important to still assess metabolic issues even
6 A. Agarwal, MD
in light of inadequate fixation as many patients fixation techniques may need to be employed to
still have some deficiencies in bone metabolism obtain improved fixation by the judicious use of
[13]. locked, fixed angle, or load-sharing devices such
Patient factors contributing to mechanical as intramedullary nails when appropriate.
problems can be related to noncompliance with Patient medical factors contributing to a bio-
weight-bearing restrictions or an error in allow- logical cause for the nonunion are many and can
ing the patient to weight bear too early. The be problematic not only from the original fracture
healing process is always a race between hard- standpoint but also for the treatment of an
ware failure and fracture healing, and thus, when established nonunion [10, 13, 25, 32, 40].
patients present with a nonunion in conjunction Established diseases such as vascular disease,
with hardware failure, the time from the original rheumatologic disease, and s/p organ transplan-
surgery is important in determining what came tation cannot be affected, but their effects on
first—the hardware failure or nonunion, as each fracture healing and subsequent management of
one can lead to the other. Often times, with plate the nonunion need to be taken into consideration.
failure there is an associated deformity through Perhaps their steroids or immunosuppressive
the nonunion site (Fig. 1.1). In cases of early agents can be held for short time period which
hardware failure, often times the patient has would allow for surgical intervention and heal-
started weight bearing too early or was allowed ing, and such decisions should be made in con-
to do so. This is more common in cases of plate junction with the patients’ appropriate other
fixation. In these situations, the fracture has not physicians. A multidisciplinary approach is nee-
healed sufficiently to handle the body weight and ded to get many of these patients healed.
the implant is taking all the stress leading to early Although there are many endocrine abnor-
failure. Failure can be in the form of screw malities that can affect the musculoskeletal sys-
loosening, implant breakage, or bending. tem, such as thyroid and parathyroid disorders,
Depending on the fracture pattern and amount of hypogonadism, and calcium imbalances to name
comminution as well as the location, it may still a few [13], diabetes has had the most attention
unite. In the lower extremity more so than the due to the high prevalence in the population.
upper, the alignment may gradually worsen as Diabetes has been shown to prolong healing
stability is lost and a mal-aligned nonunion can times for fractures [40, 41]. It is also well doc-
develop. In some instances, especially where umented that patients with diabetes have
there is comminution, as the angulation worsens increased complications when dealing with
resulting in more bony contact, the fracture may musculoskeletal conditions, especially with
unite resulting in a malunion. In late cases of fractures [32, 42, 43]. In a nationwide population
hardware failure, the fracture may have healed based study out of Taiwan, diabetics were found
sufficiently to handle some weight in addition to to have an increased incidence of fractures as
the implant and may have maintained the align- well as more adverse events and a higher mor-
ment. After a while, the implant undergoes fati- tality after fractures [42]. The addition of neu-
gue failure as the micromotion from the loading ropathic complications can make even simple
leads to failure of the implant at a stress riser fractures that require surgery end up being dis-
such as a hole in the plate. The alignment is often astrous for the patient. Wukich et al. [43] showed
times maintained, but the patient has pain and that patients with ankle fractures that had com-
discomfort which necessitates surgical interven- plicated diabetes had a 3.8 times increased risk of
tion. Loss of fixation can also occur without overall complications and a 3.4 times increased
weight-bearing issues. This is often the case in risk of malunion and nonunion compared to
patients with poor bone quality such as in those uncomplicated diabetic patients. These patients
with comorbidities such as diabetes or osteo- were also 5 times more likely to require revision
porosis. It is important to know whether patients surgery or arthrodesis. Diabetics need to under-
have these conditions as special surgical and stand that glucose control is extremely important
Fig. 1.1 Patient with right

ankle injury treated with open
reduction and internal fixation
of fibula and closed treatment
of distal tibia fracture.
Referred for nonunion after
progressive deformity
developed. a–c Three views
(anteroposterior [AP],
mortise, lateral) of the right
ankle show failure of the
fibula hardware and
mal-alignment with nonunion
of both the tibia and fibula.
Patient underwent hardware
removal and cultures. d–
f Three views (AP, lateral,
and mortise) of the right ankle
after hardware removal. Due
to the malalignment and stiff
nonunion, a Taylor spatial
frame (TSF) was applied to
allow correction and healing
of the nonunion. g–h AP and
lateral after TSF applied to
right ankle prior to correction.
i–j AP and lateral with TSF
showing full correction of the
deformity and realignment of
the limb. k–m Three views
(AP, lateral, and mortise) of
the right ankle 1 year after
consolidation of nonunion
and removal of TSF
8 A. Agarwal, MD
for them to avoid diabetic complications of end supplementation in all fracture patients, the cost
organ damage, neuropathy, nephropathy, and of an 8-week course of treatment was determined
peripheral arterial disease to minimize further and compared to the cost savings assuming just a
musculoskeletal complications [32]. Diabetics 5% reduction in nonunions. This would result in
should be treated with prolonged immobilization a potential cost savings of $65,866 annually [47].
and delayed weight bearing compared to the Many dosages of replacement therapy are avail-
nondiabetic to aid in avoiding complications. able, but the authors’ preference is for high-dose
Additionally, many of these patients require (50,000 IU) vitamin D weekly for six months
additional fixation for otherwise straightforward along with calcium supplementation. The target
fractures to try and prevent the late complications is to obtain a 25-OH level in the 40–60 range.
that occur with these injuries. Patients with low vitamin D can also develop
Vitamin D deficiency or insufficiency has secondary hyperparathyroidism and should also
been linked to nonunions, but a clear causal link have a parathyroid hormone (PTH) level drawn
is difficult to establish [13, 40]. Both the 25-OH when evaluating for a nonunion. The high PTH
vitamin D and 1, 25 OH2 vitamin D levels can be can contribute to the development of a nonunion
monitored, but the 25-OH level is the one that [13]. In most cases, the high PTH will resolve
is important. Patients with 25-OH levels <20 are with appropriate vitamin D replacement therapy.
considered insufficient and between 20 and 30 Osteoporosis has also been linked to the de-
deficient. It is not clear however whether higher velopment of nonunions [32]. The issues with
levels than simply above the 30 level are needed osteoporotic bone healing are both biologic and
in patients with fractures. Brinker et al. [13] mechanical [48]. By definition, osteoporotic
showed that a preponderance of their nonunion bone is bone with less bone mass and as such is
patients had vitamin D deficiency. They had 37 at an increased risk for fracture. The biologic
patients that were evaluated for a metabolic or changes that occur with osteoporosis, including a
endocrine abnormality of which 68% (25 of 37) diminished level of mesenchymal stem cells and
had vitamin D deficiency. It has become thus osteoblasts, a decrease in the chondrogenic
increasingly clear that many patients are vitamin potential of the periosteum and other alterations
D deficient or insufficient. In a meta-analysis of in the fracture healing pathway results in a less
the literature, it was found that the pooled than robust fracture healing process [32, 40, 48].
prevalence of hypovitaminosis was 77.5% in Additionally, because of the lower bone mass,
young trauma patients and 73% in geriatric fra- the fixation in such bone can be problematic and
gility fracture patients [44]. In a follow-up study, as such can lead to inadequate fixation and
the same authors showed that there is a lack of fracture stability. The result can be a nonunion.
consensus in prescribing vitamin D to fracture Many specialized techniques have been descri-
patients. They found that 66% of surgeons ten- bed in the management of osteoporotic fractures
ded to prescribe vitamin D to fragility fracture and should be employed when dealing with
patients compared to 25.7% to nonfragility nonunions especially if mechanical failure was a
fracture patients [45]. The lack of prescribing in significant contributor to the development of the
this population needs to be re-examined since the nonunion. Locked plating, use of load-sharing
prevalence of low vitamin D in young trauma devices, use of fixed angle devices, augmentation
patients is high. Low vitamin D is more prevalent of fixation with cement or bone graft substitutes,
than previously thought and is widespread in adjunctive use of structural bone grafts, and
patients of all orthopedic subspecialties and not preservation of soft tissue can assist in the
just orthopedic trauma [46]. Management of management of these fractures and nonunions
vitamin D is easily done via replacement therapy [48]. Although most osteoporotic individuals are
and has been shown to be successful in raising elderly, age is an independent factor which can
serum levels [44]. In a study to evaluate the cost negatively affect fracture healing also resulting in
benefit of both calcium and vitamin D delayed unions or nonunions [49]. This decline
in healing potential can be attributed to hormonal hard to make a recommendation on the use of
changes, changes at the cellular level of fracture NSAIDs both in terms of timing and duration
healing signaling, and diminished mesenchymal immediately after a fracture. Due to the lack of
stem cells which all may also occur with osteo- guidelines and unknown true effects on fracture
porosis. The true etiology still requires much healing, the author’s practice is to avoid
more investigation due to the complex interplay NSAIDS for the first 4–6 weeks after a fracture.
that occurs in fracture healing and the overlap in This is especially true of Indomethacin. Addi-
physiology with aging and osteoporosis. Another tionally, we do not use Toradol (intravenous or
confounding factor is that patients with osteo- per os) intra-operatively or immediately postop-
porosis are often being pharmacologically treated eratively for acute fracture cases. NSAID use
for their osteoporosis. The most common are the after repair of nonunions has not been investi-
bisphosphonates, which are anti-resorptive gated to our knowledge.
agents and inhibit osteoclast function. The Since inflammation is one of the initiating
interference with remodeling of the bone has factors for bone healing, it has been suggested
resulted in an unwanted side effect and resultant that perhaps healing may be altered in the poly-
“atypical” femoral fractures. It is advised that trauma patient as well [32, 37, 40, 50]. These
these medications be discontinued during the patients undergo a prolonged state of inflamma-
fracture healing process [32, 48]. It is also tion [40]. It is thought that the increased
important to evaluate all their medications and inflammation could delay fracture healing
the potential effects that they may have on bone through a variety of cellular responses [20, 50].
metabolism. Additionally, many of these patients also have
One of the most common class of medications multiple fractures that may require operative
that many patients take, both prescription and intervention. The post-op rehab protocol on one
over the counter, are the nonsteroidal fracture may result in delayed stimulation of
anti-inflammatory drugs (NSAIDs). Recently, the another fracture with resultant delayed healing or
use of NSAIDs during fracture healing has come even nonunion. Other system injuries also may
under intense scrutiny [49–51]. Early reports of have an effect on fracture healing as well. A lit-
the use of NSAIDs in animal fracture healing erature review by Hildebrand et al. [37] found
models showed a clear deleterious effect [32, 49– that isolated hemorrhagic shock, chest trauma,
51]. The doses required were very high. The severe soft tissue injury, and systemic inflam-
mechanisms by which they are theorized to mation can all affect fracture healing. Finally, it
inhibit fracture healing include inhibition of has been suggested that the American Society of
prostaglandin synthesis and reduction of osteo- Anesthesiologist (ASA) classification, which
blast activity, both of which result in an impaired indicates overall health, was associated with
fracture healing response [32]. Prostaglandins are nonunion development—the higher the ASA, the
needed during the inflammatory phase of fracture increase in probability of a nonunion [23].
healing and help start the osteogenic response [6, Smoking has been clearly shown to inhibit
31, 38, 49]. Although a few clinical studies have fracture healing and result in both delayed unions
shown a loose association between the use of and nonunions as well as increase the overall
NSAIDs and nonunions, it is controversial [50]. complications in the management of fractures [10,
Kurmis et al. [51] performed a systematic anal- 21, 24, 28, 29, 32, 46, 49, 52–54]. Smoking has
ysis of over 300 relevant papers and concluded also been linked to an increase in fracture rates
that there was not significant evidence to indicate of the hip, distal radius, spine, and other osteo-
a negative effect on fracture healing from the porotic fractures [52]. The exact mechanism and
short-term use of NSAIDs after a fracture. Most offending agent has not been clearly elucidated.
of the clinical studies published in relation to Nicotine is one of more than 4000 chemicals that
NSAIDs in fracture healing were Level 5 evi- exist in cigarette smoke [46]. It has been shown to
dence or expert opinion only [28]. Therefore, it is cause vasoconstriction (resultant hypoxia),
10 A. Agarwal, MD
platelet adhesion, and reduced cell proliferation The first step is acknowledging that smoking is
for healing. All of these physiologic changes bad for one’s health. Matuszewski et al. [55]
result in a negative effect on both wound and performed a cross-sectional cohort survey study
fracture healing [52]. However, it is not clear and found that smokers did not understand the
exactly which chemical is responsible for all the negative effects of smoking on their general
negative effects. Animal studies have had health or on fracture care. On a positive note, the
conflicting results with some studies, with nico- orthopedic trauma patients surveyed seemed
tine alone, not showing the negative effects that interested in smoking cessation more so than
are seen with smoking, whereas others have what was expected. They recommended formal
shown deleterious effects [49]. Nevertheless, the education for smoking cessation. It is well
clinical literature overwhelmingly supports the accepted and has been shown that preoperative
increased risk of delayed unions, nonunions, and smoking cessation can reduce both pulmonary
wound complications seen with smoking [21, 24, and wound complications postoperatively [46].
52–54]. Scolaro et al. [54], in a systematic review Educating the patients on the ill effects of smok-
of the literature, showed that smokers were over 2 ing on fracture healing is part of our “discussion”
times more likely (statistically significant) to with the patient being evaluated for nonunions. It
develop a nonunion than nonsmokers. This was is the author’s policy to not perform nonunion
especially true in open fractures and tibia frac- surgery on active smokers as long as the man-
tures. There was also a trend for longer healing agement can be done on an elective basis (aseptic
times and infections (deep and superficial) in the nonunions). Both serum and urinary levels of
smoking group. In a separate systematic review of cotine and nicotine are monitored to insure patient
the literature, Patel et al. [53] also found a nega- compliance. Although many feel that smoking
tive effect of smoking on bone healing. They also cessation is the primary care physician’s respon-
looked at each study in relation to the bone or sibility, as an orthopedist it behooves us to play
procedure in question. All the tibia fracture an active role to help maximize the patient’s
studies, except for one treated with external fix- outcome and minimize complications from any
ation, showed a clear increased risk of nonunion surgical intervention.
from smoking. This was also true in distraction When evaluating a patient for a nonunion
osteogenesis, fibula fractures, ulna osteotomy management, one must assess for the presence of
healing, subtalar and ankle arthrodesis, and the risk factors above. There are certainly more
elective foot surgery. Fractures of the femoral comorbidities than can affect fracture healing, but
diaphysis were not statistically significantly these are the most prevalent. These risk factors
affected by smoking. In contrast, Hernigou and and/or co-morbidities should be improved upon
Schuind [21], in their retrospective study looking or corrected if feasible. Many are injury or
at diaphyseal fractures, found that smoking was treatment related, but knowing those details can
significantly associated with nonunions (OR help devise an appropriate treatment plan for the
8.25) in the femur, as well as the tibia and nonunion.
humerus. Westgeest et al. [29] found that in a
prospective cohort study of open long-bone
fractures, smoking (OR 1.73) was significantly 1.4 General Principles
associated with developing a nonunion. Murray
et al. [24] looked at their series of diaphyseal 1.4.1 Diagnosis
clavicular fractures. They found that smoking was
the strongest predictor of a nonunion (OR 3.76) The diagnosis of a nonunion is highly contro-
and recommended that smoking cessation be an versial because no gold standard exists for heal-
integral part of any treatment. However, getting ing assessment [6, 15]. In a multinational survey
patients to stop smoking is extremely difficult. of orthopedic surgeons, there was a 73 and 53%
consensus that a lack of standardization in the found to be the number of cortices bridged by
definition for a delayed union and nonunion, callus.
respectively, existed. However, they did agree Despite the issues with radiological criteria,
(88%) that the diagnosis should be done based on standard orthogonal views (anteroposterior and
clinical evaluation and plain radiographs [3]. lateral) of the bone in question should be
Pain on weight bearing was felt to be the most obtained. If the patient is referred in, previous
consistent predictor of delayed union and studies are desired for comparison. In some
nonunion. cases, the fracture is actually progressing and
The diagnosis of a nonunion should be made reassurance is all that is needed. They may have
on a series of radiographs in addition to the a delayed union, but radiographic evidence of
clinical picture. Often the fracture healing may be healing is occurring. The length, alignment, and
delayed, but critical evaluation of radiographs 6– rotation of the limb should be appropriately
8 weeks apart may show some improvement evaluated. In the lower extremity, if there is an
indicating progress. If the X-rays show no pro- associated deformity, then additional full-length
gress on two sets of consecutive images and the radiographs (± ruler) from the hip to the knee are
patient is having pain, then nonunion has prob- obtained to assess the mechanical axis of the
ably been established assuming sufficient time limb (Fig. 1.2). Restoring the mechanical axis of
has initially passed. The problem arises in the the limb can aid in healing of the nonunion and
patient without symptoms but clear radiographic should be part of the preoperative plan. If it
evidence of a nonunion. Many of these patients, appears to be short, then a scannogram (Fig. 1.3)
because they lack symptoms, may not return in or full-length radiographs with a ruler should be
fear of needing surgery. The problem occurs obtained (see Fig. 1.2). Oblique radiographs can
when they return after hardware failure with aid in the diagnosis as well, if the standard an-
new-onset pain and/or deformity. The time pas- teroposterior and lateral do not clearly show the
sed based on the FDA definition may not have nonunion due to the obliquity of the original
been reached, but if cessation of all healing is fracture or because of overlying hardware (in-
indicated by plain radiographs and the patient is ternal or external fixation). Such views can also
symptomatic, then intervention is probably better define the plane of maximum deformity
warranted. when that plane is not in the usual sagittal or
coronal plane. Rotation can be assessed clinically
in some situations; otherwise, a CT scan may be
1.4.2 Radiographic Evaluation needed (Fig. 1.4).
and Scoring When looking at the plain radiographs, the
absence of bridging bone or callous at the frac-
After the history and physical, evaluation should ture site, sclerotic fracture edges, bone resorp-
always begin with plain radiographs. It still tion, or persistent fracture lines all may indicate a
remains the most common method of assessing nonunion. It is imperative to also critically assess
for fracture union. However, just as in the lack of the implants and the initial fixation strategy to
standardization of definitions, there is a lack of insure that the original type of healing wanted—
consensus on radiographic criteria as well. primary versus secondary—was being achieved.
Dijkman et al. [56] reviewed the literature to look Often times absolute stability was desired, yet
at radiographic criteria used in studies. They there is callus formation on the radiographs
found that bridging of the fracture by bone, cal- (Fig. 1.5). This may indicate either excessive
lus, or trabeculae was used 53% of the time. motion suggesting hardware failure or that the
Bridging of the fracture across three cortices 27% fixation was not as rigid as one wanted, allow-
of the time and loss of fracture lines was 18% of ing sufficient motion for callous formation. The
the time. The best interobserver reliability was fracture however may go on to heal. In other
situations, it may go on to a nonunion with or
12 A. Agarwal, MD
Fig. 1.2 Patient with a long

leg film obtained for
evaluation of his mechanical
axis with a left femoral neck
nonunion. A ruler can be used
also to evaluate for leg length
discrepancy
without hardware loosening or breakage. Addi- visible fracture line. A two is the presence of
tionally, the radiographs should be assessed for callus, but the fracture line is still visible. A three
periosteal reaction, loosening/lysis around hard- is for callus and the absence of a fracture line.
ware, and broken implants. Comparison to pre- The minimum score is 4, and the maximum is 12.
vious radiographs cannot be overemphasized. This has shown to improve agreement for
As mentioned above, plain radiography alone assessing union only in tibia fractures treated
is often times not a reliable tool for assessing with intramedullary nails. Whelan et al. [59]
fracture healing due to the lack of consistency showed an overall inter-observer reliability of
among surgeons and interpretation of the films. It 86% and intra-observer reliability of 88%.
is clear that better ways of assessing fracture The RUST score has not been correlated with
healing are needed [6]. Several clinical trials all functional outcomes to date.
have shown poor agreement between surgeons A similar scoring system was developed by
[15, 56]. Many have proposed criteria to stan- Bhandari et al. [57] for use in hip fractures. The
dardize fracture healing assessment [57, 58]. One Radiographic Union Score in Hip Fractures or
such assessment tool is the Radiographic Union RUSH was developed to improve agreement in
Scale for Tibial (RUST) developed by Koolstra the assessment of femoral neck fractures. In a
and his colleagues [58]. This scoring system similar manner to the RUST, the RUSH evalu-
assesses the presence or absence of fracture cal- ates cortical bridging on each of the four cortices
lus and the visibility of the fracture line on each as well as disappearance of the fracture line and
of the four cortices. The scale is from 1 to 3 and an independent score is given. A one is given for
based on callus and fracture line visibility at each no cortical bridging, two for some cortical
cortex. A one is the absence of callus and a bridging, and a three for complete cortical
Fig. 1.3 Patient with a right

ankle injury and delayed
presentation. Scannogram
was obtained to evaluate the
amount of leg length
discrepancy
bridging. If the fracture line is visible, a one is et al. [60] extended the RUSH score to evaluate
given, a two for some evidence of the fracture intertrochanteric hip fractures and evaluate
line, and a three for no evidence of the fracture agreement between radiologists and orthopedic
line. Two other aspects of femoral neck fractures surgeons. They found that the RUSH score did
are scored, the trabecular index based on con- improve the overall agreement regarding fracture
solidation and the disappearance of the fracture healing from fair to substantial between the two
line. A score of 1–3 is assigned as well to each specialties.
component. The overall minimum is 10 and Although scoring systems can be beneficial in
maximum is 30. Their initial study showed that determining union and providing a more objec-
the RUSH improved agreement among reviewers tive measurement over time, the real issue is their
regardless of subspecialty, but their agreement use in predicting a nonunion. Recently, Frank
did not improve over time. A very important et al. [19] did a study to assess the utility of the
shortcoming was that the reviewer’s assessment RUSH score to help define femoral neck fracture
was found to be potentially inaccurate without nonunion. They retrospectively pulled 250 cases
information regarding the time of the radiograph. from the FAITH hip fracture trial all of which
They had 6 of 7 patients deemed as being healed had 6-month hip radiographs. They determined
at 2 weeks, which is not possible. Chiavaras the RUSH score at 6 months for each case. They
14 A. Agarwal, MD
Fig. 1.4 (a–c. Computed

tomography scan images
showing how to measure
malrotation of left tibia—24°
internal rotation compared to
right side
found that if the RUSH score at 6 months neglected fractures. They evaluated the present-
was <18, it had 100% specificity and a positive ing nonunion film for a radiographic index they
predictive value of 100% for a nonunion. They called the neck resorption ratio (NRR) to deter-
all had a 10 times greater risk of undergoing mine whether that could predict nonunion of the
reoperation for a nonunion. If the patient does valgus intertrochanteric osteotomy. The NRR is
develop a nonunion of the femoral neck, a valgus determined by measuring the length of the frac-
intertrochanteric osteotomy is an option to obtain tured head and neck fragment and comparing it
union. Varghese et al. [61] evaluated a group of to the length of the intact neck on the contralat-
40 patients who underwent the procedure for a eral side (measured from the tip of the head to the
femoral neck nonunion developing after intertrochanteric line). The NRR was found to be
Fig. 1.5 a, b. Injury radiographs (anteroposterior [AP] Follow-up radiographs (AP and lateral) showing unin-
and lateral) of patient with left humerus fracture after a tended callus formation due to micromotion despite
motor vehicle collision. c, d Postoperative radiographs attempt at rigid fixation—infection workup was negative
(AP and lateral) after open reduction and internal fixation and patient went on to consolidate
performed in an effort to obtain absolute stability. e, f
the most important factor in predicting union in they developed a logistic regression model to
their series. All patients that had a pre-op NRR include seven of these factors. They assigned
of >0.52 had union. Taking this parameter into points to these seven factors. The NURD score
consideration before making treatment decisions gave 1 point for male gender, 2 points for open
in femoral neck nonunions may allow one to fractures, 3 points for chronic conditions, 4
consider a more definitive treatment and avoid a points for compartment syndrome, and 5 points
repeat nonunion situation. for flaps. Additionally, 1 point per ASA grade
Although utilizing a score to predict nonunion was given as well as for each 25% reduction of
after a reconstructive procedure can be useful, a cortical contact (100% = 0; 75% = 1; 50% = 2;
score to predict nonunions for acute fractures 25% = 3). If the injury was low energy or spiral,
would have greater applicability. The Nonunion one point was subtracted for each factor. They
Risk Determination (NURD) Score was devel- found that a NURD score of 0–5 had a 2%
oped by O’Halloran et al. [26]. The authors ret- chance of nonunion versus a 61% chance if the
rospectively reviewed all tibial shaft fractures at score was >12. The score was felt to be a
their institution over a 7-year period treated with potential nonunion prediction model that clini-
an intramedullary nail. They had 382 patients cians could utilize to determine which patients
with 56 nonunions. Factors were evaluated and had a higher risk of nonunion. If such scores
16 A. Agarwal, MD
could be developed and validated for other additional cost. Three-dimensional ultrasound is
bones, prediction of nonunions could be com- a newer technology that may have added benefits
monplace and allow for earlier intervention. of being able to measure the vascularity not only
In our practice, comparative plain radiographs in the surrounding soft tissue but the fracture
over time, and the clinical picture and evaluation itself, as well as providing more information on
of the patient are sufficient to diagnose a non- the progression of healing [63].
union. However, in some situations, plain radio- Fluoroscopy is another imaging modality that
graphs may not allow complete evaluation of the can be used primarily to assess motion at a
nonunion site because of the hardware. In these fracture site to determine healing. This is most
cases, a CT scan, with metal suppression if hard- useful in the patient treated without internal fix-
ware is present, can be obtained to further evaluate ation and when there is a question of the healed
the nonunion site as well as look for areas of status of the injury. It can also be useful in cases
sequestered dead bone or areas of bone deficits where external fixation has been used since the
that may require bone grafting. CT scans have external fixation can be loosened without com-
been shown to have high sensitivity but moderate plete removal and the fracture site stressed. If
specificity with about a 90% accuracy for the there is motion, the external fixation can easily be
detection of nonunions [62]. Sagittal and coronal “tightened” and “reset.” In our practice, this is
reconstructions to include 3D reconstructions can usually done in conjunction with anticipated
help with visualization (Fig. 1.6). Many times, external fixation removal after definitive man-
these fractures are “clinically” healed, but patients agement of a fracture or in reconstructive cases,
have symptomatic hardware. A CT scan can also where determining the “laxity” of a nonunion can
aid in looking at the integrity of the bony con- guide treatment.
solidation and for defects within the “healed” Magnetic resonance imaging (MRI) with
construct. In some situations, the patient can be gadolinium can be used to assess the nonunion
considered as having an implant dependent union; site for infection and more importantly for
e.g., there is some central bone loss but sufficient devascularized bone or a sequestrum [31, 65].
bridging that the bone has healed around these Additionally, because of its ability to detect
deficiencies, but the strength of the bone may be marrow changes, it is very sensitive for
reliant upon the associated hardware. osteomyelitis. Osteomyelitis usually shows
Ultrasonography (US) has been shown to decreased marrow signal on T-1 images but
have some utility in diagnosing nonunions [63]. increased signal on T-2 images (Fig. 1.7a–d).
In a study by Moed et al. [64] in which tibia The MRI also allows one to determine the extent
fractures treated with an intramedullary nail were of bony involvement [31] in such cases because
evaluated, the authors showed a sensitivity of of the marrow changes which is crucial in
100% and a positive predictive value of 97% in determining the best reconstructive option based
detecting healing of the fracture site. They also upon the anticipated length of resection required
could predict healing of these injuries much to eradicate the osteomyelitis.
earlier (38 days versus 127 days) than plain Nuclear medicine studies (Fig. 1.7e) have
radiography. Chachan et al. [14] in their been historically used to aid in the detection of
prospective diagnostic follow-up study showed infection as well, but over time their utility has
that ultrasound was able to predict fracture been questioned [66]. They are still of use in
healing 2 weeks earlier than plain radiographs. evaluating the nonunion site for infection and/or
More importantly, it was able to predict nonbiologic activity [31]. Leukocyte-labeled studies
unions 8.5 weeks earlier. Despite the earlier have been shown to have appropriate diagnostic
detection for a nonunion, US has not become accuracy for osteomyelitis in the peripheral
widespread in its use. The benefits of no radia- skeleton [67]. The traditional technetium bone
tion have not outweighed the primary issues of scan will have increased signal on any biological
user dependency, time required for the study and bone activity, and hence, any fracture site that is
Fig. 1.6 Patient referred for nonunion 9 months after f Computed tomography scan images (axial, coronal,
treatment for tibial plateau and tibia shaft fracture treated sagittal, and 3D reconstructions) which show the subtle
with open reduction and internal fixation. a, b Antero- hypertrophic nature of the nonunion. g, h One year after
posterior (AP) and lateral radiographs of the tibia show treatment of hypertrophic nonunion with hardware
consolidation of the plateau component. There is hard- removal and subsequent reamed nailing (AP and lateral)
ware failure and nonunion of the tibial shaft. c–
biologically active should have uptake. Thus, it is usually for the suspected infected cases when
really is not used for the evaluation of healing the clinical signs of an infection are absent but
although, in cases of avascular or nonviable laboratory markers—erythrocyte sedimentation
fractures sites, e.g., the atrophic nonunion, rate (ESR), C-reactive protein (CRP), or white
decreased or no uptake may be the case. Our use blood cell (WBC) count—may be elevated. In
18 A. Agarwal, MD
Fig. 1.7 a, b Anteroposterior and lateral radiographs of signal on T2 image indicating osteomyelitis. e Nuclear
patient with infected nonunion of right tibia. c, d medicine studies showing increased uptake on indium
Magnetic resonnance images of tibia showing increased study suggesting the presence of infection
these cases, a bone scan is obtained which is suspicion for infection. If it is negative for uptake
usually positive. If by chance it is negative for at the site, then infection is less likely but
uptake, then no other imaging is done and con- unfortunately never completely ruled out. The
cern becomes for an atrophic nonunion. After a final study done after a positive Indium scan, is
positive bone scan, an indium (tagged WBC) the sulfa colloid marrow scan. The areas of
scan is performed. If this is positive at the uptake are then compared to the indium scan. If
fracture/nonunion site, then there is increased the areas of uptake are concordant with the
indium scan, the uptake is deemed to be sec- As mentioned before, in cases where the orig-
ondary to the associated marrow changes and not inal surgery was deemed to be highly contributory
infection. Conversely, if the areas of uptake on to the development of the nonunion, more exten-
the indium scan do not coincide with uptake on sive laboratory studies may not be needed. In the
the sulfa colloid (discordant), then it is thought to cases where the technical aspects seemed to be
be suggestive of an infection [31]. The specificity sound and the reason for the nonunion unclear,
and sensitivity of such imaging studies has been other laboratory studies may point to an underly-
controversial. Stucken et al. [66] showed that not ing metabolic abnormality as the etiology [13].
utilizing the nuclear medicine tests actually These patients would probably benefit from an
improved their predicted probabilities of infec- endocrinology workup if feasible. Often times in
tion based on laboratory studies alone. The latest our practice, these are unfunded trauma patients
imaging modality, which has shown some pro- and the workup is often left to the orthopedic
mise to aid in the detection of infection or os- trauma surgeon to do the full evaluation. Many of
teomyelitis, has been the positron emission these patients may also have sustained fragility
tomography (PET) scan ± CT scan. A fractures which also warrant laboratory
fluorodeoxyglucose PET scan has been shown to workup. These underlying metabolic disorders
have the highest diagnostic accuracy for include vitamin D deficiency, hypothyroidism,
excluding or confirming the diagnosis of chronic hypogonadism, hypocalcemia, and overall poor
osteomyelitis [31, 67]. This could aid in the nutritional status. Brinker et al. [13] showed that
evaluation of the presence of infection in a 31 of 37 of their patients with a nonunion had some
nonunion. type of metabolic abnormality with vitamin D
deficiency being the most common. The labora-
tory studies, in addition to the above, should
1.4.3 Laboratory Evaluation include serum 25-hydroxy-vitamin D, calcium,
phosphorus, alkaline phosphatase, thyroid func-
Laboratory studies can assist in determining the tion tests, parathyroid hormone level, hormone
etiology of the nonunion or at least look at levels (testosterone, estrogen, and follicle stimu-
conditions that may have contributed to the lating hormone), and albumin and cortisol levels
development of the nonunion. All patients should [13]. Vitamin D deficiency has been set at 20–
be evaluated with a CBC with differential, ESR, 30 mg/dl, and <20 are considered insufficient.
and CRP. These are utilized to evaluate for In cases of infected nonunions, it is helpful to
infection but realizing that the ESR and CRP are obtain results of previous cultures if available to
simply indicators of inflammation and can be determine the previous organism(s). At the time of
elevated in the absence of an infected nonunion. surgery, especially in cases of staged procedures,
Conversely, normal markers do not necessarily which is often the situation in dealing with infected
rule out an infection either and are usually the nonunions [66], deep tissue cultures and bone
case in indolent infections. A standardized pro- biopsies can help determine the presence or
tocol to rule out infection was assessed by absence of an infection as well as the offending
Stucken et al. [66] to evaluate the efficacy of organism. Preoperative antibiotics should be with-
laboratory studies (WBC, CRP, ESR) and held until after intra-operative cultures are obtained.
nuclear medicine studies. They found that the It is also recommended to cease any antibiotics for
ESR and the CRP were both independently at least two weeks, if possible, to maximize the
accurate predictors of infection. With all three chance of identifying the organism. The first stage
tests being positive, the predicted probability of is usually to remove previous hardware, to obtain a
an infection was 100%. If the nuclear medicine better idea of the nonunion site, and to get biopsies
studies were included, the probability went down and cultures. Antibiotic beads can be placed in the
to 86% for three positive tests. interim prior to the second stage.
20 A. Agarwal, MD
There has been an increasing interest in compared them to a group with normal bone
looking for serologic markers that may help to healing. They found temporal variations of these
predict fracture healing and therefore potentially cytokines in the three groups, with high expres-
predict nonunions [16, 20, 22, 27, 30, 68]. sions of IGF-1 corresponding to a successful
Although a full review and discussion of these Masquelet treatment. They demonstrated signif-
markers is beyond the scope of this chapter, it is icant differences in cytokine expression between
important to mention that they exist and have normal fracture healing and the nonunion treat-
future implications in predicting fracture healing. ment groups. If the time profiles of each of these
These biomarkers are either factors that regulate markers can be fully understood, then perhaps
the healing process itself or bone turnover variations in these markers from what may be
markers that are extracellular matrix components considered the normal in fracture healing may
related to degradation or production during the provide insight into which fractures will go on to
repair process [15]. The local or systemic factors a nonunion [27]. Earlier detection and subse-
regulating the healing process include vascular quent earlier treatment could result in substantial
endothelial growth factor (VEGF) and trans- cost savings [20].
forming growth factor-beta (TGF-b). Serum
TGF-b has been found to be an indicator of
healing versus nonhealing with 4 week levels 1.5 Definitions and Classification
being much lower in a group of patients that had
a delayed union [20]. The bone turnover markers As mentioned previously, the US FDA defined a
can be divided into one of three categories: 1. nonunion as a fracture that is at least nine months
bone formation markers, 2. bone resorption old and has not shown any signs of healing
markers, and 3. osteoclast regulatory proteins progression for at least three consecutive months
[16, 20, 22]. The bone formation markers indi- [1]. This however cannot be applied to every
cate osteoblastic activity and as such are frag- fracture, and all nonunions are not the same.
ments of type-I and type-III pro-collagen that are Harwood and Ferguson [31] proposed more
released during the formation of type-III collagen sensible definitions. They suggested that a non-
(PIIINP, PICP, PIIINP). Osteocalcin (OC) and union be defined as “a symptomatic fracture with
bone-specific alkaline phosphatase (BSAP) are no potential to heal without intervention.” A
also measures of osteoblastic activity. Bone delayed union was defined as “a fracture in
resorption markers include those that measure the which healing has not occurred in the expected
degradation of type-I collagen (CTX, NTX, time and the outcome remains uncertain.”
ICTP, pyridinoline, deoxypyridinoline). The most common classification was original
Tartrate-resistant acid phosphatase (TRAcP) and described by Weber and Cech [69] in 1976 and
cathepsin K (CK) are noncollagenous markers has survived for 40 years. It was based on the
that also measure bone resorption but are osteo- viability and healing potential of the nonunion.
clast regulatory proteins. Other osteoclast regu- From a vascular viewpoint, that corresponds to
latory proteins include receptor activator of either a hypervascular or avascular environment
nuclear factor NF-kB ligand (RANKL) and os- [2, 31, 65, 69, 70]. This is based on the appear-
teoprotegerin (OPG). The marker activity of only ance of the fracture site on plain radiographs after
a handful of these have been evaluated in various a period of time when improvement in the frac-
fractures and shown some promise in predicting ture healing has ceased. The hypervascular non-
fracture healing [30]. Fischer et al. [18] evaluated unions have been further subdivided into a
a number of cytokines—TGF-b, platelet-derived descriptive classification as an “elephant foot,”
growth factor (PDGF-AB), insulin-like growth “horse hoof/foot,” and oligotrophic nonunion.
factor 1 (IGF-1)—in patients with long-bone The avascular nonunions have been further sub-
nonunions treated both successfully and unsuc- divided into the torsion wedge, comminuted,
cessfully with the Masquelet technique and defect, or atrophic nonunion. In addition, the
Fig. 1.8 a, b Anteroposterior and lateral radiographs of a patient with a low energy left tibia fracture treated with cast
immobilization that went onto a hypertrophic (“elephant foot”) nonunion
pseudarthrosis has been described [2, 31, 69]. site. Revision of the fixation is dependent upon
Any of these types of nonunions can be aseptic the integrity of the hardware and need for cortical
or septic. If septic, the infection has to be erad- apposition. All three of these nonunions gener-
icated and any osteomyelitis addressed usually ally require revision fixation with the aim of
with bone resection. Additionally, these may or improving stability. Bone grafts and other bio-
may not have a deformity that is associated with logic adjuncts are not needed except possibly in
it, and if present, any management needs to the case of the oligotrophic nonunion [70].
address the malalignment. All of the avascular nonunion subtypes can be
The hypervascular “elephant foot” nonunion considered as having atrophic ends as all are
(Fig. 1.8) is based on the appearance of the bone deficient in callus formation, have undergone
ends. These hypertrophic nonunions exhibit some resorption, or have significant bone loss at
abundant callus formation and are due to exces- the time of injury [2, 31, 65, 69, 70] (Fig. 1.11).
sive motion at the fracture site from inadequate These generally require a biologic stimulus to
stability [31, 65]. The motion precludes union of heal the nonunion with varying degrees of fixa-
the fracture ends. These are well vascularized tion (or revision fixation) and/or soft tissue
and generally do not require a bone graft. These reconstruction [70]. If the hardware placed
require enhanced mechanical stability, which appears to be intact and appropriate, then a bio-
may involve revision of the hardware or addi- logic stimulus may be all that is needed. This is
tional fixation. The “horse hoof/foot” nonunion usually in the form of autogenous bone grafting
is also hypertrophic but much less so. It usually although various bone graft substitutes have been
occurs in a situation of inadequate or unstable used. Other adjunctive treatments have also been
plate fixation constructs but can occur with nails described [70] and will be discussed later.
[2] (see Figs. 1.6 and 1.9). The “oligotrophic” A pseudarthrosis is a nonunion that chroni-
nonunion albeit hypervascular is not hyper- cally develops into a joint-like appearance with a
trophic on radiographic appearance (Fig. 1.10). hypertrophic callus or can be atrophic on radio-
The callus is absent, and some absorption occurs graphs with gross motion [2] (Fig. 1.12). Despite
but the ends are viable [2]. It is often times due to the obvious instability, these are surprisingly
inadequate reduction or distraction at the fracture nonpainful. In fact, this was defined by Harwood
22 A. Agarwal, MD
and Ferguson [31] as “a painless fracture that has Fig. 1.9 Patient with a right grade I open tibia fracture c
failed to unite and has no potential to do so treated initially with irrigation and debridement and
reamed intrameduallary (IM) nailing. a, b Injury, antero-
without intervention.” These all require surgery. posterior (AP) and lateral. c, d Postop AP and lateral,
The cavity at the fracture site is usually filled follow-up after 10 months showing development of
with a synovial lining creating a “false joint.” a hypertrophic nonunion and subsequent exchange nail-
These require resection of this cavity along with ing with union. e, f Nonunion AP and lateral. g, h
Exchange IM nail, AP and lateral. i, j Healed AP and
stabilization and bone grafting. lateral
Other classification schemes have been
reported as well [71–73]. The classical Ilizarov classify nonunions and dictate the level of care
description has been to define nonunions based that the nonunion requires. This scoring system
upon the amount of motion at the site, as stiff, takes into consideration the bone, soft tissues,
slack, or lax [71]. These correspond to the pre- and the patient to determine the best course of
viously described radiographic appearances as action. The maximum score would be 100
well. It is however important to take into con- (scored points 2). The scoring system is very
sideration that motion can only be assessed in the comprehensive and looks at all the issues pre-
absence of intact hardware or adjacent intact viously mentioned including the fracture char-
structures, e.g., an intact fibula in the case of the acteristics, adequacy of original treatment,
tibia. The stiff nonunion (hypertrophic) generally defects, alignment, soft tissue integrity, and
has no detectable motion on stress examination. patient risk factors. The higher the score, the
The slack nonunion (oligotrophic-hypertrophic) more difficult it was felt to obtain union. Those
has some motion hinging at the fracture site. The with a score up to 25 were felt to have a
lax nonunion (atrophic) has free movement at the straightforward nonunion that could be managed
fracture site. This classification is often used in by standard techniques. Those with scores from
the management of nonunions with external fix- 26 to 50 should have more specialized care. In
ation [71, 74–76]. addition to specialized care, specialized treat-
Biasibetti et al. [72] reported their classifica- ment was also required if the score was 51–75.
tion based on radiographic evaluation. Their They recommended consideration for amputa-
preference is for the use of external fixation in tion for any score above 75. Although this score
the management of these nonunions. They looks to have some promise, it has not been
defined nonunions as types 1–4. The type 1 validated to our knowledge.
nonunions are the classic hypertrophic nonunions Careful assessment of the radiographs over
that require mechanical stabilization by com- time can help classify the type of nonunion. The
pression. The type 2 nonunions are those with type of nonunion can then help determine the
large oblique fragments where axial compression cause of the nonunion suggesting either a bio-
would result in shear and torsion with negative logic or mechanical etiology. Taking all of these
affects on consolidation. Type 3 nonunions are previous factors that have been discussed can
those that were comminuted injuries, have sig- help determine the best course of action to take in
nificant defects, or are atrophic. These require managing the nonunion. Classification and
both mechanical stability and biologic stimula- scoring systems can certainly be helpful.
tion. The type 4 is the infected nonunion.
The management of nonunions is extremely
complicated, and failure rates have been repor- 1.6 Management Principles
ted around the 20% level. Despite classification
schemes and scoring systems to better provide In general, the management principles for the
improved agreement on when to diagnose a treatment of the nonunion are common to all sites
nonunion, treatment guidelines are lacking. In and are based on the classification. The goals in
an effort to provide such guidelines, Calori et al. treatment of the nonunion are universal: 1. heal-
[73] in 2008 proposed a new scoring system to ing the nonunion, 2. restoring function, and 3.
24 A. Agarwal, MD
Fig. 1.10 Patient with

a segmental tibial shaft
treated with an intramedullary
nail but with distraction noted
at the proximal fracture. a, b
Anteroposterior (AP) and
lateral. Patient developed
an oligotrophic nonunion at
8 months at both sites. c, d
Nonunion, AP, and lateral.
Patient underwent
dynamization with removal of
both distal locking screws and
subsequently healed. e, f
Healed, AP and lateral
Fig. 1.11 Patient with left clavicle fracture without atrophic nonunion. c Healed clavicle fracture after
shortening and minimal elevation initially treated treatment with open reduction and internal fixation and
non-operatively. a Injury radiograph. b Nonunion radio- bone graft
graph after 3 months showing resorption and established
eliminating pain. There are two basic tenets in point of view as well as from the entire limb point
accomplishing these goals—maximize the biol- of view if u will. Keep in mind that improvement
ogy and re-establish appropriate mechanical of the local mechanical stability also improves the
integrity of the nonunion environment. Maxi- local biology at the nonunion site to promote
mizing the biology of the environment can be union. It has also been suggested that nonunions
looked at from two perspectives: local and sys- should be treated with polytherapy, insuring the
temic. Locally, it is important to enhance the nonunion site is enhanced with osteoprogenitor
biology at the nonunion site and eradicate infec- cells, growth factors, and an adequate osteocon-
tion if present. Systemically, the patient’s ductive scaffold in cases of adequate stability
comorbidities must be minimized or corrected if [77]. This is certainly an aggressive approach and
feasible. The mechanical integrity of the non- may be warranted if these three aspects of the
union environment can be looked at from a local diamond concept [11] are lacking.
26 A. Agarwal, MD
Fig. 1.12 Patient with left tibia fracture treated with intramedullary nailing of pseudarthrosis after resection
closed management and development of pseudarthrosis of synovial cavity with subsequent union.(c, d) Healed AP
tibia but healed fibula. (a, b) Pseudarthrosis anteroposte- and lateral
rior (AP) and lateral; Patient treated with reamed
1.6.1 Biological Environment: options for bony defects are discussed in

Systemic Chap. 15, but include the Masquelet technique
with massive bone grafting, distraction osteoge-
After a careful evaluation of the patient and nesis (bone transport) [71, 74], and vascularized
causes for the nonunion, any metabolic abnor- or nonvascularized bone grafts (Figs. 1.13 and
malities should be addressed. Vitamin D defi- 1.14). Struijs et al. [79] reviewed the literature on
ciency or insufficiency should be corrected with the management of infected long-bone non-
replacement vitamin D therapy. Our preference is unions. The majority were case series, and
to start patients at 50,000 units of vitamin D definitive conclusions and recommendations
weekly for at least 6 months. Levels should be could not be made. However, it was clear that
obtained after 4–6 weeks to insure a proper appropriate debridement is universally required
response. In patients with associated secondary as a basis for any further treatment. The majority
hyperparathyroidism, vitamin D replacement of the first-stage treatment methodologies, when
should solve the high PTH level. Patients should significant bone resection is required for associ-
also be given calcium supplementation along ated osteomyelitis, included bone transport
with the vitamin D. Smoking cessation counsel- techniques with the Ilizarov fixator after the
ing should be initiated in efforts to minimize or debridement with 70–100% union results. In a
even stop smoking to aid in the healing process retrospective review of utilizing a single-stage
after reconstruction. Diabetes should be as well treatment protocol for “presumptive” aseptic
controlled as feasible. All comorbidities should nonunions, the authors had success in preventing
be optimized prior to intervention if time allows. secondary surgery in 72% of culture positive
cases [80]. If preoperatively, the history and
clinical examination do not indicate an infection,
1.6.2 Biological Environment: Local a single-stage protocol of withholding antibi-
otics, removing the implant, open debridement or
Infection should be ruled in or out prior to any canal reaming, sending cultures followed by
definitive management. If infection is present, the antibiotics and revision ORIF or exchange nail-
decision between a one-stage and two-stage ing was performed. They had positive cultures in
treatment plan must be made [78]. If a 28.7% and out of those 28% needed secondary
two-stage approach is deemed necessary, the first surgery. Overall, they felt that a single-stage
step is to remove existing hardware, evaluate the protocol is warranted in cases where the non-
nonunion site, and obtain cultures and/or biop- union is considered aseptic preoperatively. In
sies to determine the presence and extent of the obvious cases of infected nonunions, a
infection. The presence of osteomyelitis must be single-stage protocol is not utilized. The best
determined as well as the extent of bony two-stage results (93–100% union with recurrent
involvement. If there is an obvious infection, the infection rate up to 18%) seemed to be with
infection must be cleared up prior to wound debridement, antibiotic beads, and planned sec-
closure and certainly before definitive manage- ondary fixation. In a study by Obremskey et al.
ment for the nonunion including bone grafting [35] where they surveyed members of the
[78]. There are varied opinions on when timing Orthopedic Trauma Association (OTA), almost
of the bone graft should occur from immediate to 90% of surgeons used some sort of antibiotic
6 weeks after the resection [35]. If osteomyelitis cement spacer before bone grafting in a seg-
is present or suspected and confirmed by biopsy, mental defect. In any event, treatment needs to be
the amount of bony resection that needs to be individualized for the patient. Such specialized
performed to eradicate the infection has to be techniques as using antibiotic impregnated
determined. As mentioned before, MRI is quite cement-coated nails for the interim stabilization
useful since marrow changes can help delineate in infected long-bone nonunions are mainstay of
the extent of osteomyelitis. Reconstruction treatment. Recently, Scolaro and Mehta [81]
28 A. Agarwal, MD
Fig. 1.13 Patient presented with a year history of osteomyelitis with plating and cement spacer placement
draining sinus tracts after open reduction and internal to create membrane in anticipation of Masquelet proce-
fixation (ORIF) of grade IIIB right open tibia fracture dure. e, f AP and lateral after ORIF and cement
concerning for infected nonunion. a, b Infected Nonunion placement. Bone grafting into membrane, which was
anteroposterior (AP) and lateral. Patient underwent infec- obtained from ipsilateral femur, using
tion workup, which was consistent with osteomyelitis. reamer-irrigator-aspirator system (RIATM). g, h AP and
The patient then had hardware removal with evaluation lateral after cement removal and bone grafting. Patient
and biopsy of the bone. c, d AP and lateral after hardware went on to heal with complete consolidation of the bone
removal. Patient then underwent resection of graft. i, j. One-year follow-up AP and lateral of tibia
Fig. 1.14 Patient referred after sustaining right harvesting from contra-lateral femur using the
Grade IIIB open femur fracture with massive bone loss. reamer-irrigator-aspirator system and then placed into
Initial treatment was irrigation and debridement and defect after cement removed. c, d Initial postop AP and
retrograde intramedullary nail with placement of bone lateral radiographs after bone grafting. Patient went on to
cement in defect in anticipation of Masquelet technique. consolidate after the bone grafting with complete healing
a, b Presenting anteroposterior (AP) and lateral showing across the defect. e, f AP and lateral showing consolida-
bone defect with cement. Patient underwent bone tion and incorporation of bone graft into defect
30 A. Agarwal, MD
described the use of antibiotic impregnated as an alternative bone graft. Since the invention
cement-coated locking plates in the use of of the RIATM system, it has been used more and
infected peri-articular nonunions. In a similar more for harvesting autogenous bone graft.
fashion to utilizing the antibiotic impregnated Those that have used it cite lower complications
cement-coated nail to stabilize the nonunion and lower comorbidity than ICBG harvesting.
temporarily while the infection is cleared, a plate Dimitrou et al. [85] performed a systematic
was used in peri-articular nonunions for the same review of the literature and found that the overall
reason with success, albeit a small case series. In complication rate for RIATM was 6% compared
addition, the infection must be addressed with to 19.4% for ICBG harvesting. They also showed
appropriate antibiotics. The duration of and mode that there were differences between the anterior
of delivery (intravenous and per os) also should and posterior crest harvest sites. The anterior
be based on the organism, bone penetration of crest had significantly higher rates of infection,
the antibiotics, and the retention of potentially hematoma formation, fracture, and hypertrophic
contaminated hardware. Much of the treatment scar formation but significantly lower rates of
plan is based on the surgeon’s experience since chronic donor site pain and sensory disturbances.
each case can be unique. Amputation should In a separate clinical study by Loeffler et al. [87],
always be discussed with patients that have they prospectively enrolled 92 patients under-
infected nonunions [71, 73]. If infection is pre- going anterior ICBG for nonunions. They had a
sent, the reconstructive path can be long and 3% infection rate and only 2% rate of chronic
difficult. Many patients have already undergone pain. They felt that anterior ICBG harvesting was
numerous surgeries over several years to no avail well tolerated. In addition to comparing compli-
in resolving the infection or nonunion. cation rates between RIATM and ICBG, there has
Atrophic or oligotrophic nonunions require a been concern that the bone graft quality (cellular
biologic stimulus to reinitiate the healing pro- constituents and biochemical characteristics)
cess. This is usually in the form of a bone graft. from the intramedullary canal is not as good.
Autogenous bone graft remains the gold standard Sagi et al. [88], in a prospective study, harvested
[31, 78, 82–84] because it is osteogenic, bone graft from both the medullary canal
osteoinductive, and osteoconductive, with the (RIATM) and the iliac crest from the same indi-
iliac crest (ICBG) as the most common site of vidual for nonunion procedures. They evaluated
harvest historically [31, 84, 85]. In a retrospec- the graft histologically and performed transcrip-
tive study of long-bone nonunions by Flierl et al. tional profiling for biochemical markers that are
[86], they compared the success rates of 5 dif- known to be expressed during fracture healing.
ferent groups: autograft, allograft, auto- The transcriptional profiles were found to be very
graft + allograft, recombinant human bone similar. The RIATM graft was found to have
morphogenetic protein-2 (rhBMP-2) ± adjunc- greater regenerative characteristics as well as
tive bone grafting. The autograft was superior in mesenchymal stem cells. This suggests that
union time, had the lowest rates of surgical RIATM bone graft may actually be better.
revisions and revision bone grafting, and had a Dawson et al. [89] looked at the union rates
lower new-onset postoperative infection rate. between RIATM and ICBG in a prospective
Obremskey et al. [35] surveyed members of the randomized study for nonunions or a
OTA and 92% use autograft for bone grafting post-traumatic defect that required operative
procedures in nonunions. The site from where intervention. They had 113 patients for the final
the bone graft was obtained varied, however, statistics, 57 patients received ICBG, and 56 had
with 50.9% of the respondents picking the RIATM grafting. The union rates were similar as
reamer-irrigator-aspirator system (RIATM), were the rates of donor site complications, but
49.9% chose anterior crest, and 24.8% for pos- the RIATM had larger volumes of graft (anterior
terior crest (more than 1 choice was allowed). ICBG SS than posterior ICBG NSS <
Only 20.8% of surgeons used allograft and BMP RIATM) and had significantly less donor site
pain. Autogenous bone graft remains the gold centrifuged by a variety of commercially avail-
standard, but the choice of harvest site is still at able proprietary systems that separate out the
the surgeons’ discretion. Adjunctive bone sub- platelet rich plasma (PRP), which has shown
stitutes are sometimes required for recalcitrant mixed results in aiding facture healing [31, 50,
nonunions or if more volume is needed. The 84]. The PRP is only considered osteoinductive.
author prefers to use RIATM especially if large A large bore needle is inserted into the iliac
volumes of bone graft are required or the canal is crest in order to aspirate the bone marrow. The
being accessed due to the implants being used aspirate can then be directly injected into the
(intramedullary nails). Other local donor sites nonunion site under fluoroscopic guidance [70,
have been utilized for small amounts of graft 94]. The technique is useful when the retained
depending on the site of the nonunion, e.g., distal hardware is intact and stable. Braly et al. [92]
radius metaphysis for forearm nonunions or published their case series in eleven consecutive
proximal tibia metaphysis for distal tibia/fibula patients that presented with delayed union or a
nonunions. If the patient is obese, the preference nonunion of the distal tibia metaphysis that were
is also for RIATM to avoid wound complications initially treated with ORIF. They had 9 of 11
with the large soft tissue envelope. patients heal within six months. They found it to
Another alternative for large bone defects be a safe and inexpensive, minimally invasive
besides massive autogenous cancellous bone treatment. However, the use of BMA is limited
grafts are the vascularized bone grafts [90]. In because of the small number of stem cells
certain situations, a vascularized graft may be the obtained [82, 84]. In an effort to increase the
best option. Not only can they provide better number of stem cells, multiple aspirations and
structural support but can promote healing due to cell concentration techniques have been described
the added blood supply. Numerous types of [83]. Other future concepts include culturing
vascularized bone grafts exist [91], but it does aspirated cells to increase the numbers and com-
require a surgeon with microvascular skills and bining these cultured cells with specific scaffolds
they tend to have more issues. Historically, it was during surgery or in the laboratory creating
felt that if large structural grafts were used hybrid constructs for implantation [83, 93, 95].
(>6 cm), then it should be vascularized. Allsopp To obtain PRP, peripheral blood is obtained
et al. [90] in reviewing the literature found no from the patient. The amount of blood required
evidence to support this perception nor did it depends on the commercial PRP concentration
support that the success was superior to nonva- centrifuge systems available. The blood is placed
scularized grafts. The technique is still useful and in the centrifuge and the PRP separates out. It
a valuable part of one’s armamentarium. It may can then be drawn into a syringe and injected at
be beneficial in particular situations such as the nonunion site [31]. The PRP contains
nonunions complicated with osteonecrosis, e.g., numerous growth factors but in low concentra-
the femoral neck or scaphoid [91]. tions [84]. The clinical outcome in the use of
One of the advantages of autogenous bone PRP has been extremely varied and thus has not
grafts is that it contains the patient’s own gained wide acceptance.
osteoprogenitor cells which aid in its osteogenic Bone morphogenetic proteins (BMPs) have
potential. Other sources that can provide osteo- been considered the most important growth fac-
genesis are bone marrow and peripheral blood. tor in bone formation and healing. Although
Bone marrow (BMA) can be aspirated providing there are many BMPs that have been described,
a source of osteoprogenitor cells that have been only three have been shown to stimulate stem
shown to provide a biologic stimulus to aid cell differentiation into the osteoblast lineage
healing in nonunions [31, 70, 82–84, 92–95]. It in vitro—BMP 2, 4, and 7 [96, 97]. Only
is considered both osteogenic and osteoinductive. recombinant BMP 2 (rhBMP-2:Infuse; Med-
Peripheral blood can be obtained and then tronic Sofamor Danek, Memphis, TN) and 7
32 A. Agarwal, MD
(rhBMP-7: Osteogenic Protein-1 [OP-1]; Stryker, osteoinductive capacity of the graft. Morison
Kalamazoo, MI) have been cleared by the FDA et al. [104] looked at the use of rhBMP-7 in
for clinical use, with rhBMP-2 indicated for open atrophic long-bone nonunions in the upper
tibia fractures and rhBMP-7 for recalcitrant long extremity. They used BMP alone but with plate
bone nonunions [98, 99]. There has been exten- fixation. However, they did state that if local
sive research looking at these molecules as a way autogenous bone was available, it was morse-
to repair nonunions and accelerate the fracture lized and added but autogenous bone graft was
healing process [96, 97, 99]. Friedlaender et al. not harvested. They had an 89% success rate.
[100] reported on their prospective, randomized As mentioned before, the other BMP that is
control multicenter study utilizing OP-1 in the available is rhBMP-2. This has been primarily
management of tibial nonunions. All patients had studied in open tibia fractures with little to no
an intramedullary rod and either OP-1 or auto- data in nonunions to our knowledge. Jones et al.
graft. At 9-month follow-up, there was a clinical [105] in the BESTT-ALL trial used BMP-2 with
success rate of 85% and a radiographically allograft versus autograft in open tibia fracture
healed rate of 84% in the autogenous bone graft bone defects. The average size of the defect was
group compared to 81% and 75%, respectively, 4 cm (1–7 cm). The success rates were not sta-
in the OP-1 group. There was no statistically tistically significantly different with a 67% union
significant difference between the two patients, rate in the autograft group and 87% in the
although 20% in the autograft group had chronic BMP-2+ allograft group. In a study by Aro et al.
pain at the donor site. OP-1 was felt to be safe [106], rhBMP-2 was used in conjunction with
and effective for tibial nonunions. Dimitriou et al. reamed intramedullary nail fixation of open tibia
[101] utilized rhBMP-7 in 25 consecutive fractures and compared to reamed nailing alone.
patients with 26 nonunions in various locations. They found that the use of BMP-2 did not
They had success in 24 of the 26 nonunions; accelerate the rate of fracture healing, despite the
however, 16 of these successes also had autograft trend toward faster healing at the 13-week mark
in addition to the BMP. Only 8 cases had in the BMP-2 group. This difference normalized
rhBMP-7 alone. In an observational retrospec- at 20 weeks, where 68% of the BMP-2 group and
tive, nonrandomized study by Ronga et al. [102], 67% of the nail alone were healed.
they also looked at the use of rhBMP-7 in Multiple reviews of the literature [97, 98, 107]
long-bone nonunions. They had an 88.8% suc- have all concluded that although there was good
cess rate with an average healing time of clinical data on the effectiveness of BMPs, it was
7.9 months. Their group was mixed in that 38 as good but not better than autogenous bone
cases had the BMP alone, 11 cases were BMP graft. The use of BMPs can be expensive, but a
with an osteoconductive agent, 50 cases with an cost–benefit analysis has shown that their use can
autograft, and a composite graft in 6. In both potentially provide a cost savings in both non-
these studies, the only conclusion that could be unions and open fractures [97–99, 103]. More
made was that the use of rhBMP-7 was safe and prospective, randomized clinical studies are
effective and could be utilized with autograft. needed to determine the true effectiveness of
Giannoudis et al. [103] specifically looked at the BMPs in both nonunions and acute fractures.
effect of BMP-7 with autograft. They retrospec- There are many other bone graft substitutes,
tively reviewed their prospective database of either derived from human sources or manmade,
patients treated for atrophic nonunions in which that are commercially available [31, 82, 84].
both BMP-7 and autograft were used in all They include the calcium phosphate substances,
patients at different anatomic sites. Revision of bioactive glass, coral, allograft, and demineral-
the fixation was also performed in 77.8% of ized bone matrix (DBM). The synthetic substi-
cases. They had a 100% union rate. They con- tutes and allograft are strictly osteoconductive,
cluded that although autograft was the gold whereas the DBM is both osteoconductive and
standard, the BMP-7 could enhance the osteoinductive, although the osteoconductive
potential is highly variable based on the company that exchange nailing was the method of choice
[84]. Most of these materials are best used as based upon better than 90% union rates. In
graft extenders in the management of nonunions. looking at the literature on femoral nonunions,
No good clinical studies exist evaluating these Crowley et al. [109] also found excellent rates
materials in the treatment of nonunions. with exchange nailing and felt that it remained
It is also important to note that in cases of the gold standard despite good results with
open injuries, soft tissue management is integral adjunctive plate fixation. Swanson et al. [110,
to the initial treatment. Poor soft tissues and 111] reported their excellent results utilizing a
inadequate vascularity may contribute to the systematic approach in both femoral and tibial
development of a nonunion. In all cases of nonunions regardless of classification. All
nonunions, the local soft tissue environment patients had correction of any metabolic or
should be appropriately assessed. If the soft tis- endocrine abnormalities. The atrophic nonunions
sues are deficient or damaged, it is important to did not have open bone grafting. The femurs
obtain good soft tissue coverage through the use underwent secondary dynamization in 28% of
of either local or free tissue flaps [31]. the cases and the tibias in 7% of cases. There
were 4 cases (9%) that had partial fibulectomy at
the same time as the exchange nailing in the tibia
1.6.3 Mechanical Environment: Local cases. They had a 100% union rate in femurs and
98% union rate in tibias. In both studies, they
The fixation may or may not need revision routinely exchanged nails with a size at least
depending on the technical considerations with 2 mm larger in diameter in static mode but used a
respect to the management of the initial fracture different manufacturer’s nail. The use of a dif-
and its integrity. If appropriate and intact, then ferent manufacturer’s nail was felt to be impor-
the biologic stimulus may be insufficient. If the tant to optimize screw purchase since the screw
original fixation was inadequate or has failed, the locations/trajectories would be different. Other
construct should be appropriately revised and the more recalcitrant long bone hypertrophic non-
need for bone graft assessed based upon the unions that have failed exchange nailing may be
initial healing response. The radiographic better off with adjunct plate fixation, which has
appearance of the nonunion should be used to been shown to be effective in these situations
classify the site, which can aid in the manage- especially for the femur [65, 108, 109]
ment and determination of the need for bone (Fig. 1.16). In cases of plate fixation and hyper-
graft. trophic nonunions, often times the entire con-
In cases of hypertrophic nonunions, stability struct needs to be removed and completely
is needed (see Figs. 1.6g, h and 1.15). The best revised. In these cases, the hardware has often
treatment is based largely in part due to the initial failed with resultant mal-alignment. The healing
implant used for the original fracture. In cases of actually may continue because of the excessive
fractures previously treated with intramedullary motion causing increased callus and a stiff non-
nails, exchange nailing is regarded as the method union. Many times, these can be managed with
of choice for both the femur [65, 108–110] and distraction osteogenesis utilizing external fixa-
tibia [107, 108, 111] (see Fig. 1.9g–j). Nail dy- tion, which also allows for correction of the
namization is best reserved for cases of static deformity at the same time and subsequent
locking and oligotrophic nonunions to stimulate healing of the nonunion [74–76, 112] (Fig. 1.1d–
the healing response [65] (see Fig. 1.10e, f). Care m). The frames can provide an excellent
must be taken in cases of comminuted or oblique mechanically stable environment to provide
fractures, where dynamization could lead to healing. Distraction osteogenesis, by applying an
unacceptable shortening or loss of rotation. In a Ilizarov circular fixator, was used in a case series
review of the literature on aseptic tibial non- of 16 hypertrophic mal-aligned nonunions [75].
unions, Kanakaris et al. [107] in 2007 concluded They had complete correction of the deformity
34 A. Agarwal, MD
b Fig. 1.15 Patient sustained a “nightstick” fracture to nonunion is if there are issues. If the nonunion
right ulna after an assault and was treated with cast does heal despite ignoring the malalignment, a
immobilization and subsequent bracing but developed
painful nonunion (a, b) anteroposterior (AP) and lateral of malunion will be created which can in and of
established hypertrophic nonunion. Patient required sta- itself be problematic for the patient [61]. It is
bility and underwent open reduction and internal fixation imperative to fully evaluate the associated
(ORIF) with plate fixation. c, d AP and lateral after ORIF. deformity with the appropriate radiographs (see
The fracture healed once stability was obtained. e, f
One-year follow-up AP and lateral showing the healed previous section), scannogram for length, and CT
ulna scan for rotational issues if warranted. All bones
in the particular limb should be assessed for any
and 100% union. Feldman et al. [112] used the mal-alignment with long limb standing films.
Taylor spatial frame (TSF) in conjunction with A detailed physical examination should be per-
bone grafting to heal 5 atrophic nonunions in formed to assess for any compensatory changes
addition to two hypertrophic nonunions with in the adjacent joints.
100% success. Schoenleber and Hutson [76]
reported their results on eight patients utilizing
either an Ilizarov fixator or the TSF for distrac- 1.6.5 Adjunct Therapies
tion osteogenesis also with 100% success. Other
times, revision of the internal fixation is war- In addition to the surgical management of non-
ranted to either a new plate construct, or in some unions, noninvasive interventions in the form of
cases depending upon the anatomic site in bone stimulators have been used to help facilitate
question, intramedullary devices can be suc- fracture healing acutely as well in cases of
cessful in obtaining union if the intramedullary delayed unions or nonunions [108]. They come
canal is still patent or can be re-established. in three forms: 1. ultrasound, 2. extracorporeal
In cases of atrophic nonunions with hardware shock waves (ESWT), and 3. electrical stimula-
failure, both the local mechanical and biological tion [70]. The clinical data are varied.
environment (see above) need to be addressed. Low-intensity pulsed ultrasound (LIPUS) has
Revision fixation is warranted if the hardware the most clinical data and has been shown to
has failed or was inadequate at the outset. enhance bone healing safely [113–119]. The lit-
Anticipation of the need for early bone grafting erature has shown that ultrasound can reduce the
in cases of bone loss can help prevent hardware healing times of fresh fractures of the radius and
failure and can lead to a successful outcome. tibia, can offset the negative effects of smoking
and age on fracture healing, and can be effective
in the treatment of delayed unions and nonunions
1.6.4 Mechanical Environment: Limb [114–119] (Fig. 1.17). Heckman et al. [115]
evaluated the use of LIPUS in acute tibia frac-
As mentioned, any associated deformity must be tures. They had 67 fractures that were all treated
addressed when dealing with the nonunion. with long leg immobilization. They showed that
Repair of the nonunion without correction of any LIPUS significantly decreased the time to clinical
deformity, especially the mechanical alignment union and overall union (clinical and radio-
in the lower extremity, will often fail to restore graphic) compared to the nontreatment
the proper biomechanics and result in persistence group. Cook et al. [116] looked at patients that
of the nonunion [31]. Correction of the biome- had either tibia fractures or distal radius fractures
chanics is crucial for many nonunions such as the and that smoked to see whether LIPUS showed a
femoral neck, where valgus intertrochanteric difference in healing with smokers. In the tibia
osteotomy can be successful as long as excessive fracture groups, they showed a significant 41%
valgus alignment is avoided [61]. The length, reduction in healing time in the smokers and a
alignment, and rotation should always be asses- 26% significant reduction in healing time in the
sed in patients and addressed at the time the nonsmokers. Smokers had a significant 51%
36 A. Agarwal, MD
b Fig. 1.16 Patient referred for nonunion of left femur that reduction in healing time and nonsmokers a
had undergone four prior surgeries over a two year period. significant 34% reduction in healing time for the
a, b Presenting anteroposterior (AP) and lateral radio-
graphs showing oligotrophic nonunion. Patient underwent patients with distal radius fractures. Nolte et al.
exchange nailing with bone graft harvesting from the [118] reported their results in 29 cases of non-
same femur using the reamer-irrigator-aspirator (RIATM) unions treated with LIPUS. They had an 86%
system and placement of bone graft at the fracture site. c, success rate with an average treatment time of
d Postop AP and lateral radiographs after exchange
intramedullary nail and bone grafting. Patient continued 22 weeks. This was a heterogeneous group of
with pain and now a hypertrophic nonunion 8 months nonunions including both a variety of anatomic
later. e, f AP and lateral showing development/conversion locations and types of nonunion. In a review of
into a hypertrophic nonunion. Patient with persistent pain the literature, Watanabe et al. [119] showed that
and instability at fracture site now with recalcitrant
nonunion. Adjunctive plating was performed to provide the reported success rates seemed to be better for
increased stability. g, h Postoperative AP and lateral more subcutaneous bones than the deeper bones
radiographs after plating; Patient subsequently had reso- in both delayed unions and nonunions. Overall
lution of his pain with complete healing of the nonunion. rates in prospective cohort studies were reported
i, j One-year follow-up AP and lateral radiographs
showing complete consolidation of the nonunion site as anywhere from 55 to 100%. It has also been
Fig. 1.17 Elderly male presented with left humeral shaft lateral radiographs showing nonunion. Discussion with
fracture after low energy fall. Patient had significant patient and family was to try alternative methods due to
medical comorbidities and decision was made to manage high surgical risks. Ultrasound bone stimulator (Exo-
the patient with bracing. a, b Injury anteroposterior genTM) was started and patient went on to heal the
(AP) and lateral radiographs in brace showing excellent fracture. e, f AP and lateral radiographs 6 months later
alignment. Patient continued with mild discomfort and showing healed fracture
radiographs showed persistent nonunion. c, d AP and
38 A. Agarwal, MD
suggested that LIPUS has utility in distraction The studies available on adjunct therapies indi-
osteogenesis and can reduce the time required for cate that LIPUS has a much more positive response
maturation of the callus [114, 119]. in delayed unions and nonunions as well as in cer-
Another form of ultrasound therapy is ESWT. tain fresh fractures. However, none of the studies
In ESWT, shock waves that are single provide guidelines as to when it should or should not
high-amplitude sound waves which are generated be used. They can be of benefit in patients that may
by various means. It has been evaluated in the not be in the best health to undergo surgical proce-
treatment of delayed unions and nonunions. Zelle dures. The clinical decision-making should be based
et al. [120] in a systematic review of the literature on one’s experience, patient’s needs and wants, and
found ten-level 4 studies using ESWT for this the type of nonunion.
purpose. The overall union rate was 76% and was
found to be significantly higher in hypertrophic
nonunions (76%) than atrophic nonunions (29%). 1.7 Summary
They concluded that the cumulative data suggests
that ESWT can stimulate the healing process; The best management in treating nonunions is their
however, further studies are warranted due to the prevention. Adhering to basic AO principles of
level of evidence in these studies. fracture fixation and limiting the soft tissue dis-
There are several types of electrical stimula- section are paramount to a good result. Iatrogenic
tion available: (1) capacitively coupled electric causes have been shown to be a significant con-
field (CCEF), (2) pulsed electromagnetic fields tributor to nonunion development [25]. The soft
(PEMF), (3) direct current (DC) (more invasive), fracture callus that begins to form right away has
and 4. combined magnetic fields (CMF) [70, healing potential as shown by Danoff et al. [125] in
121]. In 2008, Mollon et al. [122] performed a an animal model. They created a mid-shaft femoral
meta-analysis of randomized control trials look- shaft fracture in rats and stabilized it with intrame-
ing at the use of electrical stimulation in long dullary nailing. They exposed the fracture site at
bone fracture healing. They could not show a seven days and created three study groups. In the
benefit of its use in improving the rate of union in first group, none of the soft callus was removed. In
fresh fractures, delayed unions or nonunions. the second, the soft callus was removed. The final
They did cite the heterogeneity in the studies as a group had the callus removed and then replaced.
reason for the lack of recommendations either The callus removal group showed significant evi-
way. In a subsequent review of the literature in dence of delayed healing. Replacing the callus
2010, Goldstein et al. [123] reviewed 4 separate mitigated the negative effect on the healing. They
meta-analyses on electrical stimulation in frac- recommended replacing the soft callus on all ORIF
ture healing. That review also concluded that no procedures. In addition to limiting the biologic
clear benefit to the use of electrical stimulation insult of surgery, all fracture patients should be
was seen. They felt that the meta-analysis by critically evaluated for comorbidities that may also
Mollon et al. [122] was the most methodologi- contribute to nonunion development as mentioned
cally rigorous. It was clear that better studies before. Early bone grafting when appropriate
were needed. Adie et al. [124] published their should be performed to aid fracture healing when
multicenter, double-blind randomized trial on the defects are present and promote healing and prevent
use of PEMF stimulation for acute tibial shaft hardware failure. If a nonunion presents, reassess-
fractures. They showed that the use of PEMF did ment of the patient is required. Critical evaluation of
not reduce the number of secondary procedures the initial treatment should be performed. If there
needed for delayed unions or nonunions. Addi- clearly were issues with the mechanical environ-
tionally, it did not improve union rates or ment, metabolic causes may not need to be sought
patient-reported functional outcomes in acute after; however, vitamin D insufficiency and suffi-
tibial shaft fractures. ciency are more prevalent than thought. If the initial
fixation was appropriate, then a metabolic workup non-union following fasciotomy for compartment
is warranted. There should be careful planning of syndrome associated with tibia fractures: a matched
cohort comparison. J Orthop Trauma. 2016;30
the treatment for the nonunion. All patient aspects (7):392–6.
of the nonunion must be addressed to include 13. Brinker MR, O’Connor DP, Monla YT, Earth-
deformities, metabolic issues, biology, and stability man TP. Metabolic and endocrine abnormalities in
of the nonunion site. patients with nonunions. J Orthop Trauma. 2007;21
(8):557–70.
14. Chachan S, Tudu B, Sahu B. Ultrasound monitoring
of fracture healing: is this the end of radiography in
References fracture follow-ups? J Ortho Trauma. 2015;29(3):
e133–8.
15. Cook, GE, Bates BD, Tornetta P, McKee MD,
1. Food and Drug Administration. U.S. Department of Morshed, Slobogean GP, Schemitsch EH. Assess-
Health and Human Services. Guidance document ment of fracture repair. J Orthop Trauma. 2015;29
for industry and CDRH staff for the preparation of (Suppl 12):S57-61.
investigational device exemptions and premarket 16. Cox G, Einhorn TA, Tzioupis C, Giannoudis PV.
approval applications for bone growth stimulator Bone-turnover markers in fracture healing. J Bone
devices; draft; availability. [Docket No. 98D-0238]. Joint Surg Br. 2010;92(3):329–34.
Rockville, MD: U. S. Food and Drug Administra- 17. Dimitriou R, Kanakaris N, Soucacos PN, Gian-
tion; 1998. noudis PV. Genetic predisposition to non-union:
2. Frölke JP, Patka P., Patka P. Definition and evidence today. Injury. 2013;44(Suppl 1):S50–3.
classification of fracture non-unions. Injury. 18. Fischer C, Doll J, Tanner M, Bruckner T, Zimmer-
2007;38(Suppl 21):S19–S22. Erratum in: Injury. mann Helbig L, et al. Quantification of TGF-B1,
2007;38(10)1224. PDGF and IGF-1 cytokine expression after fracture
3. Bhandari M, Fong K, Sprague S, Williams D, treatment vs. non-union therapy via masquelet.
Petrisor B. Variability in the definition and per- Injury. 2016;47(2):342–9.
ceived causes of delayed unions and nonunions. 19. Frank T, Osterhoff G, Sprague S, Garibaldi A,
J Bone Joint Surg Am. 2012;94(15):e1091–6. Bhandari M. Slobogean GP; FAITH Investigators.
4. Antonova E, Le TK, Burge R, Mershon J. Tibia The radiographic union score for hip (RUSH)
shaft fractures: costly burden of nonunions. BMC. identifies radiographic nonunion of femoral neck
213:14:42. fractures. Clin Orthop Relat Res. 2016;474
5. Kanakaris NK, Giannoudis PV. The health eco- (6):1396–404.
nomics of the treatment of long-bone non-unions. 20. Hankenson KD, Zimmerman G, Marcucio R. Bio-
Injury. 2007;38(Suppl 2):S77–84. logical perspectives of delayed fracture healing.
6. Hak DJ, Fitzpatrick D, Bishop JA, Marsh JL, Injury. 2014;45(Suppl 2):S8–15.
Tilp S, Schnettler R, et al. Delayed union and 21. Hernigou J, Schuind F. Smoking as a predictor of
nonunions: epidemiology, clinical issues, and finan- negative outcome in diaphyseal fracture healing. Int
cial aspects. Injury. 2014;45(Suppl2):S3–7. Orthop. 2013;37(5):883–7.
7. Brinker MR, Hanus BD, Sen M, O’Connor P. The 22. Kurdy NM. Serology of abnormal fracture healing:
devastating effects of tibial nonunion on the role of PIIINP, PICP, and BsALP. J Orthop
health-related quality of life. J Bone Joint Surg Trauma. 2000;14(1):48–53.
Am. 2013;95(24):2170–6. 23. Malik MH, Harwood P, Diggle P, Khan SA. Factors
8. Schottel PC, O’Connor P, Brinker MR. Time affecting rates of infection and nonunion in
trade-off as a measure of health-related quality of intramedullary nailing. J Bone Joint Surg Am.
life: long bone nonunions have a devastating impact. 2004;86(4):556–60.
J Bone Joint Surg Am. 2015;97(17):1406–10. 24. Murray IR, Foster CJ, Eros A, Robinson CM. Risk
9. Wichlas F, Tsitsilonis S, Disch AC, Haas NP, factors for nonunion after nonoperative treatment of
Hartmann C, Graef F, Schwabe P. Long-term func- displaced midshaft fractures of the clavicle. J Bone
tional outcome and quality of life after successful Joint Surg Am. 2013;95(13):1153–8.
surgical treatment of tibial nonunions. Int 25. Niikura T, Lee SY, Sakai Y, Nishida K, Kuroda R,
Orthop. 2015;39(3):521–5. Kurosaka M. Causative factors of fracture non-
10. Bishop JA, Palanca AA, Bellino MJ, Lowenberg DW. union: the proportions of mechanical, biological,
Assessment of compromised fracture healing. J Amer patient-dependent, and patient-independent factors.
Acad Orthop Surg. 2012;20(5):273–82. J Ortho Sci. 2014;19(1):120–4.
11. Giannoudis PV, Einhorn TA, marsh D. Fracture 26. O’Halloran K, Coale M, Costales T, Zerhusen T Jr,
healing: The diamond concept. Injury. 2007;38 Castillo RC, Nascone JW, O’Toole RV. Will my
(Suppl 4):S3-S6400-7. tibial fracture heal? Predicting nonunion at the time
12. Blair JA, Stoops TK, Doarn MC, Kemper D, of definitive fixation based on commonly available
Erdogan M, Griffing R, Sagi HC. Infection and
40 A. Agarwal, MD
variables. Clin Orthop Relat Res. 2016;474 experimental evidence. J Orthop Trauma. 2016;30
(6):1385–95. (3):104–112.
27. Pountos I, Georgouli T, Pneumaticos S, Giannoud- 38. Marsell R, Einhorn TA. The biology of fracture
is PV. Fracture non-union: can biomarkers predict healing. Injury. 2011;42(6):551–5.
outcome? Injury. 2013;44(12):1725–32. 39. Tzioupis C, Giannoudis PV. Prevalence of
28. Santolini E, West R, Giannoudis P. Risk factors for long-bone non-unions. Injury. 2007;38(Suppl 2):
long bone fracture non-union: a stratification S3-9. Erratum in Injury. 2007;38(10):1224.
approach based on the level of existing scientific 40. Borrelli J Jr, Pape C, Hak D, Hsu J, Lin S,
evidence. Injury. 2015;46(Suppl 8):S8–19. Giannoudis P, Lane J. Physiological challenges of
29. Westgeest J, Kin BSC, Weber D, Dulai SK, bone repair. J Orthop Trauma. 2012;26(12):708–11.
Bergman JW, Buckley R, Beaupre LA. Factors 41. Loder RT. The influence of diabetes mellitus on the
associated with development of nonunion or healing of closed fractures. Clin Orthop Relat Res.
delayed healing after an open long bone fracture: 1988;232:210–6.
a prospective cohort study of 736 subjects. J Orthop 42. Liao CC, Lin CS, Shih CC, Yeh CC, Chang YC,
Trauma. 2016;30(3):149–55. Lee YW, Chen TL. Increased risk of fracture and
30. Zimmermann G, Muller U, Wentzensen A. The postfracture adverse events in patients with dia-
value of laboratory and imaging studies in the betes: two nationwide population-based retrospec-
evaluation of long-bone non-unions. Injury. tive cohort studies. Diab Care. 2014;37(8):2246–
2007;38(Suppl 2):S33–7. 52.
31. Harwood PJ, Ferguson DO. An update on fracture 43. Wukich DK, Joseph A, Ryan M, Ramirez C,
healing and non-union. Orthop Trauma. 2015;29 Irrgang J. Outcomes of ankle fractures in patients
(4):228–42. with uncomplicated versus complicated diabetes.
32. Calori, GM, Albisetti W, Agus A, Iori S, Tagli- Foot Ankle Int. 2011;32(2):120–30.
abue L. Risk factors contributing to fracture 44. Sprague S, Petrisor B, Scott T, Devji T, Phillips M,
non-unions. Injury. 2007;38S:S11–S18. Spurr H, et al. What is the role of vitamin D
THERE IS NO CITATION 19. supplementation in acute fracture patients? A sys-
33. Lin CA, Swiontkowski M, Bhandari M, Walter SD, tematic review and meta-analysis of the prevalence
Schemitsch EH, Sanders D, Tornetta P 3rd. Ream- of hypovitaminosis D and supplementation efficacy.
ing does not affect functional outcomes after open J Orthop Trauma. 2016;30(2):53–63.
and closed tibial shaft fractures: the results of a 45. Sprague S, Bhandari M, Devji T, Scott T,
randomized controlled trial. J Orthop Trauma. Petrisor B, McKay P, Slobogean GP. Prescription
2016;30(3):142–8. of vitamin D to fracture patients: a lack of
34. Obremskey W, Molina C, Collinge C, Nana A, consensus and evidence. J Orthop Trauma.
Tornetta P 3rd, Sagi C, et al. Evidence-based quality 2016;30(2):e64–9.
value and safety committee orthopaedic trauma 46. Stephens BF, Murphy A, Mihalko WM. The effects
association, writing committee. Current practice in of nutritional deficiencies, smoking, and systemic
the management of open fractures among orthopae- disease on orthopaedic outcomes. J Bone Joint Surg
dic trauma surgeons. Part A: Initial management. Am. 2013;95(23):2153–7.
A survey of orthopaedic trauma surgeons. J Orthop 47. Childs BR, Andres BA, Vallier HA. Economic
Trauma. 2014;28(8):e198–202. benefit of calcium and vitamin D supplementation:
35. Obremskey W, Molina C, Collinge C, Tornetta P does it outweigh the cost of nonunions? J Orthop
III, Sagi C, Schmidt A, et al. Evidence-based Trauma. 2016;30(8):e285–8.
quality value and safety committee orthopaedic 48. Hobby B, Lee MA. Managing atrophic nonunion in
trauma association, writing committee current prac- the geriatric population: incidence, distribution, and
tice in the management of open fractures among causes. Orthop Clin North Am. 2013;44(2):251–6.
orthopaedic trauma surgeons: Part B: Management 49. Kwong FNK, Harris MB. Recent developments in
of segmental long bone defects. A survey of the biology of fracture repair. J Am Acad Orthop
Orthopaedic Trauma Association members. Surg. 2008;16(11):619–25.
J Orthop Trauma. 2014;28(8):e203–7. 50. Giannoudis PV, Hak D, Sanders D, Donohoe E,
36. Dye NB, Vagts C, Manson TT, O’Toole RV. Tosounidis T, Bahney C. Inflammation, bone
Estimation of tibial shaft defect volume using healing, and anti-inflammatory drugs: an update.
standard radiographs: development and validation J Orthop Trauma. 2015;29(Suppl):S6–9.
of a novel technique. Injury. 2015;46(2):299–307. 51. Kurmis AP, Kurmis TP, O’Brien JX, Dalen T. The
37. Hildebrand F, Griensven MV, Huber-Lang, M, effect of nonsteroidal anti-inflammatory drug
Flohe SB, Andruszkow H, Marzi I, Pape HC; administration on acute phase fracture-healing: a
Trauma research network for the German Society of review. J Bone Joint Surg Am. 2012;94(9):815–23.
Trauma, DGU. Is there an impact of concomitant 52. Lee JJ, Patel R, Biermann S, Dougherty PJ. The
injuries and timing of fixation of major fractures on musculoskeletal effects of cigarette smoking. J Bone
fracture healing? A focused review of clinical and Joint Surg Am. 2013;95(9):850–9.
53. Patel RA, Wilson RF, Patel PA, Palmer RM. The the femoral shaft. A review. Injury. 2012;43
effect of smoking on bone healing. Bone Joint Res. (7):980–8.
2013;2(6):102–11. 66. Stucken C, Olszewski DC, Creevy WR,
54. Scolaro JA, Schenker ML, Yannascoli S, Bald- Murakami AM, Tornetta P. Preoperative diagnosis
win K, Mehta S, Ahn J. Cigarette smoking increases of infection in patients with nonunions. J Bone Joint
complications following fracture. J Bone Joint Surg Surg Am. 2013;95(15):1409–12.
Am. 2014;96(8):674–81. 67. Termaat MF, Raijmakers PG, Scholten HJ,
55. Matuszewski PE, Boulton CL, O’Toole RV. Bakker FC, Patka P, Haarman HJ. The accuracy
Orthopaedic trauma patients and smoking: knowl- of diagnostic imaging for the assessmemt of chronic
edge deficits and interest in quitting. Injury. osteomyelitis: a systemic review and meta-analysis.
2016;47(6):1206–11. J Bone Joint Surg Am. 2005;87(11):2464–71.
56. Dijkman BG, Sprague S, Schemitsch EH, Bhan- 68. Stoffel K, Engler H, Kuster M, Riesen W. Changes
dari M. When is a fracture healed? Radiographic in biochemical markers after lower limb fractures.
and clinical criteria revisited. J Orthop Trauma. Clin Chem. 2007;53(1):131–4.
2010;24(Suppl 3):S76–80. 69. Weber BG, Cech O. Pseudarthrosis, pathology,
57. Bhandari M, Chiavaras M, Ayeni O, Chakraver- biomechanics, therapy, results. Bern: Hans Huber;
rty R, Parasu N, Choudur H, Bains S, Sprague S, 1976.
Petrisor B. Assessment of radiographic fracture 70. Panagiotis M. Classification of non-union. Injury.
healing in patients with operatively treated femoral 2005;36(Suppl 4):S30–7.
neck fractures. J Orthop Trauma. 2013;27(9):e213– 71. Atkins RM. Principles of management of septic
9. non-union of fracture. Injury. 2007;38(Suppl 2):
58. Koolstra BW, Dijkman BG, Sprague S, S23–32.
Schemitsch EH, Bhandari M. The radiographic 72. Biasibetti A, Aloj D, Di Gregorio G, Masse A,
union scale in tibia fractures: reliability and validity. Salomone C. Mechanical and biological treatment
J Orthop Trauma. 2010;24(Suppl 1):S81–6. of long-bone non-unions. Injury. 2005;36(Suppl 4):
59. Whelan DB, Bhandari M, Stephen D, Kreder H, S45–50.
McKee MD, Zdero R, Schemitsch EH. Develop- 73. Calori GM, Phillips M, Jeetle S, Tagliabue L,
ment of the radiographic union score for tibial Giannoudis PV. Classification of non-union: need
fractures for the assessment of tibial fracture healing for new scoring system? Injury. 2008;39(Suppl 2):
after intramedullary fixation. J Trauma. 2010;68 S59–63.
(3):629–32. 74. Kanellopoulos AD, Soucacos PN. Management of
60. Chiavaras MM, Bains S, Choudur H, Parasu N, nonunion with distraction osteogenesis. Injury.
Jacobson J, Ayeni O, et al. The radiographic union 2006;37(Suppl 1):S51–5.
score for hip (RUSH): the use of a checklist to 75. Kocaoglu M, Eralp L, Sen C, Cakmak M, Din-
evaluate hip fracture healing improves agreement cyurek H, Goksan SB. Management of stiff hyper-
between radiologists and orthopaedic surgeons. trophic nonunions by distraction osteogenesis.
Skeletal Radiol. 2013;42(8):1079–88. J Orthop Trauma. 2003;17(8):543–8.
61. Varghese VD, Boopalan PR, Titus VTK, Oom- 76. Schoenleber SJ, Hutson JJ Jr. Treatment of hyper-
men AT, Jepegnanam TS. Indices affecting outcome trophic distal tibia nonunion and early malunion
of neglected femoral neck fractures after valgus with callus distraction. Foot Ankle Int. 2015;36
intertrochanteric osteotomy. J Orthop Trauma. (4):400–7.
2014;28(7):410–6. 77. Calori GM, Mazza E, Colombo M, Ripamonti C,
62. Bhattacharyya T, Bouchard KA, Phadke A, Tagliabue L. Treatment of long bone non-unions
Meigs JB, Kassarjian A, Salamipour H. The with polytherapy: indications and clinical results.
accuracy of computed tomography for the diagnosis Injury. 2011;42(6):587–90.
of tibial nonunion. J Bone Joint Surg Am. 2006;88 78. Egol KA, Nauth A, Lee M, Pape HC, Watson JT,
(4):692–7. Borrelli J Jr. Bone grafting: sourcing, timing,
63. Augat P, Morgan EF, Lujan TJ, MacGillivray TJ, strategies, and alternatives. J Orthop Trauma.
Cheung WH. Imaging techniques for the assessment 2015;29(Suppl 12):S10–4.
of fracture repair. Injury. 2014;45(Suppl 2):S16–22. 79. Struijs PAA, Poolman RW, Bhandari M. Infected
64. Moed BR, Subramanian S, van Holsbeeck M, nonunion of the long bones. J Orthop Trauma.
Watson JT, Cramer KE, Karges DE, et al. Ultra- 2007;21(7):507–11.
sound for the diagnosis of tibial fracture healing 80. Amorosa LF, Buirs LD, Bexkens R, Wellman DS,
after static interlocked nailing without reaming: Kloen P, Lorich DG, Helfet DL. A single-stage
clinical results. J Orthop Trauma. 1998;12(3):206– treatment protocol for presumptive aseptic diaphy-
13. seal nonunions: a review of outcomes. J Orthop
65. Gelalis ID, Politis AN, Arnaoutoglou CM, Trauma. 2013;27(10):582–6.
Korompilias AV, Pakos EE, Vekris MD, et al. 81. Scolaro JA, Mehta S. Stabilisation of infected
Diagnostic and treatment modalities in nonunions of peri-articular nonunions with an antibiotic
42 A. Agarwal, MD
impregnated cement coated loaking plate: technique 96. De Biase P, Capanna R. Clinical applications of
and indications. Injury. 2016;47(6):1353–6. BMPs. Injury. 2005;36(Suppl 3):S43–6.
82. Mahendra A, Maclean AD. Available biological 97. Giannoudis PV, Dinopoulos HT. BMPs: options,
treatments for complex non-unions. Injury. 2007;38 indications, and effectiveness. J Orthop Trauma.
(Suppl 4):S7–12. 2010;24(Suppl 3):S9–16.
83. Rosset P, Deschaseaux F, Layrolle P. Cell therapy 98. Obert L, Deschaseaux F, Garbuio P. Critical anal-
for bone repair. Orthop Traumatol Surg Res. ysis and efficacy of BMPs in long bones non-union.
2014;100(Suppl 1):S107–S112. Review. Injury. 2005;36(Suppl 3):S38–42.
84. Sen MK, Miclau T. Autologous iliac crest bone 99. Schmidmaier G, Schwabe P, Wildemann B,
graft: should it still be the gold standard for treating Haas NP. Use of bone morphogenetic proteins for
nonunions? Injury. 2007;38(Suppl; 1):S75–S80. treatment of non-unions and future perspectives.
85. Dimitriou R, Mataliotakis GI, Angoules AG, Injury. 2007;38(Suppl 4):S35–41.
Kanakaris NK, Giannoudis PV. Complications 100. Friedlaender GE, Perry CR, Cole JD, Cook SD,
following autologous bone graft harvesting from Cierny G, Muschler GF, et al. Osteogenic protein-1
the iliac crest and using the RIA: a systematic (Bone morphogenetic protein-7) in the treatement of
review. Injury. 2011;42(Suppl 2):S3–15. tibial nonunions. J Bone Joint Surg Am. 2001;83-A
86. Flierl MA, Smith WR, Mauffrey C, Irgit K, Wil- (Suppl 1 Pt 2):S151–S158.
liams AE, Ross E, et al. Outcomes and complication 101. Dimitriou R, Dahabreh Z, Katsoulis E,
rates of different bone grafting modalities in long Matthews SJ, Branfoot T, Giannoudis PV. Appli-
bone fracture nonunions: a retrospective cohort cation of recombinant BMP-7 on persistent upper
study in 182 patients. J Ortho Surg Res. 2013;8:33. and lower limb non-unions. Injury. 2005;36(Suppl
87. Loeffler BJ, Kellam JF, Sims SH, 4):S51–9.
Bosse MJ. Prospective observational study of 102. Ronga M, Baldo F, Zappala G. Cherubino P;
donor-site morbidity following anterior iliac crest BMP-7 Italian Observational Study (BIOS)
bone-grafting in orthopaedic trauma reconstruction Group. Recombinant human bone morphogenetic
patients. J Bone Joint Surg Am. 2012;94(18):1649– protein-7 for treatment of long bone non-union: an
54. observational, retrospective, non-randomized study
88. Sagi HC, Young ML, Gerstenfeld L, Einhorn TA, of 105 patients. Injury. 2006;37(Suppl 3):S51–6.
Tornetta P. Qualitative and quantitative differences 103. Giannoudis PV, Kanakaris NK, Dimitriou R, Gill I,
between bone graft obtained from the medullary Kolimarala V, Montgomery RJ. The synergistic
canal (with a reamer/irrigator/aspirator) and the iliac effect of autograft and BMP-7 in the treatment of
crest of the same patient. J Bone Joint Surg Am. atrophic nonunions. Clin Orthop Relat Res.
2012;94(23):2128–35. 2009;467(12):3239–48.
89. Dawson J, Kiner D, Gardner W II, Swafford R, 104. Morison Z, Vicente M, Schemitsch EH,
Nowotarski PJ. The reamer- irrigator-aspirator McKee MD. The treatment of atrophic, recalcitrant
(RIA) as a device for harvesting bone graft long-bone nonunion in the upper extremity with
compared with iliac crest bone graft (ICBG): Union human recombinant bone morphogenetic protein-7
rates and complications. 2014;28(10):586–90. (rhBMP-7) and plate fixation: a retrospective
90. Allsopp BJ, Hunter-Smith DJ, Rozen WM. Vascu- review. Injury. 2016;47(2):356–63.
larized versus nonvascularized bone grafts: what is 105. Jones AL, Bucholz RW, Bosse MJ, Mirza SK,
the evidence? Clin Orthop Relat Res. 2016;474 Lyon TR, Webb LX, et al. BMP-2 evaluation in
(5):1319–27. surgery for Tibial Trauma-Allograft (BESTT-ALL)
91. Soucacos PN, Dailiana Z, Beris AE, Johnson EO. Study Group. Recombinant human BMP-2 and
Vascularized bone grafts for the management of allograft compared with autogenous bone graft for
non-union. Injury. 2006;37(Suppl 1):S41–50. reconstruction of diaphyseal tibial fractures with
92. Braly HL, O’Connor DP, Brinker MR. Percuta- cortical defects. A randomized, controlled trial.
neous autologous bone marrow injection in the J Bone Joint Surg Am. 2006;88(7):1431–41.
treatment of distal meta-diaphyseal tibial nonunions 106. Aro HT, Govender S, Pate AD, Hernigou P, de
and delayed unions. J Orthop Trauma. 2013;27 Gregorio AP, Popescu GI, et al. Recombinant
(9):527–34. human bone morphogenetic protein-2: a random-
93. Gómez-Barrena E, Rosset P, Lozano D, Stanovici J, ized trial in open tibial fractures treated with reamed
Ermthaller C, Gerbhard F. Bone fracture healing: nail fixation. J Bone J Surg Am. 2011;93(9):801–8.
cell therapy in delayed unions and nonunions. 107. Kanakaris NK, Paliobeis C, Manidakis N, Gian-
Bone. 2015;70:93–101. noudis PV. Biological enhancement of tibial dia-
94. Homma Y, Zimmermann G, Hernigou P. Cellular physeal aseptic non-unions: the efficacy of
therapies for the treatment of non-union: the past, autologous bone grafting, BMPs and reaming
present and future. Injury. 2013;44(Suppl 1):S46–9. by-products. Injury. 2007;38(Suppl 2):S65–S75.
95. Tseng SS, Lee MA, Reddi AH. Nonunions and the Erratum in Injuryl;207;28(10):1224.
potential of stem cells in fracture healing. J Bone 108. Brinker MR, O’Connor DP. Management of aseptic
Joint Surg Am. 2008;90(Suppl 1):92–8. tibial and femoral diaphyseal nonunions without
bony defects. Orthop Clin N Am. 2016;47(1):67– 117. Rubin C, Bolander M, Ryaby JP, Hadjiargyrou M.
75. The use of low-intensity ultrasound to accelerate the
109. Crowley DJ, Kanakaris NK, Giannoudis PV. healing of fractures. J Bone Joint Surg Am. 2001;83
Femoral diaphyseal aseptic non-unions: is there an (2):259–70.
ideal method of treatment? Injury. 2007;38(Suppl 118. Nolte PA, van der Krans A, Patka P, Janssen IM,
2):S55–63. Ryaby JP, Albers GH. Low-intenisty pulsed ultra-
110. Swanson EA, Garrard EC, Bernstein DT, O’Connor sound in the treatment of nonunions. J Trauma.
DP, Brinker MR. Results of a systematic approach 2001;51(4):693–702.
to exchange nailing for the treatment of aseptic 119. Watanabe Y, Matsushita T, Bhandari M, Zdero R,
femoral nonunions. J Orthop Trauma. 2015;29 Schemitsch EH. Ultrasound for fracture healing:
(1):21–7. current evidence. J Orthop Trauma. 2010;24(Suppl
111. Swanson EA, Garrard EC, O’Connor DP, 3):S56–61.
Brinker MR. Results of a systematic approach to 120. Zelle BA, Gollwitzer H, Zlowodzki M, Buhren V.
exchange nailing for the treatment of aseptic tibial Extracorporeal shock wave therapy: current evi-
nonunions. J Orthop Trauma. 2015;29(1):28–35. dence. J Orthop Trauma. 2010;24(Suppl 3):S66–70.
112. Feldman DS, Shin SS, Madan S, Koval KJ. Correc- 121. Karamitros AE, Kalentzos VN, Soucacos PN. Elec-
tion of tibial malunion and nonunion with six-axis tric stimulation and hyperbaric oxygen therapy in
analysis deformity correction using the Taylor the treatment fo nounions. Injury. 2006;37(Suppl
spatial frame. J Orthop Trauma. 2003;17(8):549– 1):S63–73.
54. 122. Mollon B, da Silva V, Busse JW, Einhorn TA,
113. Khan Y, Laurencin CT. Fracture repair with Bhandari M. Electrical stimulation for long-bone
ultrasound: clinical and cell-based evaluation. fracture-healing: a meta-analysis of randomized
J Bone Joint Surg Am. 2008;90(Suppl 1):138–44. controlled trials. J Bone Joint Surg Am. 2008;90
114. Malizos KN, Hantes ME, Protopappas V, (11):2322–30.
Papachristos A. Low intensity pulsed ultrasound 123. Goldstein C, Sprague S, Petrisor BA. Electrical
for bone healing: an overview. Injury. 2006;37 stimulation for fracture healing: clinical evidence.
(Suppl 1):S56–62. J Orthop Trauma. 2010;24(Suppl 3):S62–5.
115. Heckman JD, Ryaby JP, McCabe J, Frey JJ, 124. Adie S, Harris IA, Naylor JM, Rae H, Dao A,
Kilcoyne RF. Acceleration of tibial fracture healing Yong S, Ying V. Pulsed electromagnetic field
by non-invasive, low-intensity pulsed ultrasound. stimulation for acute tibial shaft fractures. J Bone
J Bone Joint Surg Am. 1994;76(1):26–34. Joint Surg Am. 2011;93(7):1569–76.
116. Cook SD, Ryaby JP, McCabe J, Frey JJ, Heck- 125. Danoff JR, Aurégan JC, Coyle RM, Burky RE,
man JD, Kristiansen TK. Acceleration of tibia and Rosenwasser MP. Augmentation of fracture healing
distal radius fracture healing in patients who smoke. using soft callus. J Orthop Trauma. 2016;30
Clin Orth Rel Res. 1997;337:198–207. (3):113–8.
Fracture Healing
2
Saam Morshed, MD, PhD, MPH and Anthony Ding, MD
interdependent processes; disruption in key steps

2.1 Introduction
can delay or terminate healing altogether.
The causative factors underlying nonunion are
Musculoskeletal injury is one of the leading
often multifactorial. Injury patterns, patient fac-
causes of disability and dysfunction worldwide.
tors, and even interventions all have substantial
In the USA alone, the burden of fracture care in
implications toward successful repair. A thor-
an aging population is projected to exceed $25
ough understanding of the normal healing pro-
billion in costs [1–3]. Achieving fracture union is
cess, and where it goes awry, is essential to the
paramount to patient recovery, return to activity,
diagnostic and therapeutic approach in treating
and quality of life following injury. While the
nonunions.
majority of fractures will heal uneventfully, a
The purpose of this chapter is to provide the
small but significant number will demonstrate
conceptual framework for understanding fracture
impaired healing [4]. When fractures fail to heal,
healing and its modulating factors in the context
they place a substantial burden on the patient and
of nonunion management. The first part dis-
on the healthcare system [5–7]. Brinker and
cusses the physiology of fracture healing—its
O’Connor [5] showed that fracture nonunion is
biology, mechanics, and assessment. The second
more burdensome than many chronic medical
part focuses on modulators of healing—
conditions, including chronic obstructive pul-
patient-related factors, comorbidities, injury pat-
monary disease and congestive heart failure.
terns, surgical intervention, and biologic aug-
Fracture healing is a complex, highly orches-
mentation—that may promote or impair fracture
trated regenerative process to restore skeletal
union.
integrity. The response following injury involves
tightly coordinated temporal and spatial interac-
tions among cytokines, growth factors, progenitor
cells, and adjacent tissues. The intricacy of frac-
2.2 Physiology of Fracture Healing
ture healing incorporates multiple pathways and
Despite its complexity, fracture healing is driven
by fundamental principles. Fractures all require a
viable pool of progenitor cells, an osteoconduc-
tive scaffold (extracellular matrix), signaling
S. Morshed (&) A. Ding molecules and their receptors, a vascular supply,
Zuckerberg San Francisco General Hospital, and a suitable mechanical milieu to heal. Failure
Orthopedic Trauma Institute, Department of in one or more of these domains impairs suc-
Orthopedic Surgery, University of California San cessful healing [8–12]. The ability to achieve
Francisco, 2550 23rd Street, Building 9, 2nd Floor,
San Francisco, CA 94110, USA fracture healing hinges on the interdependency
e-mail: saam.morshed@ucsf.edu between the mechanics and the biology at the

46 S. Morshed, MD, PhD, MPH and A. Ding, MD
fracture site. The mechanical environment dic- must be rigidly fixed, allowing minimal inter-
tates the biologic response to skeletal injury, and fragmentary strain (<5%) [15–19]. Gaps must be
there must be sufficient stability to promote bony small, less than 1 mm [17]. Because these con-
healing. As healing progresses, extracellular ditions usually do not occur naturally, direct
matrix is laid across the fracture site, which lends healing is primarily achieved by operative fixa-
further mechanical support to the fracture. tion [9]. These fixation methods include com-
pression plating, lag screw fixation (Fig. 2.1),
and multiplanar external fixation. Failure to meet
2.2.1 Biology of Fracture Healing the above conditions can impair the healing
process. Achieving rigid stability in the setting of
The healing response depends on the temporal comminution or a large fracture gap prohibits
and spatial interactions among four main tissue callus formation across the fracture site. Failing
types: cortical bone, bone marrow, periosteum, to respect the biology around the fracture site
and surrounding soft tissue. Ossification, the through extensive dissection and excessive soft
process of bone tissue formation both in normal tissue stripping likewise discourages healing
development and in skeletal injury, is a key (Fig. 2.2).
process in fracture healing. Endochondral ossi- Contact healing occurs in the absence of
fication utilizes a cartilage scaffold to form bone, gapping, where cortices are directly apposed.
whereas intramembranous ossification forms “Cutting cones” lay down new osteons longitu-
bone without a cartilage scaffold. dinally across the fracture site. Osteoclasts form
There are two main pathways of fracture the tip of the cone, resorb injured bone, and
healing: direct healing and indirect healing. create new Haversian canals (Fig. 2.3) [8]. New
Direct, or primary, healing allows for direct blood vessels, branching from endosteal and
remodeling of lamellar bone. It involves only periosteal circulation, penetrate the canals and
intramembranous ossification in the formation of deliver osteoblastic precursors. Osteoblasts form
bone. Indirect, or secondary, healing relies on the end of the cutting cone unit, laying down new
forming a cartilage callus scaffold, through bone that will eventually mature into its lamellar
which bone forms and remodels into its mature structure (Fig. 2.4) [8, 9, 13]. There is limited
lamellar structure. Whether a fracture heals by contribution from the surrounding periosteum
direct or indirect means is determined early by its and soft tissues.
biologic and physical environment [13, 14]. Ini- Gap healing occurs with small gaps less than
tial stability influences the inflammatory 0.8–1 mm under similar rigid conditions. Unlike
response following injury and can thus influence in contact healing, hematoma initially fills the
the mode of repair. Rigid stability follows a gap. It is quickly replaced with woven bone in
direct healing pathway, whereas relative stability the first 1–2 weeks. Woven bone is then
leads to indirect healing. Additionally, as with replaced by lamellar repair bone, though this
most biologic phenomenon, fracture healing interposed bone is oriented perpendicular to the
represents a spectrum with varying degrees of long bone axis. While stronger than cartilage,
direct and indirect healing happening simultane- this bone bridge is biomechanically weaker at
ously, depending on the anatomical location and its interface with the normal bone due to its
the mechanical environment. orthogonal orientation. At 6–8 weeks, the repair
bone undergoes secondary remodeling. Cutting
2.2.1.1 Direct Fracture Healing cones from the neighboring cortices traverse
Direct or primary healing regenerates lamellar and replace the repaired bone to reconstitute the
bone across the fracture without a cartilage canalicular system, recreate the longitudinal
scaffold. To do so, several conditions must exist. lamellar structure, and ultimately restore skeletal
First, the cortical bone must be anatomically integrity. No cartilaginous callus is formed [9,
reduced and apposed. Second, the fragments 20].
2 Fracture Healing 47
Fig. 2.1 Primary healing with absolute stability. The with lag screw fixation, neutralization plate. f, g 3-month
patient is a 26-year-old woman who was struck by a follow-up, showing progressive healing of tibia without
motor vehicle and sustained a Grade III open right distal callus formation and healing of fibula with callus. h,
tibia fracture. a Injury radiographs. b, c Initial irrigation i 1-year follow-up showing complete healing of tibia and
and debridement of the fracture site, spanning external fibula
fixation, and lag screw fixation. d, e Definitive fixation
2.2.1.2 Indirect Fracture Healing anatomical reduction. Rather, micromotion, to an

Indirect fracture healing regenerates bone extent, stimulates the healing response. Indirect
through a cartilage callus scaffold (Fig. 2.5) [13]. healing is the predominant mechanism in most
It still requires a relatively stable environment, fractures treated by nonoperative means. It is also
but it does not require rigid stability or achieved by interventions that allow for relative
Fig. 2.2 Impaired healing with absolute stability. The persistent fracture lines with little evidence of healing as
patient is a 41-year-old man who sustained an open right well as hardware failure, consistent with nonunion. c,
distal tibia fracture that was initially treated with open d Nonunion repair with removal of hardware and
reduction internal fixation at an outside facility. a, intramedullary nailing. e, f 6-month postoperative radio-
b 6-month postoperative radiographs demonstrate graphs with healing of fracture
Fig. 2.3 Cutting cones. Low

power photomicrograph of a
“cutting cone” in direct bone
healing and remodeling.
Multinucleated osteoclasts
(right) form the leading edge
of the cone, followed by
osteoblasts (left) forming new
bone. From Einhorn [8], with
permission
stability. These include intramedullary nailing of granulation tissue, and ultimately callus. Fur-
long bone fractures (Fig. 2.6), external fixation thermore, the hematoma itself contains progeni-
(Fig. 2.7), bridge plating (Fig. 2.8), and splint- tor cells, cytokines, and growth factors that
ing, bracing, or casting. directly participate in the healing process [22,
Three fundamental phases of indirect healing 23]. Recent studies have identified higher levels
have been described [21]: inflammatory, repara- of factors and signaling molecules in fracture
tive, and remodeling. Trauma initiates the acute hematoma. These include macrophage
inflammatory phase, and, through the release of colony-stimulating factor (M-CSF), transforming
mediators, cytokines, and growth factors, recruits growth factor-beta (TGF-b), and interleukins
progenitor cells responsible for initiating repair. (IL), all of which have important roles in stim-
In the reparative phase, progenitor cells lay down ulating fracture healing (Table 2.1) [24–27].
cartilaginous and bony callus, facilitate neoan- The initial inflammatory response occurs
giogenesis, and replace callus with woven bone. immediately after injury and lasts several days.
The remodeling phase replaces the woven bone The response is marked by infiltration of mac-
with a mature lamellar bone structure. rophages, platelets, polymorphonuclear leuko-
cytes, and lymphocytes into the fracture site.
Inflammatory Phase
These secrete proinflammatory cytokines
Injury disrupts skeletal architecture, blood ves-
including interleukins (IL-1, IL-6),
sels, periosteum, and adjacent soft tissue. The
platelet-derived growth factor (PDGF), and
response to injury initiates the inflammatory
tumor necrosis factor-alpha (TNF-a). These fac-
phase, characterized by the release of cytokines
tors recruit other inflammatory cells, promote
and chemoattractants that together initiate heal-
angiogenesis, recruit progenitor stem cells, and
ing and recruit progenitor cells.
induce their differentiation.
Following injury, hematoma occupies the Reparative Phase
fracture site. Fracture hematoma serves two key The reparative phase is characterized by the
functions. It provides a physical scaffold for deposition of extracellular matrix across the
subsequent occupation by progenitor cells, fracture site. It involves a tightly regulated
Fig. 2.4 Healing of stabilized fracture. Progressive heal- presence of new bone formation (green), there is minimal
ing of a stabilized tibia fracture in a mouse model staining for collagen type IIa expression (red), a marker of
demonstrates no callus formation on serial radiographs chondrogenesis. (SO/FG Safranin O/Fast Green stain).
(day 4 through day 21) or on histological staining. In the From Thompson et al. [13], with permission
sequence of events that ultimately stabilizes the Recruitment of Mesenchymal Stem Cells
fracture site with bridging bone. Following the
inflammatory phase, this phase begins with the The recruitment of MSCs is an essential com-
recruitment of mesenchymal stem cells. These ponent of fracture healing. MSCs reside
progenitors differentiate into osteogenic and throughout the body, including the periosteum,
chondrogenic cell lines, which produce soft car- bone marrow, trabecular bone, muscle, and sys-
tilaginous callus as a scaffold for bone healing. temic circulation [28]. Periosteal- and bone
Vascular ingrowth prompts the maturation of the marrow-derived MSCs were traditionally thought
fracture callus; the soft callus undergoes miner- to be the primary sources of progenitor cells in
alization, resorption, and ultimately replacement early fracture repair [29]. However, current data
by hard callus. The end result provides a stable suggests that other sources of MSCs, namely
bridge of bone across the fracture site. from muscle and systemic circulation, may also
Fig. 2.5 Healing in unstabilized fractures. In contrast to staining. Safranin O/Fast Green staining demonstrates
stabilized fractures, progressive healing of a stabilized abundant collagen type IIa expression (red), consistent
tibia fracture in a mouse model demonstrates abundant with robust chondrogenesis. From Thompson et al. [13],
callus formation on serial radiographs and on histological with permission
contribute to the progenitor cell population [28, factor (VEGF) production [31]. TNF-a and IL-6
30]. promote recruitment and differentiation of
Inflammation at the time of injury releases a muscle-derived stromal cells. TNA-a, at low
number of chemokines, growth factors, and sig- concentrations, also stimulates chondrogenic and
nals to recruit MSCs and other inflammatory osteogenic differentiation [32–34] (see
cells. In the early phase, TNF-a, IL-1, and IL-6 Table 2.1). In vivo injection of TNF-a acceler-
play key roles in chemotaxis, mesenchymal stem ates fracture healing and callus mineralization
cell (MSC) recruitment, and osteogenic and [32]. Conversely, the absence of TNF-a signal-
chondrogenic differentiation [14]. Peak levels of ing appears to delay both chondrogenic differ-
IL-1 and IL-6 are reached within the first 24 h, entiation and endochondral resorption [14, 24,
and then decline precipitously after 72 h. IL-1 34].
and IL-6 contribute to chemotaxis of other Emerging evidence has also supported the
inflammatory cells and of MSCs and promote role of stromal cell-derived factor (SDF-1) in
angiogenesis via vascular endothelial growth skeletal repair. SDF-1 is a potent chemoattractant
Fig. 2.6 Secondary healing with intramedullary device. formation. g, h 9-month follow-up, with progressive
The patient is a 23-year-old man who was struck by a callus formation and bone bridging across the tibial
motor vehicle at high speed and sustained right tibial and fracture. There is some callus at the fibula fracture ends,
fibular shaft fractures with associated compartment syn- but no bone bridging across the fracture site. i, j 3-year
drome. a, b Initial injury radiographs. c, d Immediate follow-up, with complete healing of tibial fracture, and
postoperative radiographs following tibia intramedullary nonunion of fibular fracture
nailing. e, f 2-month follow-up, demonstrate callus
expressed at sites of injury to recruit MSCs from upregulated in periosteum at the fracture site and
both circulating and local sources. Kitaori recruits MSCs that participated in the healing
demonstrated that SDF-1 expression is process. Additionally, blocking the function of
Fig. 2.7 Secondary healing with external fixator. The spanning external fixation. e, f 10-week follow-up, with
patient is a 51-year-old man who was struck by a vehicle interval removal of external fixator and cast application.
and sustained a Schatzker VI left tibial plateau fracture. a, There is bridging bone and progressive healing across the
b Initial injury radiographs. c, d Definitive treatment with fracture site
b Fig. 2.8 Secondary healing with bridge plating. The

Formation of Soft Cartilaginous Callus
patient is a 62-year-old man who was involved in a
motorcycle crash. He sustained a Grade I open left tibia By this time, the fracture hematoma has been
fracture. a, b Initial injury radiographs. c, d Initial converted to granulation tissue, containing
management consisted of external fixation, followed by
bridge plating across the fracture. e, f 17-month follow-up
inflammatory cytokines and growth factors that
after bridge plating, demonstrating bone healing across stimulate MSC differentiation, proliferation, and
fracture site production of extracellular matrix. The formation
of cartilaginous callus marks the initial attempts
SDF-1 significantly reduced bone formation, at achieving fracture union. The result is a cal-
indicating SDF-1 has a crucial role in fracture cified cartilaginous bridge that both provides
healing [35]. stability and creates a template for further
remodeling.
Table 2.1 Cytokines and their roles in fracture healing

Cytokine Effect
IL-1 Stimulates chemotaxis of inflammatory cells, MSCs
Promotes VEGF production and angiogenesis
IL-6 Stimulates chemotaxis of inflammatory cells, MSCs
Promotes VEGF production and angiogenesis
PDGF Released by platelets and inflammatory cells
Stimulates chemotaxis of inflammatory cells and osteoblasts
TNF-a Recruits MSCs during inflammatory phase
Regulates chondrocyte apoptosis, resorption of cartilage callus
Regulates bone remodeling, osteoclastogenesis
Stimulates chondrogenic and osteogenic differentiation
FGF Promote differentiation of fibroblasts, chondrocytes, myocytes, and osteoblasts
TGF-b Stimulates chemotaxis and proliferation of MSCs
Stimulates proliferation of chondrogenic and osteogenic cells
Induces production of extracellular matrix
MMP Degrades chondral and osseous extracellular matrix
VEGF Mediates neoangiogenesis
angiopoietin Regulates formation of larger vessels and branching of collateral branches from existing vessels
BMP Promote osteoblast differentiation and osteogenesis
Upregulates extracellular matrix production
Stimulate VEGF production
M-CSF Secreted by osteoblasts to induce osteoclast differentiation and proliferation
Upregulates RANK expression
OPG Inhibits osteoclast differentiation and activation
Inhibits osteoclast-mediated resorption
RANKL Stimulates osteoclastogenesis, osteoclast activation through its receptor RANK
Sclerostin BMP antagonist
IL interleukin; PDGF platelet-derived growth factor; TNF-a tumor necrosis factor-alpha; FGF fibroblast growth factor;
TGF-b transforming growth factor-beta; MMP matrix metalloproteinase; VEGF vascular endothelial growth factor;
BMP bone morphogenetic protein; OPG osteoprotegerin; RANK receptor-activated NF-jb; RANKL receptor-activated
NF-jb ligand. From Tsiridis et al. [24] with permission
Cartilaginous callus formation is driven by bone formation require adequate oxygen

growth factors, chondrocytes, fibroblasts, and delivery. Failure to do so leads to delayed
mechanical stimulation across the fracture site. healing. Revascularization is thus critical for
TGF-b and IGF-1 play primary roles in this stage progressive healing and bone formation [9, 11,
of chondrogenesis and endochondral bone for- 12, 36–38].
mation, stimulating the recruitment, proliferation, Two main molecular pathways regulate this
and differentiation of MSCs. BMPs also promote process: an angiopoietin-dependent pathway and
chondrogenesis. Several days after fracture, a VEGF-dependent pathway. Angiopoietins
chondrocytes derived from MSCs proliferate and promote formation of larger vessels and collat-
synthesize collagen. Starting from the periosteum eral vessels off existing vessels. VEGF promotes
and the fractured ends, chondrogenesis pro- endothelial cell differentiation, proliferation, and
gresses by appositional replacement of adjacent neoangiogenesis, and it mediates the principal
granulation tissue with cartilage matrix [29]. vascularization pathway [11, 24].
Fibroblasts produce fibrous tissue in areas with Inflammatory cytokines from early fracture
limited cartilage production. Micromotion across healing, particularly TNF-a, induce expression of
the fracture stimulates callus formation, and angiopoietin, allowing for early vascular
increased callus formation provides more ingrowth from existing periosteal vessels [9, 33].
mechanical stability to the fracture. When suffi- However, the primary vascularization process is
cient callus and stability have been attained, driven by VEGF. Following calcification of
roughly 2 weeks after fracture, chondrocytes cartilage callus, osteoblasts and hypertrophic
undergo hypertrophic differentiation. Prolifera- chondrocytes housed in callus express high
tion ceases. Collagen synthesis is downregulated. levels of VEGF, stimulating neoangiogenesis
Hypertrophic chondrocytes release vesicular into the avascular chondral matrix [36, 38, 39].
stores containing calcium, proteases, and phos- Concurrently, matrix metalloproteinases (MMPs)
phatases into the surrounding matrix. As the degrade calcified cartilage to facilitate ingrowth
collagen matrix is degraded, released phosphate of new vessels [40].
ions bind with calcium to promote cartilage cal-
cification. These calcium and phosphate deposits Hard Callus Formation
become the nidus for hydroxyapatite crystal
formation [8]. With the onset of neoangiogenesis, the next
At the same time, intramembranous ossifica- event is characterized by the transition from soft
tion occurs in areas of low strain, beneath the callus to hard callus: the removal of calcified
periosteum, and directly adjacent to the fractured cartilage and its replacement with woven bone
cortices. Within 24 h following injury, MSCs matrix. This process is mediated by MMPs,
from the bone marrow differentiate into BMPs, osteoclasts, chondroclasts, and osteo-
osteoblastic phenotypes. Proliferation and dif- blasts [36, 40, 41].
ferentiation peak at day 7–10. Woven bone is Osteoclasts have historically been considered
formed in these regions without a cartilage the key cell type in soft callus resorption. How-
scaffold. ever, more recent evidence suggests that resorp-
tion is nonspecific and mediated by multiple cell
Revascularization and Angiogenesis lines, including osteoclasts and chondroclasts
alike, and by MMP expression [40, 41]. This has
Fracture healing begins in a relatively hypoxic been supported by findings that impaired osteo-
environment; injury to vessels, periosteum, and clast function does not necessarily impair heal-
soft tissue compromises local blood supply ing. In an osteoclast-deficient osteopetrosis
[22]. Early cartilage callus can form in this mouse model, there was no difference in callus
hypoxic environment. However, as healing remodeling or union rates compared with control
progresses, subsequent callus remodeling and mice [42].
Cartilage callus is removed and subsequently progenitors express the complementary receptor,
replaced by woven bone. Mature osteoblasts RANK. Upon contact, RANKL interacts with
secrete osteoid, a combination of type I collagen, RANK to induce fusion of osteoclast progenitors
osteocalcin, and chondroitin sulfate. Collagen and thus produce mature multinucleated osteo-
fibrils are randomly oriented, producing an clasts. Alternatively, osteoblasts can also secrete
irregular structure known as woven bone [41]. OPG, which acts as a decoy by binding RANK
and consequently disrupts RANKL–RANK
Remodeling Phase
interactions. By modulating RANKL and OPG
While woven bone provides more biomechanical
expression, osteoblasts can tightly regulate
stability than fibrous tissue and soft callus, its
osteoclast activation. Osteoblasts express and
irregular and disordered structure is mechanically
secrete M-CSF, which induces osteoclast pre-
inferior to native cortical bone. Further remod-
cursor proliferation and differentiation. Addi-
eling is required to restore structural integrity.
tionally, M-CSF upregulates the expression of
The final phase of fracture healing converts
RANK on osteoclast precursors [43–45].
irregular woven bone into structured lamellar
bone. The process encompasses both catabolic Metaphyseal Fracture Healing
and anabolic mechanisms, regulated by the The principles underlying fracture healing have
coordinated relationship between osteoblasts and largely been based on diaphyseal models. By
osteoclasts. Whereas the earlier phases take place comparison, the existing literature for metaphy-
over the course of days to weeks, this final phase seal healing is limited. Metaphyseal bone differs
spans months to years after injury [9]. from diaphyseal bone in anatomy and biologic
activity. Periosteum is thicker around the meta-
Remodeling is characterized by woven bone physis. Blood supply is richer to the metaphysis
resorption followed by lamellar bone formation. [12]. Additionally, metaphyseal bone has a larger
Osteoclasts are multinucleated polarized cells that active bone surface area with consequently
attach to mineralized surfaces. At sites of attach- higher bone turnover rates [46].
ment, osteoclasts form ruffled borders, effectively
increasing surface area through which lysosomal Diaphyseal bone healing hinges on the inter-
enzymes and hydrogen ions are secreted. Enzymes relationship between biomechanics and biology.
degrade the organic collagen components, while Early in the healing process, the mechanical
the acidic milieu demineralizes the bone matrix. environment determines the biologic response,
The erosive pits left by the osteoclasts are termed whether healing will proceed by direct or indirect
“Howship’s lacuna.” Following resorption, means. In stable situations, healing proceeds
osteoblasts form new bone within these lacunae. directly to osteogenesis. In unstable conditions,
This process progresses along the length of hard healing begins with chondrogenesis. The same
callus, layer upon layer, replacing woven bone holds true for metaphyseal healing. Under rigidly
with lamellar bone [43, 44]. stable conditions, newly formed bone bridges the
Activation and regulation of remodeling fracture gap with minimal chondrogenic tissue,
depends on intimate coupling between osteo- similar to direct healing. Under more flexible
blasts and osteoclasts. Osteoblasts initiate conditions, bone intermixed with islands of
remodeling by producing factors to stimulate chondrogenic tissue forms across the gap, anal-
osteoclastogenesis and osteoclast function. The ogous secondary healing. Interestingly, both sit-
principle cytokines secreted by osteoblasts are uations do not generate a significant amount of
M-CSF, receptor-activated NF-jb ligand external callus [47]. Whereas progenitor cells
(RANKL), and osteoprotegerin (OPG). M-CSF need to be recruited in diaphyseal healing, the
and RANKL are essential for osteoclast forma- metaphysis houses a large reservoir of precursor
tion. Osteoblasts express RANKL on their cell cells, obviating the need for a large periosteal
membranes, whereas mononuclear osteoclast reaction and MSC recruitment [48].
2.2.2 Biomechanics of Fracture histomorphometric and finite element analyses

Healing [15, 49, 50]. In models of indirect healing,
intramembranous bone formation occurs at the
The relationship between mechanics and biology periosteum and directly adjacent to the cortex,
is well established in skeletal physiology. areas characterized by low strain. Cartilaginous
Wolff’s law stipulates that bone structurally callus developed between the fractured ends, in
adapts to its loading conditions. Likewise, areas of high strain. Increasing the mechanical
biomechanics plays a central role in skeletal stress and strain, by early loading or delayed
repair. Following injury, the mechanical envi- stabilization, impairs bone bridging and delayed
ronment influences the biologic healing response. healing across the fracture [51, 52]. Histological
This response in turn attempts to restore skeletal analysis in these animal models of delayed sta-
integrity. Understanding how biomechanical bilization demonstrated higher proportions of
factors affect healing is therefore fundamental to cartilage and fibrous tissue in the fracture site
fracture treatment. The existing body of literature compared to fractures that were stabilized early
has identified three mechanical parameters that (Fig. 2.9) [53]. Similarly, Augat demonstrated in
impact fracture healing: interfragmentary strain, a sheep model that higher gap sizes and higher
gap size, and hydrostatic pressure. The degree to strains led to lower amounts of bone formation
which these parameters affect healing, and the and higher proportions of connective tissue and
timing at which they are applied, will be dis- fibrocartilage formation across the fracture
cussed in this section. (Fig. 2.10) [49].
2.2.2.1 Interfragmentary Strain 2.2.2.2 Fracture Gap

Perren’s strain theory proposes that “a tissue While the strain theory accounts for some of the
cannot be produced under strain conditions which clinical observations seen in fracture healing,
exceed the elongation at rupture of the given tis- further work has shown that strain is not the only
sue element” [16]. Thus, bone can only form in determinant of tissue differentiation. Fracture gap
low strain environments, while fibrous tissue can is as important, if not more important, than strain.
form in high strain environments. In stable frac- Augat et al. and Claes et al. examined the effects
tures, a low strain environment allows for primary of increasing gap size (1, 2, and 6 mm) and
osteogenesis across the fracture gap. However, in different strains (7 vs. 31%) on bone healing and
unstable fractures, high strains preclude direct mechanical strength. Augat demonstrated in a
bone formation. Instead, precursor tissues must sheep model that higher gap sizes and higher
first bridge the gap, providing adequate mechan- strains led to lower amounts of bone formation
ical stability for osteogenesis to ultimately occur. and higher proportions of connective tissue and
Such is the case with endochondral bone forma- fibrocartilage formation across the fracture
tion. Cartilage callus first bridges the gap and (Fig. 2.10) [49]. Increasing gap correlated with
provides provisional stability across the fracture. less bone formation. Cases in which bone failed
When sufficient stability has been attained, the to bridge the fracture gap were only observed for
cartilage callus can then undergo calcification, gaps >2 mm. Regardless of interfragmentary
and woven bone can replace the chondral matrix. strain, gaps of 6 mm never healed. Strain played
If strain is still too high, more callus is produced, a more subtle role. While there was no difference
increasing its diameter and effectively increasing in mechanical properties between strain groups,
its strength. If strain still remains too high, bone those that experienced higher strain (31%) had
bridging may not occur and a fibrous nonunion higher cartilage and fibrous tissue content, and
may develop instead. lower bone content [49, 50]. Additionally,
The relationship between strain and tissue hydrostatic pressure and local stress play a role in
differentiation correlates with both tissue differentiation.
Fig. 2.9 Histological findings in impaired healing. Nonstabilized fractures (e) demonstrate increased cartilage
formation compared to stabilized fractures (d). From Miclau et al. [53] with permission
2.2.2.3 Timing in Fracture Healing healing outcome [14]. Immediate and early full
Fracture healing involves a complex temporal weight bearing in a sheep model has been shown
and spatial sequence of events. The timing at to delay healing, demonstrating lower bone
which mechanical stimulation is introduced content compared to delayed weight bearing
appears to affect the outcomes of skeletal repair. [51]. Others have likewise shown that early or
The initial mechanical environment is an early immediate mechanical loading led to decreased
determinant of tissue differentiation and of bone formation and inferior mechanical
Fig. 2.10 Influence of fracture gap size and strain on tissue and fibrocartilage at the fracture site and within the
tissue differentiation. Tissue differentiation as a function callus. Conversely, low strains and gaps had higher
of fracture gap size and strain. With higher gaps and amounts of bone formation. From Augat et al. [49], with
strains, there is an increasing proportion of connective permission
properties [52, 54]. These same studies also of fracture healing, and new scoring systems are
showed that delayed loading led to higher pro- refining our ability to assess fracture healing.
portions of bone formation and improved
biomechanical properties. Miclau et al. showed 2.2.3.1 Clinical Criteria
that delayed stabilization for even 24 h in mice Physical examination and clinical evaluation
led to higher cartilage callus formation and lower remain the cornerstone of fracture healing
bone content compared to those who had assessment. Weight bearing status has been
immediate stabilization [53]. Taken together, shown to correlate with fracture tissue stiffness
these findings demonstrate that timing of [58], though the clinicians’ ability to assess
mechanical loading impacts fracture healing. stiffness is not reliable [59]. Weight bearing
When loading occurs prematurely or exceeds without pain is the most commonly endorsed
tolerable amounts, it can disrupt early healing factor, used in over half of all published studies
and have deleterious effects. However, with cal- to assess healing [57]. Pain at the fracture site
lus providing some inherent stability across the and tenderness to palpation are also important
fracture site, loading is better tolerated and may signs in assessing healing. Conversely, the lack
stimulate further callus formation and bony of weight bearing is considered the most impor-
healing. tant clinical criteria for impaired healing.
2.2.3.2 Radiologic Scores

2.2.3 Assessment of Fracture Healing The Radiographic Union Score for Hip (RUSH)
and the Radiographic Union Score for Tibia
The accurate assessment of fracture union is (RUST) were developed to provide standard-
often a difficult undertaking, but nonetheless ized, reliable radiographic measures of fracture
fundamental to clinical practice and research. healing [60–63]. These scoring systems evaluate
Nonunions can be a source of significant dis- healing on the basis of cortical bridging and
ability, and its early diagnosis and treatment is fracture line visibility on AP and lateral views
paramount to improving patients’ quality of life (Table 2.2; Figs. 2.11 and 2.12). Both RUST and
and return to function [55]. The definition of RUSH have high interobserver agreement, with
nonunion provided by the United States Food intraclass correlation coefficients of 0.86 and
and Drug Administration (FDA) requires a min- 0.85, respectively. Compared to subjective
imum of at least nine months to elapse since the assessment, these scores increase reliability and
initial injury and no signs of healing for the final agreement among clinicians in assessing radio-
three months. Yet, there are no standardized graphic progression of fracture healing [62–65].
methods of assessing fracture union, and there The lack of consensus in the orthopedic
still remains considerable variability among community limits the ability to establish consis-
clinicians and researchers alike [56, 57]. How- tent criteria to define union. Most practices use a
ever, advances in imaging techniques, improved combination of clinical and radiographic criteria
knowledge about the biology and biomechanics to assess fracture healing. Additionally, several
serologic markers of bone metabolism and
Table 2.2 Calculation of Score per cortex Callus Fracture line

RUST and RUSH scores
1 Absent Visible
2 Present Visible
3 Present Invisible
The RUST and RUSH scores are based on radiographic findings on AP and lateral
projections. Each cortex is scored according to the presence of callus and visibility of
fracture line, with a maximum score of 12 for 4 cortices
cytokines, including TGF-b, have been identified evidence has thus far been limited and inconclusive.
as candidate biomarkers for tracking healing D’Ippolito et al. [71] demonstrated lower numbers
progression [8, 66]. Tools to measure mechanical of MSCs with osteogenic potential in adult human
properties in healing bone are also being devel- vertebrae. In contrast, Stenderup et al. [72] found no
oped. As our understanding of fracture healing age-related decrement in the number of osteogenic
continues to evolve, so too will our ability to stem cells from iliac crest marrow. The effects of
gauge the healing process. age on fracture healing in humans, independent of
other associated variables such as metabolic bone
diseases, require further investigation.
2.3 Modulation of Fracture Healing
2.3.1.2 Metabolic Bone Disease
2.3.1 Comorbidities Osteoporosis
Osteoporosis is the most common metabolic
2.3.1.1 Aging bone disease, affecting over 200 million people
Aging has profound effects on bone health, worldwide [73]. Unlike normal age-related
modeling, and repair. Bone mass declines with changes, osteoporosis is a metabolic disease
advancing age, owing in part to hormonal characterized by decreased bone mass, decreased
changes, limited physical activity, and altered mineral content, increased porosity, and com-
biologic responses. Additionally, elderly patients promised microarchitecture. On a cellular level,
have a higher prevalence of comorbidities and the balance between anabolic and catabolic pro-
take more medications, some of which may cesses is unhinged to favor net bone resorption.
directly impact bone healing. Clinically, the weakened architecture predisposes
Animal studies have demonstrated decreased to fragility fractures. Almost half of women with
fracture healing capacity with increasing age osteoporosis will sustain at least one fragility
[67]. Compared to adults, juveniles exhibit faster fracture in their lifetime [73].
healing rates and remodeling potential [68]. In
murine models, juveniles had more robust peri- Osteoporotic fractures are challenging to treat.
osteal responses, higher chondrocytic and Appropriate management requires an apprecia-
osteoblastic differentiation, and faster healing tion of how osteoporosis affects bone health, bone
rates [67]. Additionally, juveniles mounted a quality, and healing. As most clinical studies have
larger angiogenic response, illustrated by higher focused on medical management and fracture
VEGF, HIF-1a, and MMP expression [69]. In prevention, there is limited data on how osteo-
contrast, adults had relative delays in endo- porosis influences fracture healing in humans.
chondral ossification, decreased periosteal thick- More recently, Nikolaou et al. assessed the effect
ness, and decreased chondrogenic potential in the of osteoporosis on healing time in patients with
periosteum [46]. Furthermore, skeletal maturity femoral shaft fractures following intramedullary
brought on a sharp drop in regenerative potential nailing. The elderly group of patients with radi-
[67]. Additionally, elderly mice demonstrated ologic evidence of osteoporosis had delayed
decreased angiogenic potential [69]. In a murine healing compared to a younger cohort
model of senile osteoporosis, bone (19.4 weeks versus 16.2 weeks, respectively),
marrow-derived MSCs had increased adipogenic though this difference is probably not clinically
and decreased osteogenic differentiation. Despite significant [74].
these abnormalities, the process of fracture Animal studies have shown that osteoporosis
healing was unchanged [70]. impairs fracture healing. In an ovariectomized rat
How aging affects fracture healing after skeletal osteoporosis model, Namkung-Matthai et al. [75]
maturity remains controversial, and the clinical demonstrated early failure in the repair process
Fig. 2.11 Radiographic union score for hip (RUSH) 4. c, d 6-week follow-up radiographs, with a RUSH = 8,
fracture healing assessment, Assignment of RUSH in a demonstrating callus on the anteroposterior view and
patient who sustained a left intertrochanteric fracture. a, lateral views, though the fracture lines are still visible
b Immediate postoperative radiographs, with a RUSH =
Fig. 2.12 Radiographic

union score for tibia (RUST)
fracture healing assessment.
Assignment of RUST in a
patient with distal tibial shaft
fracture at 3 months. a At
4 weeks, there is healing
callus along the medial,
lateral, and anterior cortices,
but fractures lines are visible.
RUST score = 8. b At
10 weeks, there is bridging
callus and no fracture line at
the anterior and medial
cortices. Fracture lines are
still visible posteriorly and
laterally. RUST score = 10
with a 40% reduction in callus size, and patients [80]. Healing time in nondisplaced
decreased bone mineral density and strength. fractures was prolonged by 87% in
Walsh et al. [76] demonstrated delayed healing non-neuropathic diabetic patients compared to
and decreased tensile and bending strength in nondiabetic patients [81].
estrogen-deficient rats. Lill et al. likewise
demonstrated decreased bending stiffness and Diabetes is a chronic inflammatory disorder;
delayed healing in their osteoporotic sheep type I is an autoimmune disorder against
model. However, final strength at the end of insulin-producing islet of Langerhans beta cells,
healing was not different from healthy sheep [77]. while type II is associated with obesity-related
To what degree osteoporosis impairs fracture inflammation. Acute inflammation plays a piv-
healing remains unclear. While the healing otal role in early fracture healing in recruiting
potential is present in patients with osteoporosis, skeletal progenitors to the site of injury. How-
it may not be as robust. Furthermore, concomitant ever, these events are tightly regulated; inflam-
comorbidities such as vitamin D deficiency or matory cytokine levels are active within the first
other disorders of calcium homeostasis in these 72 h after injury, and at specific points in the
patients may also impair the healing response. healing cascade. Continued inflammation and
continued cytokine expression, left unchecked,
2.3.1.3 Endocrine Disorders can halt the progression of bone remodeling and
Hyperparathyroidism, thyroid disorders, and fracture healing [82, 83].
hypogonadism have also been shown to impair Recent evidence from animal studies suggests
fracture healing [78, 79]. In patients with unex- that uncontrolled diabetes may directly impact
plained nonunions, Brinker et al. found a high callus formation, chondrocyte survival, and
prevalence of these metabolic and endocrine osteoclast activity. Hyperglycemia upregulates
disorders that had previously been unrecognized. the expression of proinflammatory factors, such
The mechanisms by which these impede the as TNF-a and VEGF [82]. Upregulation of
healing process are still undetermined. However, TNF-a stimulates chondrocyte apoptosis. Addi-
medical management of the underlying abnor- tionally, diabetes is associated with premature
mality, in conjunction with surgical fixation, resorption of the cartilaginous callus and
successfully treats the majority of cases [78]. increased osteoclastogenesis. Impaired matrix
While routine screening is not indicated in the synthesis, chondrocyte dysfunction, and prema-
acute setting, impaired healing in otherwise ture resorption all decrease callus formation.
appropriately treated fractures warrants further These mechanisms may explain its weaker
evaluation for metabolic abnormalities. biomechanical strength in diabetic fracture heal-
ing [83–86].
Diabetes Mellitus Glycemic control should be the cornerstone of
Diabetes mellitus poses significant challenges to fracture management in diabetic patients. It has
fracture management through impairment of repeatedly been shown to reduce or prevent the
healing, protective sensation, and host immunity. aforementioned issues with bone healing [87].
These effects are mediated by incompetent Successful fracture healing in these patients often
microcirculation, and in severe cases, they may requires prolonged immobilization and weight
also be associated with peripheral vascular dis- bearing precautions [80]. Soft tissue management
ease. Delayed fracture healing in diabetic patients is also paramount, particularly in those with
has been well documented. Early observations by peripheral neuropathy. Surgical interventions
Cozen showed significantly delayed fracture likewise should respect soft tissue coverage;
healing and nonunions in a series of diabetic aggressive dissection and inattentiveness to soft
tissue handling may further compromise the nicotine replacement may be safe and would
already tenuous blood supply in diabetic patients reduce exposure to inhaled CO and other chem-
[81, 85, 88]. icals that may pose more physiologic harm.
2.3.2.2 Alcohol Consumption

2.3.2 Habits Alcoholism and binge drinking are
well-documented risk factors for traumatic inju-
2.3.2.1 Smoking ries, disrupted bone metabolism, and impaired
Smoking is well known to impair fracture heal- fracture healing. Not only does alcohol abuse
ing. In multiple clinical trials, smoking has con- confer higher fracture risk [97], but it also pro-
sistently been associated with nonunion, longs healing times. Nyquist et al. [98] showed
pseudarthrosis, and delayed healing. In the that alcohol abusers with transverse tibia frac-
Lower Extremity Assessment Project (LEAP), tures required longer healing times than nonal-
smokers, both former and active, were 32 and coholic patients. Alcoholic patients have lower
37% more likely to develop nonunion, respec- bone mineral density and abnormal bone turn-
tively. Smokers also required longer healing over markers consistent with defective bone
times [89, 90]. For midshaft clavicle fractures, formation and osteoblast dysfunction [99, 100].
smoking was the strongest risk factor for non- Furthermore, alcoholism is frequently paired
union [91]. Among distal tibia fractures treated with smoking and malnutrition, which may fur-
with two-ring hybrid external fixators, smoking ther compromise bone health and bone repair
delayed union by 10 weeks [92]. Additional, [101].
smoking has been associated with higher com- Alcohol exposure predominantly affects early
plication, reoperation, and infection rates [89, repair and bone formation [102–104]. In vitro
93]. osteoblast cultures demonstrate decreased pro-
Cigarette smoke contains hundreds of chemi- liferation and osteoid synthesis when exposed to
cals and gases, among them nicotine, carbon ethanol. Additionally, rodent models have
monoxide, and carcinogens. Carbon monoxide demonstrated decreased mechanical properties in
impairs oxygen delivery, creating a hypoxic fracture repair tissue following alcoholic inges-
environment for tissues. Nicotine induces vaso- tion [103]. In ethanol-fed rats, there was absence
constriction, likewise impairing oxygen delivery of mineralized callus on radiographs while in
to tissues. Recent studies have found a bimodal ethanol-free controls there was complete healing
dose-dependent effect of nicotine on osteoblasts. [105]. Recent evidence demonstrates that pro-
At high concentrations, nicotine had an inhibi- duction of inflammatory cytokines, including
tory effect on osteoblast proliferation and differ- IL-1 and TNF-a, increased oxidative stress, and
entiation, but at lower doses, it actually impaired Wnt signaling may mediate these
stimulated osteoblast activity [94]. While con- effects [104].
sidered the addictive constituent in cigarettes, the Just as acute ingestion can lead to impaired
role of nicotine in impaired fracture healing has healing, abstinence can lead reversal of its effects
undergone re-evaluation [95, 96]. Tobacco [102, 103]. Laitinen et al. [99] found that bone
extract without nicotine reduced the mechanical formation markers improved to near control
strength in healing femoral fractures compared to levels after two weeks of abstinence. More recent
nicotine alone [96]. The negative effects of evidence also suggests a role for antioxidant
smoking toward fracture healing are likely due to treatment with N-acetylcysteine in reversing the
other constituents in cigarette smoke rather than negative healing effects of alcohol consumption
from nicotine itself. These studies suggest that [106].
2.3.3 Medications 115, 116]. Aspirin, at doses equivalent to

325 mg, similarly delayed fracture healing,
2.3.3.1 Nonsteroidal Anti-inflammatory though smaller doses did not demonstrate any
Drugs radiographic or mechanical differences compared
Inflammation is critical in fracture healing. As with controls [116].
part of the inflammatory cascade, cyclooxyge- The importance of COX-2 and prostaglandins
nase (COX) converts arachidonic acid into in fracture healing has been clearly established.
prostaglandins [107, 108]. Downstream, pros- While the mounting evidence in animal studies
taglandin E2 (PGE2) stimulates bone metabolism, supports the effect of NSAIDs in suppressing
bone formation, and maintenance [108, 109]. fracture healing, translation of these effects to
Deficient PGE2 signaling conversely leads to human subjects remains less convincing. As
osteopenia and impaired bone healing [110]. such, there is currently inadequate clinical evi-
Additionally, COX-2 is essential to fracture dence to prohibit their routine use in acute frac-
healing, mediating repair through osteogenesis. ture care. NSAIDs remain an important feature in
COX-2 knockout mice fail to form mineralized the development of a multimodal, opiate-sparing
matrix during endochondral ossification, where approach to postinjury and postsurgical pain
COX-1 knockout mice display no disruption in regimen, and further clinical work is paramount
healing [111]. in understanding its effects in orthopedic
Nonsteroidal anti-inflammatory drugs patients.
(NSAIDs) exert their analgesic effect by inter-
fering with prostaglandin production and COX 2.3.3.2 Bisphosphonates
function. NSAIDs have long been used as pro- Bisphosphonates are a mainstay of antiresorptive
phylaxis for heterotopic ossification, and human osteoporosis treatment. This class of drugs acts
studies suggest adverse effects of NSAIDs on by inhibiting osteoclast-mediated resorption,
fracture repair. However, these clinical studies improving bone mass and mineralization. How-
are all level III-IV data, have been retrospective, ever, there have been concerns about the hypo-
and have produced conflicting results [107, 108, thetical risk that bisphosphonates may impair
112, 113]. Giannoudis et al. correlated NSAID bone healing. The reparative process relies on
use > 4 weeks with higher rates of nonunion in osteoclast-mediated remodeling of hard callus
femoral shaft fractures treated with intramedul- into woven bone and woven bone into mature
lary nailing. Even short-term use demonstrated lamellar bone.
delayed union [112]. However, this study was Clinical studies have reported mixed results.
largely limited by its retrospective nature and In a retrospective review of humeral fractures,
lack of controls; whether true causality exists Solomon reported a higher nonunion rate with
cannot be extrapolated from these results. bisphosphonate use in the postfracture period.
In animal studies, NSAIDs do appear to However, the conclusions of this study should be
negatively affect skeletal repair [107, 108, 111, tempered with its limitations, including the rare
114]. The earliest of these studies demonstrated occurrence of fractures (0.4%) and its retrospec-
that indomethacin treatment not only reduced the tive design [117]. Rozental et al. explored the
mechanical properties of rat femora during frac- effect of bisphosphonate use on distal radius
ture healing, but also created fibrous tissue rather fracture healing time. Patients treated with bis-
than callus between fractured ends [114]. Sub- phosphonates had slightly longer healing times
sequent studies have also shown that the use of (55 days versus 49 days), but this difference,
both nonselective and COX-2 selective NSAIDs while statistically significant, was not considered
decreases bone formation and cortical bridging, clinically significant [118]. More recently, Gong
prolongs healing times, and increases rates of similarly investigated the impact of bisphospho-
nonunion [111, 115]. These effects do appear to nate treatment on healing in distal radius frac-
be both time and dose dependent [14, 107, 108, tures after surgical fixation. There was no
difference in time to union, or in radiographic or The short-term results of bisphosphonate use

clinical outcomes [119]. In a randomized, postfracture are encouraging. Clinical and basic
double-blind, placebo-controlled trial using science studies have not shown major differ-
zoledronic acid after hip fracture, Lyles et al. did ences in healing with bisphosphonate use.
not find any evidence of delayed healing. Fur- However, its long-term effects remain unclear.
thermore, if administered within 90 days after Furthermore, the emergence of atypical femur
surgical fixation, zoledronic acid improved sur- fractures associated with long-term bisphos-
vival and reduced the incidence of new clinical phonate use has raised safety concerns
fractures [120]. (Fig. 2.13) [125]. These fractures have a
Thus far, animal studies have been largely reported prolonged healing course [126, 127].
reassuring and have not demonstrated a detri- As these fractures occur in the subtrochanteric
mental effect of bisphosphonates on fracture region, an area subject to high stress and prone
healing. Rather, animals treated with bisphos- to malunion, it is difficult to ascertain whether
phonates had increased callus formation and these healing issues are a result of the fracture
mineralization. Others have demonstrated some or a result of the drug effect. Additionally, while
evidence of delay in callus remodeling and true causality has yet to be determined, the FDA
resorption, though there was no long-term impact has proposed offering a drug holiday for certain
on healing [121–124]. lower risk patients, though concrete guidelines
Fig. 2.13 Atypical femur fracture related to bisphospho- sustained a low-energy right femur subtrochanteric
nate use. The patient is a 43-year-old woman with a oblique fracture after twisting that leg, consistent with
history of metastatic breast cancer status postlumpectomy an atypical femur fracture. a Imaging of the contralateral
and hormone therapy. She had a long history of bispho- leg demonstrated stress reaction in the subtrochanteric
sphonate use. A recent positron emission tomography region, concerning for an impending pathologic fracture
scan did not demonstrate any bony metastases. She (b)
defining appropriate candidates have not been functional scores, with lower pain scores and
established [128]. faster “Timed Up and Go” testing compared to
untreated patients [137].
2.3.3.3 Parathyroid Hormone Analogs
Parathyroid hormone regulates serum calcium
homeostasis via intestinal absorption, renal 2.4 Conclusion
secretion, and bone metabolism. In the skeletal
system, PTH binds to and stimulates osteoblasts In conclusion, fracture healing is a highly com-
to form new bone. Continuous PTH stimulation plex temporally and spatially coordinated process
increases RANKL expression and decreases to restore mechanical integrity to bone following
OPG expression, increasing osteoclast formation trauma. Appropriate management of both acute
and catabolic function. However, intermittent fractures and nonunions requires a comprehen-
PTH exposure preferentially stimulates anabolic sive understanding of the principles that govern
osteoblast activity [124, 129]. healing. This includes the biologic factors, the
Teriparatide, the biologically active 1–34 mechanical factors, and their interdependence.
fragment of recombinant human PTH, is the first Previous work has concentrated on optimizing
anabolic medication approved for osteoporosis the mechanical environment for healing to occur,
[124], and its applications in fracture care are driving new innovations in implant design and
currently being investigated [130]. Animal stud- function. More recently, the focus has shifted
ies in both rodent and simian models support toward optimizing the biologic environment. The
PTH’s role in enhancing fracture healing. In goal of fracture care is to achieve union in order
rodent models, PTH appears to accelerate healing to restore patients’ functionality and livelihood.
during chondrogenesis. PTH treatment elevates To this end, our treatment strategies in fracture
chondrogenic gene expression, cell recruitment, care will continue to evolve in stride with our
and differentiation, while osteogenic gene growing understanding of fracture healing as well
expression was not significantly increased. as its impact on patient-important outcomes such
Additionally, PTH stimulates earlier chondrocyte as health related quality of life and function.
hypertrophy and maturation of cartilage callus
[131, 132]. Andreassen demonstrated increased
fracture site strength and improved bone mineral References
content with PTH administration in a
dose-dependent manner [133, 134]. Similarly, in 1. Spiegel DA, Gosselin RA, Coughlin RR,
monkeys, higher dose PTH treatment had smaller Joshipura M, Browner BD, Dormans JP. The bur-
den of musculoskeletal injury in low and
callus sizes, consistent with accelerated remod- middle-income countries: challenges and opportu-
eling of callus to lamellar bone [135]. nities. J Bone Joint Surg Am. 2008;90(4):915–23.
Early clinical results, while limited, have also 2. United States Bone and Joint Initiative. Injuries.
been encouraging. In a prospective, randomized The burden of musculoskeletal diseases in the
United States (BMUS), 3rd ed. Rosemont, IL; 2014.
control trial, placebo, 20 lg teriparatide or 40 lg Available at http://www.boneandjointburden.org.
teriparatide was administered following distal Accessed 31 Aug 2015.
radius fracture. Interestingly, median time to 3. Burge R, Dawson-Hughes B, Solomon DH,
cortical bridging was significantly shorter in the Wong JB, King A, Tosteson A. Incidence and
economic burden of osteoporosis-related fractures
20 lg group (7.4 weeks) compared to both plain the United States, 2005–2025. J Bone Miner Res.
cebo (9.1 weeks) and 40 lg (8.8 weeks) groups 2007;22(3):465–75.
[136]. In a prospective clinical trial of pelvic 4. Tzioupis C, Giannoudis PV. Prevalence of
fractures using CT to evaluate fracture union, long-bone non-unions. Injury. 2007;38(Suppl 2):
S3–9. Erratum in Injury. 2007;38(10):1224.
PTH treatment decreased healing time to 5. Brinker M, O’Connor DP. The incidence of
7.8 weeks, compared to 12.6 weeks for controls. fractures and dislocations referred for orthopaedic
Additionally, PTH-treated patients had better
services in a capitated population. J Bone Joint haematomas found at the site of human fractures
Surg. 2004;86-A(2):290–7. contain progenitor cells with multilineage capacity.
6. Dahabreh Z, Dimitriou R, Giannoudis PV. Health J Bone Joint Surg Br. 2007;89(1):133–8.
economics: a cost analysis of treatment of persistent 24. Tsiridis E, Upadhyay N, Giannoudis P. Molecular
fracture non-unions using bone morphogenetic aspects of fracture healing: which are the important
protein-7. Injury. 2007;38(3):371–7. molecules? Injury. 2007;38(Suppl 1):S11–25.
7. Kanakaris NK, Giannoudis PV. The health eco- 25. Sarahrudi K, Thomas A, Albrecht C, Aharinejad S.
nomics of the treatment of long-bone non-unions. Strongly enhanced levels of sclerostin during
Injury. 2007;38(Suppl):S77–84. human fracture healing. J Orthop Res. 2012;30
8. Einhorn TA. The cell and molecular biology of (10):1549–55.
fracture healing. Clin Orthop Relat Res. 1998;355 26. Sarahrudi K, Thomas A, Mousavi M, Kaiser G,
(Suppl):S7–21. Kottstorfe J, Kecht M, et al. Elevated transforming
9. Marsell R, Einhorn TA. The biology fracture growth factor-beta 1 (TGF-ß1) levels in human
healing. Injury. 2011;42(6):551–5. fracture healing. Injury. 2011;42(8):833–7.
10. Giannoudis PV, Einhorn TA, Marsh D. Fracture 27. Sarahrudi K, Mousavi M, Thomas A, Eipeldauer S,
healing: the diamond concept. Injury. 2007;38(3 Vecsei C, Pietschmann P, Aharinejad S. Elevated
Suppl):S3–6. levels of macrophage colony-stimulating factor in
11. Hankenson KD, Dishowitz M, Gray C, human fracture healing. J Orthop Res. 2010;28
Schenker M. Angiogenesis in bone regeneration. (5):671–6.
Injury. 2011;42(6):556–61. 28. Bielby R, Jones E, McGonagle. The role of
12. Rhinelander FW. Tibial blood supply in relation to mesenchymal stem cells in maintenance and repair
fracture healing. Clin Orthop Relat Res. of bone. Injury. 38(Suppl 1):S26–32.
1974;105:34–81. 29. Malizos KN. Paptheodorou. The healing potential
13. Thompson Z, Miclau T, Hu D, Helms JA. A model of the periosteum. Injury. 2005;36(Suppl 3):S13–9.
for intramembranous ossification during fracture 30. Shah K, Majeed Z, Jonason J, O’Keefe RJ. The role
healing. J Orthop Res. 2002;20(5):1091–8. of muscle in bone repair: the cells, signals, and
14. Pape HC, Marcucio R, Humphrey C, Colnot C, tissue responses to injury. Curr Osteoporos
Knobe M, Harvey EJ. Trauma-induced inflamma- Rep. 2013;11(2):130–5.
tion and fracture healing. J Orthop Trauma. 2010;24 31. Yang X, Ricciardi BG, Hernandez-Soria A, Shi Y,
(9):522–5. Pleshko Camacho N, Bostrom MP. Callus miner-
15. Claes LE, Heigele CA. Magnitudes of local stress alization and maturation are delayed during fracture
and strain along bony surfaces predict the course healing in interleukin-6 knockout mice. Bone.
and type of fracture healing. J Biomech. 1999;32 2007;41(6):928–36.
(3):255–66. 32. Glass GE, Chan JK, Freidin A, Feldmann M,
16. Perren SM. Evolution of the internal fixation of long Horwood NJ, Nanchahal J. TNF-alpha promotes
bone fractures. J Bone Joint Surg Br. 2002;84 fracture repair by augmenting the recruitment and
(8):1093–10. differentiation of muscle-derived stromal cells. Proc
17. Rahn BA, Gallinaro P, Baltensperger A, Perren SM. Nat Acad Sci U S A. 2011;108(4):1585–90.
Primary bone healing. An experimental study in the 33. Lehmann W, Edgar CM, Wang K, Cho TJ, Bar-
rabbit. J Bone Joint Surg. 1971;53(4):783–6. nes GL, Kakar S, et al. Tumor necrosis factor alpha
18. Schenck R, Willenegger H. On the histological picture (TNF-A) coordinately regulates the expression of
of so-called primary healing of pressure osteosynthesis specific matrix metalloproteinases (MMPs) and
in experimental osteotomies in the dog. Experientia. angiogenic factors during fracture healing. Bone.
1963;19:593–5. [article in German]. 2005;36:300–10.
19. Perren SM. Physical and biological aspects of 34. Gerstenfeld LC, Cho TJ, Kon T, Aizawa T, Tsay A,
fracture healing with special reference to internal Fitch J, et al. Impaired fracture healing in the
fixation. Clin Orthop Relat Res. 1979;138:175–96. absence of TNF-alpha signaling: the role of
20. Shapiro F. Cortical bone repair: the relationship of TNF-alpha in endochondral cartilage resorption.
the lacunar-canalicular system and intercellular gap J Bone Miner Res. 2003;18(9):1584–92.
junctions to the repair process. J Bone Joint Surg. 35. Kitaori T, Ito H, Schwarz EM, Tsutsumi R,
1988;70(7):1067–81. Yoshitomi H, Oishi S, et al. Stromal cell-derived
21. Cruess RL, Dumont J. Fracture healing. Can J Surg. factor 1/CXCR4 signaling is critical for the recruit-
1975;18(5):403–13. ment of mesenchymal stem cells to the fracture site
22. Kolar P, Schmidt-Bleed K, Schell H, Gaber T, during skeletal repair in a mouse model. Arthritis
Toben D, Schmidmaier G, et al. The early fracture Rheum. 2009;60(3):813–23.
hematoma and its potential role in fracture healing. 36. Schindeler A, McDonald MM, Bokko P, Lit-
Tissue Eng Part B Rev. 2010;16(4):427–34. tle DG. Bone remodeling during fracture repair:
23. Oe K, Miwa M, Sakai Y, Lee SY, Kuroda R, the cellular picture. Semin Cell Dev Biology.
Kurosaka M. An in vitro study demonstrating that 2008;19(5):459–66.
37. Lu C, Saless N, Wang X, Sinha A, Decker S, formation in a rat osteotomy model. Clin Orthop
Kazakia G, et al. The role of oxygen during fracture Relat Res. 2011;469(11):3094–101.
healing. Bone. 2013;52(1):220–9. 53. Miclau T, Lu C, Thompson Z, Choi P, Puttlitz C,
38. Keramaris NC, Calori GM, Nikolaou VS, Marcucio R, Helms JA. Effects of delayed stabi-
Schemitsch EH, Giannoudis PV. Fracture vascular- lization on fracture healing. J Orthop Res. 2007;25
ity and bone healing: a systematic review of the role (12):1552–8.
of VEGF. Injury. 2008;39(Suppl 2):S45–57. 54. Weaver AS, Su YP, Begun DL, Miller JD,
39. Gerber HP, Vu TH, Ryan AM, Kowalski J, Werb Z, Alford AI, Goldstein SA. The effects of axial
Ferrara N. VEGF couples hypertrophic cartilage displacement on fracture callus morphology and
remodeling, ossification and angiogenesis during MSC homing depend in the timing of application.
endochondral bone formation. Nat Med. 1999;5 Bone. 2010;47(1):41–8.
(6):623–8. 55. Brinker MR, Hanus BD, Sen M, O’Connor DP. The
40. Behonick DJ, Xing Z, Lieu S, Buckley JM, Lotz JC, devastating effects of tibial nonunion on
Marcucio RS, et al. Role of matrix metallopro- health-related quality of life. J Bone Joint Surg
teinase 13 in both endochondral and intramembra- Am. 2013;95(24):2170–6.
nous ossification during skeletal regeneration. 56. Bhandari M, Fong K, Sprague S, Williams D,
PLoS ONE. 2007;2(11):e1150. Petrisor B. Variability in the definition and per-
41. Shapiro F. Bone development and its relation to ceived causes of delayed unions and nonunions: a
fracture repair. The role of mesenchymal osteoblasts cross-sectional, multinational survey of orthopaedic
and surface osteoblasts. Eur Cell Mater. surgeons. J Bone Joint Surg Am. 2012;94(15):
2008;15:53–76. e1091–6.
42. Flick LM, Weaver JM, Ulrich-Vinther M, Abuzza- 57. Corrales LA, Morshed S, Bhandari M, Miclau T
hab F, Zhang X, Dougall WC, et al. Effects of 3rd. Variability in the assessment of fracture-healing
receptor activator of NFkappaB (RANK) signaling in orthopaedic trauma studies. J Bone Joint Surg
blockade on fracture healing. J Orthop Res. 2003;21 Am. 2008;90(9):1862–8.
(4):676–84. 58. Joslin CC, Eastaugh-Waring SJ, Hardy JR, Cun-
43. Raisz LG. Physiology and pathophysiology of bone ningham JL. Weight bearing after tibial fracture as a
remodeling. Clin Chem. 1999;45(8 Pt 2):1353–8. guide to healing. Clin Biomech (Briston, Avon).
44. Sims NA, Gooi JH. Bone remodeling: multiple 2008;23(3):329–33.
cellular interactions required for coupling of bone 59. Webb J, Herling G, Gardner T, Kenwright J,
formation and resorption. Semin Cell Dev Biol. Simpson AH. Manual assessment of fracture stiff-
2008;19(5):444–51. ness. Injury. 1996;27(5):319–20.
45. Teitelbaum SL. Bone resorption by osteoclasts. 60. Whelan DB, Bhandari M, Stephen D, Kreder H,
Science. 2000;289(5484):1504–8. (Review). McKee MD, Zdero R, Schemitsch EH. Develop-
46. Fan W, Crawford R, Xiao Y. Structural and cellular ment of the radiographic union score for tibial
differences between metaphyseal and diaphyseal fractures for the assessment of tibial fracture healing
periosteum in different aged rats. Bone. 2008;42 after intramedullary fixation. J Trauma. 2010;68
(1):81–9. (3):629–32.
47. Uhthoff HK, Rahn BA. Healing patterns of meta- 61. Kooistra BW, Dijkman BG, Busse JW, Sprague S,
physeal fractures. Clin Orthop Relat Res. Schemitsch EH, Bhandari M. The radiographic
1981;160:295–303. union scale in tibial fractures: reliability and
48. Claes L, Reusch M, Göckelmann M, Ohnmacht M, validity. J Orthop Trauma. 2010;24(Suppl 1):S81–
Wehner T, Amling M, et al. Metaphyseal fracture 6.
healing follows similar biomechanical rules as dia- 62. Bhandari M, Chiavaras M, Ayeni O, Chakraver-
physeal healing. J Orthop Res. 2011;29(3):425–32. rty R, Parasu N, Choudur H, et al. Assessment
49. Augat P, Margevicius K, Simon J, Wolf S, Suger G, Group for Radiographic Evaluation and Evidence
Claes L. Local tissue properties in bone healing: (AGREE) Study Group (AGREE Investigators
influence of size and stability of the osteotomy Writing Committee). Assessment of radiographic
gap. J Orthop Res. 1998;16(4):475–81. fracture healing in patients with operatively treated
50. Claes LE, Heigele CA, Neidlinger-Wilke C, Kas- femoral neck fractures. J Orthop Trauma. 2013;27
par D, Seidl W, Margevicius KJ, Augat P. Effects of (9):e213–9.
mechanical factors on the fracture healing process. 63. Bhandari M, Chiavaras MM, Parasu N, Choudur H,
Clin Orthop Relat Res. 1998;(355 Suppl):S132–47. Ayeni O, Chakravertty R, et al. Radiographic union
51. Augat P, Merk J, Ignatius A, Margevicius K, score for hip substantially improves agreement
Bauer G, Rosenbaum D, Claes L. Early full between surgeons and radiologists. BMC Muscu-
weightbearing with flexible fixation delays fracture loskelet Disord. 2013;14:70.
healing. Clin Orthop Relat Res. 1996;328:194–202. 64. McClelland D, Thomas PB, Bancroft G, Moor-
52. Willie BM, Blakytny R, Glockelmann M, croft CI. Fracture healing assessment comparing
Ignatius A, Claes L. Temporal variation in fixation stiffness measurements using radiographs. Clin
stiffness affects healing by differential cartilage Orthop Relat Res. 2006;457:214–9.
65. Davis BJ, Roberts PJ, Moorcroft CI, Brown MF, 80. Cozen L. Does diabetes delay fracture healing? Clin
Thomas PB, Wade RH. Reliability of radiographs in Orthop Relat Res. 1972;82:134–40.
defining union of internally fixed fractures. Injury. 81. Loder R. The influence of diabetes mellitus on the
2004;35(6):557–61. healing of closed fractures. Clin
66. Cox G, Einhorn TA, Tzioupis C, Giannoudis PV. Orthop. 1988;232:210–6.
Bone-turnover markers in fracture healing. J Bone 82. Albowi J, Tian C, Siqueira MF, Kayal RA, McKen-
Joint Surg Br. 2010;92(3):329–34. zie E, Behl Y, et al. Chemokine expression is
67. Lu C, Miclau T, Hu D, Hansen E, Tsui K, Puttlitz C, upregulated in chondrocytes in diabetic fracture
Marcucio RS. Cellular basis for age-related changes healing. Bone. 2013;53(1):294–300.
in fracture repair. J Orthop Res. 2005;23(6):1300–7. 83. Kayal RA, Siqueira M, Alblowi J, McLean J,
68. Desai BJ, Meyer MH, Porter S, Kellam JF, Krothapalli N, Faibish D, et al. TNF-alpha mediates
Meyer RA Jr. The effect of age on gene expression diabetes-enhanced chondrocyte apoptosis during
in adult and juvenile rats following femoral fracture. fracture healing and stimulates chondrocyte apop-
J Orthop Trauma. 2003;17(10):689–98. tosis through FOXO1. J Bone Miner Res. 2010;25
69. Lu C, Sapozhnikova A, Hu D, Miclau T, Marcu- (7):1604–15.
cio RS. Effect of age on vascularization during 84. Kayak RA, Tsatsas D, Bauer MA, Allen B,
fracture repair. J Orthop Res. 2008;26(10):1384–9. Al-Sebaei MO, Kakar S, et al. Diminished bone
70. Egermann M, Heil P, Tami A, Ito K, Janicki P, Von formation during diabetic fracture healing is related
Rechenberg B, et al. Influence of defective bone to the premature resorption of cartilage associated
marrow osteogenesis on fracture repair in an with increased osteoclast activity. J Bone Miner
experimental model of senile osteoporosis. Res. 22(4);560–8.
J Orthop Res. 2009;28(6):798–804. 85. Kagel EM, Einhorn TA. Alterations of fracture
71. D’Ippolito G, Schiller PC, Ricordi C, Roos BA, healing in the diabetic condition. Iowa
Howard GA. Age-related osteogenic potential of Orthop J. 16:147–52.
mesenchymal stromal stem cells from human ver- 86. Follak N, Klöting I, Merk H. Influence of diabetic
tebral bone marrow. J Bone Miner Res. 1999;14 metabolic state on fracture healing in spontaneously
(7):1115–22. diabetic rats. Diabetes Metab Res Rev. 2005;21
72. Stenderup K, Justesen J, Eriksen EF, Rattan S, (3):288–96.
Kassem M. Number and proliferative capacity of 87. Gandhi A, Beam HA, O’Connor JP, Parsons JR,
osteogenic stem cells are maintained during aging Lin SS. The effects of local insulin delivery on
and in patients with osteoporosis. J Bone Miner diabetic fracture healing. Bone. 2005;37(4):482–90.
Res. 2001;16(6):1120–9. 88. Chaudhary SB, Liporace FA, Gandhi A, Don-
73. International Osteoporosis Foundation. Epidemiol- ley BG, Pinzur MS, Lin SS. Complications of ankle
ogy. http://www.iofbonehealth.org/epidemiology. fracture in patients with diabetes. J Am Acad
Accessed 27 Aug 2015. Orthop Surg. 2008;16(3):159–70.
74. Nikolaou VS, Efstathopoulos N, Kontakis G, 89. Castillo RC, Bosse MJ, MacKenzie EJ. Patterson
Kanakaris NK, Giannoudis PV. The influence of BM; LEAP Study Group. Impact of smoking on
osteoporosis in femoral fracture healing time. fracture healing and risk of complications in
Injury. 2009;40(6):663–8. limb-threatening open tibia fractures. J Orthop
75. Namkung-Matthai H, Appleyard R, Jansen J, Hao Trauma. 2005;19(3):151–7.
Lin J, Maastricht S, Swain M, et al. Osteoporosis 90. Adams CI, Keating JF, Court-Brown CM. Cigarette
influences the early period of fracture healing in a smoking and open tibial fractures. Injury. 2001;32
rat osteoporotic model. Bone. 2001;28(1):80–6. (1):61–5.
76. Walsh WR, Sherman P, Howlett CR, 91. Murray IR, Foster CJ, Robinson CM. Risk factors
Sonnabend DH, Ehrlich MG. Fracture healing in a for nonunion after nonoperative treatment of dis-
rat osteopenia model. Clin Orthop Relat Res. placed midshaft fractures of the clavicle. J Bone
1997;342:218–27. Joint Surg Am. 2013;95(13):1153–8.
77. Lill CA, Hesseln J, Schlegel U, Eckhardt C, 92. Ristiniemi J, Flinkkila T, Hyvonen P, Lakovaara M,
Goldhahn J, Schneider E. Biomechanical evaluation Pakarinen H, Biancari F, Jalovaara P. Two-ring
of healing in a non-critical defect in a large animal hybrid external fixation of distal tibial fractures: a
model of osteoporosis. J Orthop Res. 2003;21 review of 47 cases. J Trauma. 2007;62(1):174–83.
(5):836–42. 93. Govender S, Csimma C, Genant HK,
78. Brinker MR, O’Connor DP, Monla YT, Earth- Valentin-Opran A, Amit Y, Arbel R, et al. BMP-2
man TP. Metabolic and endocrine abnormalities in Evaluation in Surgery for Tibial gtrauma (BESTT)
patients with nonunions. J Orthop Trauma. 2007;21 Study Group. Recombinant human bone morpho-
(8):557–70. genetic protein-2 for treatment of open tibial
79. Lancourt JE, Hochberg F. Delayed fracture healing fractures: a prospective, controlled, randomized
in primary hyperparathyroidism. Clin Orthop Rel study of four hundred and fifty patients. J Bone
Res. 1977;124:214–8. Joint Surg Am. 2002;84-A(12):2123–34.
94. Shen Y, Liu HX, Ying XZ, Yang SZ, Nie PF, 110. Li M, Healy DR, Li Y, Simmons HA, Crawford DT,
Cheng SW, et al. Dose-dependent effects of nicotine Ke HZ, et al. Osteopenia and impaired fracture
on proliferation and differentiation of human bone healing in aged EP4 receptor knockout mice. Bone.
marrow stromal cells and the antagonistic action of 2005;37(1):46–54.
vitamin C. J Cell Biochem. 2013;114(8):1720–8. 111. Simon AM, Manigrasso MB, O’Connor
95. Lee JJ, Patel R, Biermann S, Dougherty PJ. The JP. Cyclo-oxygenase 2 function is essential for
musculoskeletal effects of cigarette smoking. J Bone bone fracture healing. J Bone Miner Res. 2002;17
Joint Surg Am. 2013;95(9):850–9. (6):963–76.
96. Skott M, Andreassen TT, Ulrich-Vinther M, 112. Giannoudis PV, MacDonald DA, Matthews SJ,
Chen X, Keyler DE, LeSage MG, et al. Tobacco Smith RM, Furlong AJ, De Boer P. Nonunion of the
extract but not nicotine impairs the mechanical femoral diaphysis: the influence of reaming and
strength of fracture healing in rats. J Orthop Res. non-steroidal anti-inflammatory drugs. J Bone Joint
2006;24(7):1472–9. Surg Br. 2000;82(5):655–8.
97. Kristensson H, Lunden A, Nilsson BE. Fracture 113. Dahners LE, Mullis BH. Effects of nonsteroidal
incidence and diagnostic roentgen in alcoholics. anti-inflammatory drugs on bone formation and
Acta Orthop Scan. 1980;51(2):205–7. soft-tissue healing. J Am Acad Orthop Surg.
98. Nyquist F, Berglund M, Nilsson BE, 2004;12(3):139–43.
Obrant KJ. Nature and healing of tibial shaft 114. Ro J, Sudmann E, Marton PF. Effect of indometha-
fractures in alcohol abusers. Alcohol. 1997;32 cin on fracture healing in rats. Acta Orthop Scand.
(1):91–5. 1976;47(6):588–99.
99. Laitinen K, Lamberg-Allardt C, Tunninen R, 115. Gerstenfeld LC, Thiede M, Seibert K, Mielke C,
Harkonen M, Valimski M. Bone mineral density Phippard D, Svagr B, et al. Differential inhibition of
and abstention-induced changes in bone and min- fracture healing by non-selective cyclooxygenase-2
eral metabolism in noncirrhotic male alcoholics. selective non-steroidal anti-inflammatory drugs.
Am J Med. 1992;93(6):642–50. J Orthop Res. 2003;21(4):670–5.
100. Diamond T, Stiel D, Lunzer M, Wilkinson M, 116. Lack WD, Fredericks D, Petersen E, Donovan M,
Posen S. Ethanol reduces bone formation and may George M, Nepola J, et al. Effect of aspirin on bone
cause osteoporosis. Am J Med. 1989;86(3):282–8. healing in a rabbit ulnar osteotomy model. J Bone
101. Hillers VN, Massey LK. Interrelationships of mod- Joint Surg Am. 2013;95(6):488–96.
erate and high alcohol consumption with diet and 117. Solomon HM, Mogun H, Schneeweiss S. The
health status. Am J Clin Nutr. 1985;41(2):356–62. relation between bisphosphonate use and
102. Chakkalakal DA. Alcohol-induced bone loss and non-union of fractures of the humerus in older
deficient bone repair. Alc Clin Exper Res. 2005;29 adults. Osteoporos Int. 2009;20(6):895–901.
(12):2077–90. 118. Rozental TD, Vazquez MA, Chacko AT, Ayogu N,
103. Chakkalakal DA, Novak JR, Fritz ED, Mollner TJ, Bouxsein ML. Comparison of radiographic fracture
McVicker DL, Garvin KL, et al. Inhibition of bone healing in the distal radius for patients on and off
repair in a rat model for chronic and excessive bisphosphonate therapy. J Hand Surg Am. 2009;34
alcohol consumption. Alcohol. 2005;36(3):1–14. (4):595–602.
104. Jung MK, Callaci JJ, Lauing KL, Otis JS, 119. Gong HS, Song CH, Lee YH, Thee SH, Lee HJ,
Radek KA, Jones MK, Kovacs EJ. Alcohol expo- Baek GH. Early initiation of bisphosphonate does
sure and mechanisms of tissue injury and repair. not affect healing and outcomes of volar plate
Alcohol Clin Exp Res. 2011;35(3):392–9. fixation of osteoporotic distal radial fractures.
105. Janicke-Lorenz J, Lorenz R. Alcoholism and frac- J Bone Joint Surg Am. 2012;94(19):1729–36.
ture healing. a radiological study in the rat. Arch 120. Lyles KW, Colón-Emeric CS, Magaziner JS,
Orthop Trauma Surg. 1984;103(4):286–9. Adachi JD, Pieper CF, Mautalen C, et al. HOR-
106. Volkmer DL, Sears B, Lauing KL, Nauer RK, IZON recurrent fracture trial. Zoledronic acid and
Roper PM, Yong S, et al. Antioxidant therapy clinical fractures and mortality after hip fracture.
attenuates deficient bone fracture repair associated N Engl J Med. 2007;357(18):1799–809.
with binge alcohol exposure. J Orthop Trauma. 121. Li J, Mori S, Kaji Y, Mashiba T, Kawanishi J,
2011;25(8):516–21. Norimatsu H. Effect of bisphosphonate (in-
107. Kurmis AP, Kurmis TP, O’Brien JX, Dalén T. The cadronate) on fracture healing of long bones in
effect of nonsteroidal anti-inflammatory drug rats. J Bone Miner Res. 1999;14:969–79.
administration on acute phase fracture-healing: a 122. McDonald MM, Dulai S, Godfrey C, Amanat N,
review. J Bone Joint Surg Am. 2012;94(9):815–23. Sztynda T, Little DG. Bolus or weekly zoledronic
108. Abdul-Hadi O, Parvizi J, Austin MA, Viscusi E, acid administration does not delay endochondral
Einhorn T. Nonsteroidal anti-inflammatory drugs in fracture repair but weekly dosing enhances delays in
orthopaedics. J Bone Joint Surg Am. 2009;91 hard callus remodeling. Bone. 2008;43(4):653–62.
(8):2019–27. 123. Saito M, Shiraishi A, Ito M, Sakai S, Kayakawa N,
109. Jee WS, Ma YF. The in vivo anabolic actions of Mihara M, Marumo K. Comparison of effects of
prostaglandins in bone. Bone. 1997;21(4):297–304. alfacalcidol and alendronate on mechanical
properties and bone collagen cross-links of callus in Wnt signaling in PTH-treated fractures. J Bone
the fracture repair rat model. Bone. 2010;46 Miner Res. 2007;22(12):1903–12.
(4):1170–9. 132. Holzer G, Majeska RJ, Lundy MW, Hartke JR,
124. Jørgensen NR, Schwarz P. Effects of Einhorn TA. Parathyroid hormone enhances frac-
anti-osteoporosis medications on fracture healing. ture healing. A preliminary report. Clin Orthop
Curr Osteoporos Rep. 2011;9(3):149–55. Relat Res. 1999;366:258–63.
125. Odvina CV, Zerwekh JE, Rao DS, Maalouf N, 133. Andreassen TT, Ejersted C, Oxlund H. Intermittent
Gottschalk FA, Pak CY. Severely suppressed bone parathyroid hormone (1–34) treatment increases
turnover: a potential complication of alendronate callus formation and mechanical strength of healing
therapy. J Clini Endocrinol Metab. 2005;90 rat fractures. J Bone Miner Res. 1999;14(6):960–8.
(3):1294–301. 134. Andreassen TT, Willick GE, Morley P, Whitfield JF.
126. Egol KA, Park JH, Rosenberg ZS, Peck V, Treatment with parathyroid hormone hpth(1–34),
Tejwani NC. Healing delayed but generally reliable hpth(1–31), and monocyclic hpth(1–31) enhances
after bisphosphonate-associated complete femur fracture strength and callus amount after withdrawal
fractures treated with IM nails. Clin Orthop Relat fracture strength and callus mechanical quality
Res. 2014;472(9):2728–34. continue to increase. Calcif Tissue Int. 2004;74
127. Weil YA, Rivkin G, Safran O, Liebergall M, (4):351–6.
Foldes AJ. The outcome of surgically treated femur 135. Manabe T, Mori S, Mashiba T, Kaji Y, Iwata K,
fractures associated with long-term bisphosphonate Komatsubara S. Human parathyroid hormone (1–
use. J Trauma. 2011;71(1):186–90. 34) accelerates natural fracture healing process in
128. Diab DL, Watts NB. Bisphosphonate drug holiday: the femoral osteotomy model of cynomolgus mon-
who, when and how long. Ther Adv Musculoskelet keys. Bone. 2007;40(6):1475–82.
Dis. 2013;5(3):107–11. 136. Aspenberg P, Genant HK, Johansson T, Nino AJ,
129. Ellegaard M, Jorgensen NR, Schwarz P. Parathyroid See K, Krohn K, et al. Teriparatide for acceleration
hormone and bone healing. Calcif Tissue Int. of fracture repair in humans: a prospective, ran-
2010;87(1):1–13. domized double-blind study of 102 postmenopausal
130. Goldhahn J, Feron JM, Kanis J, Papapoulos A, women with distal radial fractures. J Bone Miner
Reginster JY, Rizzoli R, et al. Implications for Res. 2010;25(2):404–14.
fracture healing of current and new osteoporosis 137. Peichl P, Holzer LA, Maier R, Holzer G. Parathy-
treatments: an ESCEO consensus paper. Calcif roid hormone 1-84 accelerates fracture-healing in
Tissue Int. 2012;90(5):343–53. pubic bones of elderly osteoporotic women. J Bone
131. Kakar S, Einhorn TA, Vora S, Miara LJ, Hon G, Joint Surg Am. 2011;93(17):1583–7.
Wigner NA, et al. Enhanced chondrogenesis and
Clavicle Nonunions
3
Laura A. Schemitsch, MA, Emil H. Schemitsch, MD, FRCS(C)
and Michael D. McKee, MD, FRCS(C)
callus has formed. In an article published in the

3.1 Introduction
American Journal of Surgery in 1941, Ghormley
et al. stated that clavicular fractures resulting in
The clavicle is a flat S-shaped bone that connects
nonunion were “unusual” [20]. Neer’s evaluation
the arm to the body and is located directly above
of 2235 patients in 1960 reported only three
the first rib. It is the first bone to ossify in utero
patients as having un-united fractures [2]. In
[1]. Fractures of the clavicle are particularly
1968, Rowe reported four nonunions within a
common, as they account for approximately 2–
series of 566 patients, thus resulting in an
5% of all fractures [2, 3] and 44% of fractures to
extremely low rate of nonunion (0.8%) [4].
the shoulder girdle [4]. These injuries tend to
Although nonunion was considered to be a seri-
occur most frequently in young and active indi-
ous complication in relation to clavicle fractures,
viduals, particularly men [5].
the understanding that it was rare had significant
Fractures of the distal third appear in roughly
implications discouraging operative intervention.
20% of cases (Fig. 3.1), mainly occurring in
Recent studies of displaced midshaft clavicle
elderly patients as a result of a low impact fall
fractures have shed new light on the reported inci-
[6–11]. In rare occurrences (5%), the medial end
dence of nonunion. In 1997, Hill et al. reviewed 52
of the clavicle bone is fractured [7, 12, 13]. In the
patients with displaced midshaft clavicle fractures
majority of cases, the midshaft of the clavicle
treated nonoperatively [21]. They reported a non-
fractures, accounting for approximately 80–85%
union rate of 15% (8 out of 52) and an unsatisfac-
of all cases [7, 14–17]. Fractures of the midshaft
tory clinical outcome in 31% of patients. In a
were traditionally treated nonoperatively due to
review of 2144 displaced midshaft fractures col-
the reported low incidence of nonunion in past
lected from the literature between 1975 and 2005,
studies in which clavicular fractures had been
Zlowodzki et al. reported a nonunion rate of 15.1%
treated conservatively [16, 18, 19]. Nonunion is
(Table 3.1) [22]. This is markedly higher than the
defined as the lack of radiographic healing at six
rate described by Neer and Rowe. Several other
months post-injury [18]. Callus formation indi-
recent studies have reinforced these findings [8, 17,
cates that a hypertrophic nonunion has occurred,
23–27]. In 2007, a study conducted by the Cana-
whereas an atrophic nonunion results when no
dian Orthopedic Trauma Society compared non-
operative treatment versus plate fixation for
displaced midshaft clavicle fractures [18]. This
L.A. Schemitsch E.H. Schemitsch (&) series found that there was a significantly lower rate
M.D. McKee of nonunion in the operative group (62 patients, 2
Department of Surgery, Division of Orthopaedic
nonunions, rate of nonunion 3%) compared with
Surgery, University of Toronto, St. Michael’s
Hospital, Toronto, Canada the patients treated conservatively (49 patients, 7
e-mail: schemitsche@smh.ca nonunions, rate of nonunion 14%). These findings

76 LauraA. Schemitsch, MA et al.
Fig. 3.1 a Preoperative radiograph of distal clavicle nonunion. b Post-operative radiograph of distal clavicle nonunion
following open reduction and internal fixation with a precontoured distal clavicle plate and a lag screw
Table 3.1 Nonunion of the clavicle following various treatments. Meta-analysis of nonoperative treatment,
intramedullary pinning, and plate fixation of displaced midshaft fractures of the clavicle from series published in
1975 through 2005 (From Zlowodzki et al. [22], with permission)
Nonunions Infections Infection Fixation
(Total) (Deep)a Failuresb
Nonoperative 15.1 N/A N/A 0
(n = 159)
Plating (n = 460) 2.2 4.6 2.4 2.2
Intramedullary pinning (n = 152) 2 6.6 0 3.9
Total (N = 771) 4.8 5.1 1.8 2.1
c c
(3.5–6.5) (3.6–7.1) (1–3.2) (1.3–3.3)
N/A = not applicable
Data are percentages with 95% confidence intervals in parentheses
a
Any infection described as deep or superficial requiring irrigation and debridement; infections of unknown significance
were not included
b
One includes refractures
c
Infection rates only include operatively treated fractures
are helpful in establishing that primary fixation may functional outcomes [8, 23, 25, 28]. Reported
benefit young and active individuals with fully risk factors for nonunion include: a “refrac-
displaced midshaft clavicle fractures. An assess- ture,” clavicle shortening (>15–20 mm),
ment of a patient’s injuries as well as their func- female gender, fracture comminution,
tional expectations remains one of the most increasing fracture displacement, older age,
important considerations regarding decisions for severe initial trauma, and unstable lateral
treatment [7]. While operative treatment is typically fracture (Neer type II) [2, 21, 28–32]. It is
effective for established nonunions, primary pre- highly likely that there are other factors that
vention of nonunion would be preferential increase risk of nonunion that requires further
(Figs. 3.2 and 3.3). clarification, such as multiple ipsilateral rib
Several recent studies have indicated that fractures or associated scapular/glenoid frac-
certain patient populations pose a greater risk tures [14, 32–34].
of nonunion, malunion, and/or poor
3 Clavicle Nonunions 77
Fig. 3.2 a Preoperative radiograph of midshaft clavicle reduction and internal fixation with a precontoured
nonunion in an active 28-year-old man fourteen months clavicle plate. Note the solid “bridge” portion of the
after fracture. b Post-operative radiograph demonstrating plate, providing extra strength at the nonunion site
union of midshaft clavicle nonunion following open
Fig. 3.3 a Preoperative radiograph of midshaft clavicle nonunion following open reduction and internal fixation
nonunion in a 39-year-old man who had failed two prior with a precontoured plate and an intercalary tricortical
attempts at operative fixation. There is significant bony iliac crest bone graft
defect. b Post-operative radiograph of midshaft clavicle
Orthopedic Trauma Unit at the Royal Infirmary of

3.2 Classification Edinburgh. The age, gender, and mechanism of
injury were recorded for each patient within 72 h
With the Allman classification from 1967 [35], of the injury. The anatomical site, configuration,
clavicle fractures can be divided into three groups; type, and extent of comminution were also
Group 1, fractures of the middle third; Group 2, recorded for each fracture. Radiographs from
fractures of the lateral third; and Group III, frac- twenty randomly selected fractures aided in the
tures of the medial third. These three groups can be development of this new classification system.
further subdivided into nondisplaced and dis-
placed fractures. Within Group 1, there is a further
subgroup that consists of fracture comminution 3.3 Epidemiology
and one or more displaced intermediary frag-
ments. In 1998, Robinson presented a new and The findings of Robinson’s epidemiological study
more comprehensive classification for clavicle estimated that in patients over 13 years of age,
fractures [17]. This was performed following a clavicle fractures occurred at a rate of 29.14 per
study of 1000 consecutive patients who were 100,000 annually [9]. Prior to this, Nordqvist and
treated for isolated clavicle fractures in the Peterson had found the incidence of clavicle
fractures to be 64 per 100,000 annually [26]. The significantly higher rates of nonunion. In 1986,
mean age for clavicular fractures was 29 years in Eskola et al. reported a nonunion rate of 3% [37].
men and 45 years in women. In that series, clav- In 1997, Hill et al. reported an exponentially
icle fractures occurred most frequently in males higher nonunion rate of 15% [21]. In 2004,
under the age of 30 years as a result of a sports Robinson et al. found an overall nonunion rate of
injury or a road traffic accident (RTA). The 6.2: 8.3% of medial fractures, 4.5% of diaphyseal
majority of fractures (89%) healed without com- fractures, and 11.5% of fractures to the lateral
plication when treated conservatively. Nonunion end [8]. Similar to Hill, Zlowodzki et al. reported
was only seen in one nondisplaced Type A frac- a nonunion rate of 15.1% in 2005 [22]. The
ture (Type 3A-1). The rate of delayed union and Canadian Orthopedic Trauma Society’s 2007
nonunion was far higher in the displaced Type B trial reported a rate of nonunion in the operative
fracture cohort, with a rate of 2.7 and 4.8%, group of 3% compared with 14% in the nonop-
respectively. They occurred almost entirely fol- erative group [18]. Most recently in 2011, Kul-
lowing a displaced diaphyseal (type 2B) or dis- shrestha et al. [38] conducted a study of 73
placed lateral end (type 3B) clavicular fracture. patients, aged 20–50 years, who were allocated
The odds ratio (OR) for delayed union or non- to operative or nonoperative treatment for their
union following a type 2B fracture compared with displaced midshaft clavicle fractures. In the
a type 2A fracture was 18 (meaning that a patient operative group, none of the 45 patients went on
with a 2B fractures is eighteen times more likely to to nonunion. In the nonoperative group, eight
develop a delayed or nonunion than a patient with patients out of twenty-eight developed nonunion
a 2A fracture pattern). The OR for delayed union which resulted in an extremely high rate of
or nonunion following a type 3B fracture com- nonunion (29%) [20]. New knowledge regarding
pared with a type 3A fracture was seventy-five. the potential rate of nonunion for this injury has
However, due to the increased prevalence of type come to influence management options for
2B fractures, more patients from this cohort were orthopedic surgeons who treat these fractures.
seen with a delayed or nonunion. High-energy
injuries such as falls from a height, RTA, and
direct violence indicated a greater incidence of 3.5 Possible Risk Factors
delayed and nonunion compared with low-energy for Nonunion
mechanisms of injury. Differences in the rate of
union were not significant amongst the age or Murray et al. [39] sought to determine risk fac-
gender cohorts. tors for nonunion in an adult population of
patients with displaced midshaft clavicle frac-
tures. They retrospectively reviewed 941 patients
3.4 Rates of Nonunion in Displaced who were at least 18 years of age, who had ini-
Clavicular Fractures tially received nonoperative management for
their injuries from January 1994 to December
Nonunion is significantly more common in dis- 2007. No nonunions were reported in the 184
placed clavicular fractures, particularly of the patients who were under 18 years of age, and
midshaft [36]. A direct correlation has been thus, they were excluded from the study. The risk
demonstrated between increased degrees of dis- of nonunion in children was zero, and age was
placement and poor functional outcomes [18]. not a significant risk factor for nonunion in the
The study conducted by Robinson et al. [8] adult population. The study population was typ-
demonstrated that patients with a displaced, ical of most midshaft clavicular fracture popula-
comminuted midshaft clavicular fracture had a tions in that it consisted mainly of young and
nonunion rate of 21%. active men. One hundred and twenty-five of
Neer and Rowe initially reported low rates of these patients went on to develop nonunion
nonunion [2, 4]. Subsequent studies have shown which constitutes a risk of nonunion of 13.3%
(95% confidence interval [CI], 11.3–15.6%). outcomes in patients with comminution had been
Using bivariate analysis, the significant factors shown in prior studies [8, 22, 27, 40]. Murray
for a higher nonunion rate were female gender, et al. also identified comorbidities that could
smoking, increased fracture displacement, over- potentially increase the risk of nonunion which
lap, translation, and comminution. In the multi- included rheumatoid disease, immunocompro-
variate analysis, the significant factors were mise, renal failure, epilepsy, and use of drugs
smoking (OR, 3.76), comminution (OR, 1.75), such as corticosteroids and those interfering with
and increased fracture displacement (OR, 1.17) vitamin D metabolism [39]. Since many patients
(Fig. 3.4). The negative effect of smoking on with a high risk of nonunion will not develop this
fracture healing has been shown in previous complication, and many patients with fewer risk
studies [40–42]. This is the first time that factors do go on to nonunion, these risk factors
smoking was identified as a risk factor for clav- are best used as a guide for surgeons when
icular nonunion since earlier studies that inves- making a plan for management of this injury.
tigated whether smoking was a risk factor This knowledge would ideally improve the
showed no correlation [21, 28]. However, poorer functional outcomes for certain patients who
Fig. 3.4 a Preoperative radiograph of a midshaft clavicle following open reduction and internal fixation with a
fracture in a 25-year-old female. b The patient was treated precontoured plate and an iliac crest bone graft. d Radio-
nonoperatively and went on to develop a symptomatic graph of healed midshaft clavicle fracture at final
atrophic nonunion with severe displacement. follow-up
c Post-operative radiograph of midshaft clavicle nonunion
would benefit from surgical intervention, while have better functional outcomes than those trea-
simultaneously avoiding unwarranted surgery for ted nonoperatively [18]. Additionally, patients
others [39]. who undergo secondary fixation due to the
development of nonunion following nonopera-
tive treatment have results that are somewhat
3.6 Treatment Options inferior to outcome following primary fixation.
[18, 45]. This would suggest that primary fixa-
Nonunion is generally defined as the lack of tion for displaced midshaft clavicle fractures may
radiographic healing at six months post-initial be the preferred course of management for cer-
injury and delayed unions are typically defined tain patients [18]. However, as Murray et al. [39]
when there is progression of healing, but the and McKee [7] suggest, it is important to identify
fracture has not achieved radiographic union at a patient’s risk factors for nonunion in order to
three months [7]. Symptomatic clavicular non- assess their suitability for surgical treatment to
union is typically painful and debilitating for optimize management.
active patients, causing local pain, persistent
deformity, and shoulder weakness resulting in
neurologic symptoms consistent with brachial 3.8 Operative
plexus impingement. Treatment options vary
from nonoperative techniques that include Methods of intervention include excision [37, 46–
symptomatic treatment with a sling or 49], intramedullary fixation [14, 50–52], and
figure-of-eight bandage, to noninvasive tech- open reduction and internal fixation (ORIF) with
niques involving electrical stimulation or cortical [53] or autogenous bone graft [32, 37, 40,
low-intensity ultrasound [5, 7, 15, 43], and sur- 54–64]. Dual plating with 100% union rates has
gical intervention. Asymptomatic patients with a also been reported [65]. Recent studies have
radiographic nonunion who have full clinical shown that young and active patients with
function do not require surgical intervention and symptomatic nonunion benefit from surgical
respond well to conservative treatment. treatment. However, residual functional impair-
ment can still occur [66]. The series presented by
Robinson et al. [8] found that the highest rate of
3.7 Nonoperative nonunion occurred in elderly female patients with
diaphyseal clavicle fractures. However, due to the
Clavicle fractures have traditionally been man- reduced functional demands of elderly patients, it
aged nonoperatively due to studies published in is possible that surgery may not be required in
the 1960s that indicated low nonunion rates [2, these individuals. This is likely not the case for
4]. This treatment involves mobilizing the active and young patients who still account for
shoulder in a sling or figure-of-eight bandage. In many of the nonunions that occur in clavicle
a randomized, controlled trial that compared the fracture cases. Poor function can hinder lifestyle,
treatment of clavicle fractures with a sling versus and for this reason, surgical intervention is usu-
a figure-of-eight bandage, the functional and ally desirable since it has been shown to signifi-
cosmetic outcomes were found to be identical cantly improve long-term shoulder function [18].
between the two groups [44]. Recent studies Clavicle nonunions are primarily treated with
have shown that the risk of nonunion following reconstructive procedures, and in rare cases, a
nonoperative treatment of midshaft clavicle salvage procedure may occur when there are few
fractures ranges from five to twenty per cent [8, or no options left for a patient [67]. Salvage
18, 21, 22]. Patients with displaced fractures are procedures have involved partial or total resection
at an even greater risk of developing nonunion of the clavicle bone (claviculectomy) as well as
[8, 21, 22]. Patients who undergo primary fixa- excision of a bony prominence in order to relieve
tion of their displaced midshaft clavicle fractures local pain caused by symptomatic nonunion.
3.8.1 Reconstructive Procedures hold the reduction while a plate is applied to the
surface can be very beneficial [7]. Due to the
The primary surgical treatment of clavicle non- complexity of the bone’s structure and the
unions involves reconstructive procedures. There multidirectional biomechanical forces that act on
are a number of methods used in the treatment of the nonunion, at least three screws should be
clavicle nonunions, with an increasing body of placed on both sides of the fracture site to sta-
literature to describe such methods and cases bilize the bone [57, 72]. In a series in which
(Table 3.2) [14, 37, 40, 50, 54–58, 60, 62, 68, clavicular nonunion were treated with short
69]. There are two main operative treatment plates (4 hole semi-tubular), there was a high
options used to achieve clavicular union: plate risk of failure reported [73]. In hypertrophic
fixation and intrameduallary screw/pin fixation. nonunions, the residual autograft from the local
Plates are typically fixated to the superior surface bone should be applied to the nonunion site
of the clavicle, but good outcomes have also followed by a standard closure. In cases of
been reported with compression plates fixated to atrophic nonunion, an autograft from the iliac
the anteroinferior surface [70]. crest should be applied to the fracture site (see
Fig. 3.2) [7]. If there is significant bone loss
with shortening, then it has been proposed that
3.9 Open Reduction Internal an intercalary graft be used (Fig. 3.3) [7, 59].
Fixation with Autogenous Bone The goal in treating these fractures is to restore
Graft the length equal to the uninjured contralateral
size. Preoperative radiographic and clinical
3.9.1 Surgical Method evaluations should be conducted to determine
the length of the uninjured side. In the instance
ORIF with a compression plate and iliac crest of clavicle shortening, it is generally accepted if
bone graft is considered the gold standard for the bone is shortened by 1 cm. If there is
treatment of clavicle nonunions [1, 7, 71]. Due significant bone loss with shortening, then it has
to the clavicle’s close proximity to the subcla- been proposed that an intercalary graft such as a
vian vascular bundle and brachial plexus, sur- tricortical autogenous iliac crest bone graft from
gical technique must be precise so that is used from the contralateral hip. Previously, it
neurovascular injury is avoided [29]. This sur- has been suggested that low-contact dynamic
gical process involves carefully dividing and compression plates are a superior choice for
preserving the myofascial layer and identifying fixation of clavicular nonunion due to its
the superior surface of the clavicle. Once this is increased ability to be contoured and preserve
accomplished, the two ends of the nonunion blood supply to the underlying bone fragments
must be identified and mobilized. Derotation of due the plate’s structured undersurface [72].
the (usually anteriorly rotated) distal fragment There have been numerous reports of favourable
typically allows the superior surface to be outcomes and high rates of success (up to 100%)
exposed which enables plate fixation on the flat with this method of intervention [54, 58, 60, 69,
superior surface. The proximal and distal ends of 74, 75]. It is important that surgeons make cer-
the nonunion are then reduced. If there is excess tain that an implant of the correct size and length
callous on the superior surface, then it should be is used in order to aid in successful union of the
ronguered away to create a flat superior surface fracture. The availability of precontoured plates
that facilitates placement of the plate on the designed specifically for the clavicle, which
bone. This excess callous should be saved, have been shown to be biomechanically equiv-
morcellized, and later inserted into the nonunion alent to compression plates, makes them the
site. The use of a lag screw or small K-wire to implant of choice in this setting [7, 30].
Table 3.2 Treatment of clavicular nonunion with various methods of operative intervention
82
Study No. of patients Male Female Age Age Hyper-tropic Atrophic Approach Complications Hardware
who achieved (y) (min) (y) (max) removal
union
Ballmer et al. 35/37 (16 patients 22 15 14 54 13 24 Decortification with plate 2 persistent 23
[54] (mean had primary osteosynthesis. symptomatic
follow-up 8.6 operative Autogenous cancellous nonunions
y) treatment) bone graft (24 cases),
tricortical, iliac crest
intercalary graft (9)
Bradbury 31/32 23 9 17 60 21 | 10 17 AO reconstruction none 13
et al. [55] +1 true pseudoarthrosis plate with autologous
(mean cancellous bone graft, 15
follow-up 7 y AO dynamic compression
post-surgery) plate with autologous
cancellous bone graft
Boyer and 7/7 5 2 17 55 0 7 Excision with 3.5-mm none N/A
Axelrod [68] pelvic construction plate
or dynamic compression
plate, lag screw for
interfragmentary
compression, cancellous
bone graft
Davids et al. 14/14 8 6 19 74 5 9 AO reconstruction plate 1 patient had 6 (1 year
[56] with iliac crest bone graft refracture 1 year post-surgery)
post-hardware
removal
Ebraheim 15/16 9 7 15 52 11 5 ORIF with reconstruction 2 cases persistent 1
et al. [57] or dynamic plate, mild pain, 1
(mean autogenous bone graft in hardware failure
follow-up 14 cases, double plating in (healed
12.9 mo) 3 cases post-revision), 1
hardware failure
(persistent
(continued)
LauraA. Schemitsch, MA et al.
3
Table 3.2 (continued)
union
nonunion, pain
subsided)
Clavicle Nonunions
Endrizzi et al. 42/47 (1 patient 33 14 12 68 N/A N/A ORIF-27 curved pelvic 3 revision surgeries 10
[58] LTF, 1 deceased) reconstruction plates, 16 due to implant
straight pelvic loosening
reconstruction plates, 4
straight dynamic
compression plates
14 cases demineralized
bone matrix, 3 autogenous
iliac crest graft, 1 resected
rib graft
Eskola et al. 20/22 16 8 22 79 – – Twenty-one cancellous 4 cases of clavicle –
[37] +2 patients treated bone grafts were executed, shortening
with resection with eighteen of these
involving rigid plate
fixations and one
Kirschner pin fixation. In
two cases, bone grafting
was used solely. In one
case, only plate fixation
was performed
Kabak et al. 31/33 19 14 19 66 8 25 ORIF with dynamic 1 incomplete 5 (CDP)
[40] compression plate brachial plexus 2 (LC-DCP)
(DCP) or low-contact palsy, 2 revision
(LC-DCP) and autogenous surgeries went on
corticancellous chips or to union
sculptured graft
Laursen and 16/16 10 6 17 62 5 11 Compression plate fixation 1 revision due to 4
Dossing [60] with autologous screw loosening, 1
cancellous bone graft
(continued)
83
Table 3.2 (continued)
84
union
(mean revision due to
follow-up 24 fracture of plate
mo)
Manske and 10/10 7 3 16 60 – – Compression plate and None
Szabo [62] iliac crest bone graft
Olsen et al. 15/16 10 6 13 55 6 |9 ORIF with 3.5-mm plate 1 persistent In patient
[69] 1 failed osteosynthesis and autologous iliac crest nonunion, 2 with
bone graft patients “fair” persistent
outcome nonunion,
hardware
removed due
to loosening
of screws
IM pinning
Capicotto 14/14 – – 18 62 – – Steinman pin fixation and None 14
et al. [14] only iliac crest bone graft
(mean
follow-up 4 y)
Enneking 13/14 10 4 19 83 3 11 ORIF with Rushpin None (aside from 1 3
et al. fixation and iliac crest case of nonunion)
[50] bone graft
(mean
follow-up 40
mo)
ORIF open reduction and internal fixation
AO Arbeitsgemeinschaft für Osteosynthesefragen
LauraA. Schemitsch, MA et al.
3.9.2 An Innovative Method In a series comparing the use of dynamic com-

pression plating (DCP) and low-contact dynamic
In 1997, Boyer and Axelrod [68] reported on an compression plating (LC-DCP), union was
innovative method for treating atrophic nonunion obtained in all cases where a LC-DCP plate was
of the clavicle. They bevelled the clavicle using used compared with 87.5% with the DCP plate
cuts at 45° to the long axis of the bone using an [40]. Better functional outcomes were obtained
oscillating saw to minimize shortening and pro- in patients who were treated with a LC-CDP
vide a flat surface of healthy viable bone. Each plate. In a series of eight patients with symp-
patient’s nonunion was fixed with a dynamic tomatic nonunion who were treated with surgical
compression plate or, in one case, a pelvic intervention, all eight patients achieved clinical
reconstruction plate. A lag screw provided and radiographic union [61]. Patients were trea-
intra-fragmentary compression in all seven cases ted using a 3.5-mm low-contact dynamic com-
and the nonunion site was grafted with autoge- pression plate (LC-DCP) and an autologous iliac
nous cancellous bone. All patients achieved crest graft. Despite a relatively short rate of
union between six and 12 weeks with full pain follow-up (average 8 months), the authors stated
relief and return to normal activity levels. The that all the patients were highly satisfied with
authors determined that each patient was very their treatment.
satisfied with their result and the slight narrowing In a retrospective review of twenty-four non-
(<2 cm) of the shoulder’s width did not cause the unions between 1994 and 2001 [75], twenty-two
patients to be dissatisfied with the cosmetic result out of twenty-four nonunions eventually healed
of the intervention. and had favourable outcomes. This series was
treated with operative fixation using a DCP or
reconstruction plate along with autogenous bone
3.9.3 Literature Review grafting. The time from injury to operation ran-
ged from three months to 29 months with the
Olsen et al. [69] reported on sixteen consecutive mean time being 10.2 months. Post-operative
cases of clavicular nonunion that were initially follow-up ranged from six to 74 months with a
treated conservatively. Each patient was treated mean follow-up of 42.1 months. Considering the
with ORIF and iliac crest bone grafting. Twelve average length of follow-up, this series provides
of the patients had an excellent result with the some indication of long-term success of opera-
remaining two receiving a good grade, one fair tive treatment for this injury.
and one failure resulting in persistent nonunion. In a series of fourteen patients with delayed
This patient had initially been treated with a and nonunion of midshaft clavicle fractures who
resection of the medial fragment. were treated with a reconstruction plate and iliac
In a series with long follow-up from the time crest bone grafting, all patients achieved radio-
of surgical intervention (7 years), thirty-one out graphic bony union at 10–30 weeks post-surgery
of thirty-two nonunions were treated successfully with excellent results [56]. This study had the
[55]. The majority of these clavicle nonunions benefit of a long period of follow-up with a mean
were the result of high-energy traumas. Fifteen of 60 months (16–101 months). In a series of ten
patients were treated with dynamic compression consecutive patients who received operative
plates and autologous bone graft, and seventeen treatment for their midshaft clavicle nonunions,
were treated with reconstruction plates and all ten patients achieved union at ten weeks
autologous cancellous bone graft. Six compres- post-operatively and had a full and painless range
sion plates and seven reconstruction plates were of motion [62].
removed with the authors recommending recon- Eskola et al. reported on an adult population
struction plates to be used in treatment of this (aged 22–79 years) with an average follow-up of
injury due to its ability to be contoured more three years. Twenty-four patients were initially
easily to the complex shape of the clavicle bone. treated nonoperatively, eventually resulting in
nonunion, and then subsequently received sur- group was 3%, and in the nonoperative group, it
gical intervention [37]. Twenty-one cancellous was 14%. At one-year follow-up, the patients in
bone grafts were performed, with eighteen of the operative group were more likely to be sat-
these involving rigid plate fixation and one isfied with the appearance of the shoulder
Kirschner pin fixation. In two cases, bone graft- (P = 0.001) and the shoulder in general
ing was used solely. In one case, only plate fix- (P = 0.002). Hardware removal was the most
ation was performed. In two cases, the medial common intervention for the surgical group.
end of the clavicle was resected. For the In 2013, Robinson et al. [76] reported on 200
twenty-two cases that were treated with fixation patients who had an isolated displaced fracture of
and bone grafting, twenty achieved bony union the middle third of the clavicle. At two weeks
of the clavicle fracture. The study also found that post-injury, 95 patients were allocated to surgery
the clavicle was shortened in four cases which and 105 to nonoperative treatment. Union was
resulted in some weakness and decreased range confirmed by 3D computer tomography (CT).
of motion. This is consistent with other studies Functional assessment was measured through
have shown decreasing functional outcomes with Short Form (SF-12), the Disabilities of the Arm,
increasing deformity. Shoulder, and Hand (DASH) questionnaire, the
It is important to remember these studies were Constant questionnaire, and shoulder motion.
performed before the routine availability of pre- There was a nonunion rate of 1% in the operative
contoured plates: Their improved biomechanical group and a significantly higher rate of 17% in
strength (compared to reconstruction plates) and the nonoperative group. Thirteen patients with
superior fit (compared to straight compression nonunion in the nonoperative group underwent
plates) make them the implant of choice for the surgical intervention and eight patients experi-
treatment of clavicular nonunion. enced delayed union, with union occurring
between 6 and 12 months post-injury. The ORIF
group had better DASH scores at 3- and
3.9.4 Multicenter, Randomized 12-month follow-up. The SF-12 physical and
Controlled Trials mental scores did not differ between groups. The
mean patient age was 33.5 and 32, respectively,
There have been two large multicenter, ran- in these studies, and follow-up rates were 84 and
domized controlled trials aimed at comparing the 89%. Both of these studies indicated that primary
results of open reduction and plate fixation with operative intervention for displaced midshaft
nonoperative treatment for displaced midshaft clavicle fractures decreases the rate of nonunion.
clavicle fractures (Table 3.3). In 2007, the
Canadian Orthopedic Trauma Society [18] con-
ducted a series involving 132 patients with this 3.10 ORIF Without Bone Grafting
injury. Final follow-up was available for
sixty-two patients allocated to operative treat- Other studies have suggested that autogenous
ment and forty-nine patients allocated to nonop- bone grafting may not be necessary in the treat-
erative treatment. Outcomes were measured ment of clavicular nonunion. Baker and Mullett
using standard clinical follow-up, the Disabilities [77] reported on a series of fifteen patients who
of the Arm, Shoulder, and Hand (DASH) ques- achieved clinical and radiographic union fol-
tionnaire, the Constant questionnaire, and plain lowing fixation with precontoured locking plates.
radiographs. DASH and Constant scores were One patient underwent hardware removal due to
significantly better in the operative group at each local discomfort but the remaining patients
time point. The rate of nonunion in the operative experienced favourable outcomes. Ramoutar
3
Table 3.3 Multicenter randomized, controlled trials comparing open reduction and internal fixation with nonoperative treatment of displaced midshaft clavicle fractures
Clavicle Nonunions
Study Total number Operative Nonoperative Mean Surgical Primary outcome measures Rate of nonunion (%) Follow-up
of patients group (n) group (n) age approach (%)
(years)
Robinson 200 95 105 32 ORIF Union confirmed by 3D ORIF: 1.2 89
et al. [76] using computer tomography (CT). Nonoperative: 17
locking Functional assessment (13 nonunions in
clavicle measured through Short nonoperative group
plate Form (SF-12), the underwent surgical
Disabilities of the Arm, intervention)
Shoulder, and Hand
(DASH) questionnaire, the
constant questionnaire, and
shoulder motion
Canadian 131 62 49 33.5 ORIF Standard clinical follow-up, ORIF: 3 84
Orthopedic using the Disabilities of the Arm, Nonoperative: 14.3
Trauma small Shoulder, and Hand (1 patient allocated to the
Society fragment (DASH) questionnaire, the operative group declined
[18] plate constant questionnaire and surgery and nonunion
plain radiographs persisted at 1 y follow-up
but was still included in the
operative group due to the
“intention to treat”
principle)
ORIF open reduction and internal fixation
87
et al. [78] reported a series of eleven patients who who are treating smokers with clavicular non-
also achieved union without bone graft. In this union. Iliac crest bone graft may be a better
series, the patients were fixated with fragment option for these patients.
decortication and compression plating.
3.12 Vascularized Bone Graft

3.11 ORIF with Allograft Bone Reconstruction
Substitute
Momberger et al. [87] presented three patients
Iliac crest bone grafting has been shown to be who had each experienced a completely dis-
very effective for nonunion healing, although placed midshaft clavicle fracture. After an aver-
various studies have shown that significant age of 3.7 procedures, each clavicle had failed to
complications can occur. These complications unite and was characterized by segmental bone
can include temporary or prolonged pain at the loss. Each patient was treated with vascularized
harvest site, pelvic fracture, infection, persistent fibula transfer. At an average follow-up of
drainage, and sensory disturbances [79–86]. In 2.8 years, each clavicle had achieved bony union
an effort to reduce these complications, certain as well as decreased pain and improved shoulder
alternatives to autogenous bone grafting have function. This method of intervention could be
been proposed. Demineralized bone matrix considered following repeated failure of clavicle
(DBM) has been shown to be effective in the fractures to unite using traditional surgical
treatment of humeral and tibial nonunions [82, methods, especially when segmental bone loss is
84]. However, the literature regarding its use and present.
efficacy for treatment of clavicle nonunions is
limited. Endrizzi et al. [58] reported on a series of
forty-seven patients who were treated for clav- 3.13 Intramedullary Pin Fixation
icular nonunion. Fourteen of these patients
received DBM, and although they reported a Intramedullary (IM) pinning with autogenous
high rate of union (93%), the study did not focus bone grafting has several potential advantages as a
solely on this treatment type. In a recent study, treatment option for patients with clavicle non-
conducted by Riggenbach et al. in 2011 [49], union. A smaller initial incision, less tissue dis-
nineteen patients were treated with ORIF and section, and less periosteal stripping are necessary
allograft bone substitute. Twelve patients [50]. A small incision for removal of the pin under
received platelet-rich plasma (PRP) with DBM, local anaesthesia (if necessary) results in
and the other seven received allograft cortico- improved cosmetic appearance and a simpler and
cancellous chips. Sixteen of the patients achieved less invasive method of hardware removal [50,
bony union, while three required revision due to 88]. It is also likely that IM pinning will cause
failure, including two “catastrophic failures.” decreased hardware irritation [7]. Boehme et al.
Two of the three persistent nonunions went on to [88] reported on twenty-one patients (from a series
heal fully. A final clinical healing rate of 95% in of fifty) who were treated with ORIF using a
this study indicates that DBM may be a viable modified Hagie intramedullary pin and autoge-
treatment option for clavicle nonunion. It is nous bone grafting. Twenty of the patients
interesting to note that the three patients who achieved union. Enneking et al. [50] used Rush pin
required revision were smokers and that non- fixation along with autogenous bone grafting in a
union was more common in smokers (P = 0.08) series of fourteen patients, with thirteen of these
compared with patients who did not smoke. This nonunions going on to heal with good results.
could have potential implications for surgeons Three patients required hardware removal.
Capicotto et al. [14] reviewed fourteen patients could be removed. In 1968, Rowe specified that
with clavicle nonunions caused mainly by for patients who had had several unsuccessful
high-velocity injuries. Three of these patients were attempts at bone grafts and were “sufficiently
treated with plating that failed, and in one case, the disabled,” resection of the clavicle could result in
fracture was pathological through irradiated bone. “surprisingly good function and cosmetic
These nonunions were treated with intramedullary appearance” [4]. In 1986, Wood [99] reported on
Steinman pin fixation along with iliac crest bone five cases of total claviculectomy and determined
graft. All fourteen nonunions went on to achieve that the procedure resulted in excellent functional
bony union and had their hardware removed outcomes. More recently, Krishnan et al. [100]
between 7 and 24 weeks post-surgery. Two conducted a retrospective chart review of six
refractures occurred through osteopenic bone, but cases of unilateral claviculectomy with a mean
this was the only complication of significance. follow-up of 5.7 years. They found that in a
Despite the increased ease of hardware relatively young cohort of patients, there was
removal with this technique, IM fixation is good restoration of range of motion, minimal
weaker biomechanically [7] and there have been strength deficits, and improved pain relief. They
reported complications such as pin migration and report that in situations where restoration of
breakage that should be taken into consideration clavicular anatomy is impossible, total clav-
[89–91]. McKee [7] recommended that a ran- iculectomy can result in good functional out-
domized, prospective trial be conducted that comes. However, it is important to note that three
compares plate and IM fixation for the treatment of the six patients had a major complication as
of displaced midshaft clavicle fractures in order well as four patients experiencing infection (two
to better assess their use in relation to this injury: deep and two superficial).
two such studies are currently underway. Wu Other studies have raised questions about
et al. [63] conducted a retrospective review resection procedures used to treat clavicle non-
comparing plate with IM fixation in thirty-three union. Connolly and Dehne [46] reported that
patients with midshaft asceptic clavicle non- resection can result in delayed problems such as
union. They found that nine of eleven patients thoracic outlet syndrome and should therefore be
treated with plate fixation healed, while sixteen avoided. In 1992, Rockwood and Wirth argued
of eighteen patients treated with IM fixation for the preservation of the clavicle due to low
healed. However, a more comprehensive study patient satisfaction following total claviculec-
with a larger patient population is required. tomy [53]. In 2007, Wessel and Schapp [101]
reported on six cases of total claviculectomy.
Good results were seen in one chronic osteitis
3.13.1 Resection case and two malignancy cases. Three
post-traumatic cases had poor results due to
Resection of the clavicle is a salvage technique persistent pain. All six patients regained full
used to alleviate pain and discomfort caused by range of motion of the shoulder. No final con-
severe or recalcitrant nonunion in low demand clusions were drawn due to the small sample size
patients. Although absolute indications for clav- and conflicting outcomes. Overall, the literature
iculectomy are rare, the primary indication is on this method of intervention for clavicle non-
multiple failed conventional surgeries and some union is sparse, with small reported sample
relative indications have included clavicular populations. Due to limited knowledge on the
tumour, vascular injury, and associated unipolar effectiveness of this procedure, the importance of
or bipolar dislocation [48, 92–98]. the clavicle for upper extremity function and
Historically, Abbot and Lucas stated that as stability, and the availability of modern methods
long as the medial and distal thirds of the clavicle of intervention, total claviculectomy should be
were left intact, the middle third of the clavicle regarded solely as a salvage procedure [7].
3.14 Conclusion 11. Robinson CM, Cairns DA. Primary nonoperative

treatment of displaced lateral fractures of the
clavicle. J Bone Joint Surg Am. 2004;86-a
Nonunion is a complication of clavicular fractures (4):778–82.
that typically results in pain and loss of shoulder 12. Seo GS, Aoki J, Karakida O, Sone S. Nonunion of a
function and mobility. Nonoperative treatment, medial clavicular fracture following radical neck
especially in displaced midshaft clavicle fractures, dissection: MRI diagnosis. Orthopedics. 1999;22
(10):985–6.
can result in nonunion that causes increased def- 13. Throckmorton T, Kuhn JE. Fractures of the medial
icits in shoulder function and pain. Due to the end of the clavicle. J Shoulder Elbow Surg. 2007;16
significant rate of nonunion in displaced fractures (1):49–54.
in young and active adults, operative intervention, 14. Capicotto PN, Heiple KG, Wilbur JH. Midshaft
clavicle nonunions treated with intramedullary
which has been shown to have high success rates Steinman pin fixation and onlay bone graft.
for pain relief and bony union, should be explored J Orthop Trauma. 1994;8(2):88–93.
as an option in cases where surgeons believe the 15. Craig EV. Fractures of the clavicle. In: Rock-
patient would benefit. If nonunion does occur, wood CA, Green DP, Bucholz RW, Heckman JD,
editors. Rockwood and Green’s fractures in adults.
plate fixation with autogenous bone grafting is a Philadelphia: Lippincott-Raven; 1996. p. 1109–61.
reliable operation with a high success rate and low 16. Neer CS. Fractures of the clavicle. In: Rock-
complication rate. wood CA, Green DP, editors. Fractures in adults.
2nd ed. Philadelphia: Lippincott; 1984. p. 707–13.
17. Robinson CM. Fractures of the clavicle in the adult.
Epidemiology and classification. J Bone Joint Surg
References Br. 1998;80(3):476–84.
18. Canadian Orthopaedic Trauma Society. Nonopera-
1. Post M. Current concepts in the treatment of tive treatment compared with plate fixation of
fractures of the clavicle. Clin Orthop Relat Res. displaced midshaft clavicular fractures. A multicen-
1989;245:89–101. ter, randomized clinical trial. J Bone Joint Surg Am.
2. Neer CS 2nd. Nonunion of the clavicle. JAMA. 2007;89(1):1–10.
1960;172:1006–11. 19. Neviaser JS. The treatment of fractures of the
3. Postacchini F, Gumina S, De Santis P, Albo F. clavicle. Surg Clin N Am. 1963;43:1555–63.
Epidemiology of clavicle fractures. J Shoulder 20. Ghormley RK, Black JR, Cherry JH. Ununited
Elbow Surg. 2002;11(5):452–6. fractures of the clavicle. Am J Surg. 51(2):343–349.
4. Rowe CR. An atlas of anatomy and treatment of 21. Hill JM, McGuire MH, Crosby LA. Closed treat-
midclavicular fractures. Clin Orthop Relat Res. ment of displaced middle-third fractures of the
1968;58:29–42. clavicle gives poor results. J Bone Joint Surg Br.
5. Crenshaw AH. Campbell’s operative orthopaedics. 1997;79(4):537–9.
St Louis: Mosby; 1996. 22. Zlowodzki M, Zelle BA, Cole PA, Jeray K,
6. Goldberg JA, Bruce WJ, Sonnabend DH, McKee MD. Treatment of acute midshaft clavicle
Walsh WR. Type 2 fractures of the distal clavicle: fractures: systematic review of 2144 fractures: on
a new surgical technique. J Shoulder Elbow Surg. behalf of the evidence-based orthopaedic Trauma
1997;6(4):380–2. working group. J Orthop Trauma. 2005;19(7):504–7.
7. McKee MD. Fractures of the clavicle. In: 23. Brinker MR, Edwards TB, O’Connor DP. Estimat-
Bucholz RU, Heckman JD, Court-Brown CM, ing the risk of nonunion following nonoperative
Tornetta P, editors. Rockwood and Green’s frac- treatment of a clavicular fracture. J Bone Joint Surg
tures in adults. 7th ed. Philadelphia: Lippincott Am. 2005;87(3):676–7; author reply 7.
Williams and Wilkins; 2010. p. 1106–43. 24. Edwards DJ, Kavanagh TG, Flannery MC. Frac-
8. Robinson CM, Court-Brown CM, McQueen MM, tures of the distal clavicle: a case for fixation.
Wakefield AE. Estimating the risk of nonunion Injury. 1992;23(1):44–6.
following nonoperative treatment of a clavicular 25. McKee MD, Pedersen EM, Jones C, Stephen DJ,
fracture. J Bone Joint Surg Am. 2004;86-a(7):1359– Kreder HJ, Schemitsch EH, et al. Deficits following
1365. nonoperative treatment of displaced midshaft clav-
9. Rockwood CA. Fractures of the outer clavicle in icular fractures. J Bone Joint Surg Am. 2006;88
children and adults. J Bone Joint Surg Br. (1):35–40.
1982;64B:642. 26. Nordqvist A, Petersson CJ, Redlund-Johnell I.
10. Webber MC, Haines JF. The treatment of lateral Mid-clavicle fractures in adults: end result study
clavicle fractures. Injury. 2000;31(3):175–9. after conservative treatment. J Orthop Trauma.
1998;12(8):572–6.
27. McKee MD, Schemitsch EH, Stephen DJ, Kreder HJ, cell metabolic activity. J Orthop Trauma. 2005;19
Yoo D, Harrington J. Functional outcome following (1):17–22.
clavicle fractures in polytrauma patients. J Trauma 43. Brighton CT, Pollack SR. Treatment of recalcitrant
Acute Care Surg. 1999;47(3):616. non-union with a capacitively coupled electrical
28. Nowak J, Holgersson M, Larsson S. Can we predict field. A preliminary report. J Bone Joint Surg Am.
long-term sequelae after fractures of the clavicle 1985;67(4):577–85.
based on initial findings? A prospective study with 44. Andersen K, Jensen PO, Lauritzen J. Treatment of
nine to ten years of follow-up. J Shoulder Elbow clavicular fractures. figure-of-eight bandage versus
Surg. 2004;13(5):479–86. a simple sling. Acta Orthop Scand. 1987;58(1):71–
29. Ebraheim NA, Xu R, Ahmed M, Frogament AD. 4.
The anatomic relation of the clavicle to the subcla- 45. Potter JM, Jones C, Wild LM, Schemitsch EH,
vian neurovascular bundle. Orthop Int Ed. McKee MD. Does delay matter? The restoration of
1987;5:279–81. objectively measured shoulder strength and
30. Martetschlager F, Gaskill TR, Millett PJ. Manage- patient-oriented outcome after immediate fixation
ment of clavicle nonunion and malunion. J Shoulder versus delayed reconstruction of displaced midshaft
Elbow Surg. 2013;22(6):862–8. fractures of the clavicle. J Shoulder Elbow Surg.
31. McKee MD, Wild LM, Schemitsch EH. Midshaft 2007;16(5):514–8.
malunions of the clavicle. J Bone Joint Surg Am. 46. Connolly JF, Dehne R. Nonunion of the clavicle
2003;85-a(5):790–797. and thoracic outlet syndrome. J Trauma. 1989;29
32. Wilkins RM, Johnston RM. Ununited fractures of the (8):1127–1132; discussion 32-3.
clavicle. J Bone Joint Surg Am. 1983;65(6):773–8. 47. Edvardsen P, Odegard O. Treatment of posttrau-
33. Stanley D, Norris SH. Recovery following fractures matic clavicular pseudarthrosis. Acta Orthop Scand.
of the clavicle treated conservatively. Injury. 1977;48(5):456–7.
1988;19(3):162–4. 48. Middleton SB, Foley SJ, Foy MA. Partial excision
34. White RR, Anson PS, Kristiansen T, Healy W. of the clavicle for nonunion in National Hunt
Adult clavicle fractures: relationship between mech- Jockeys. J Bone Joint Surg Br. 1995;77(5):778–80.
anism of injury and healing. Orthop Trans. 49. Riggenbach MD, Jones GL, Bishop JY. Open
1989;13:514–5. reduction and internal fixation of clavicular non-
35. Allman FL Jr. Fractures and ligamentous injuries of unions with allograft bone substitute. Int J Shoulder
the clavicle and its articulation. J Bone Joint Surg Surg. 2011;5(3):61–7.
Am. 1967;49(4):774–84. 50. Enneking TJ, Hartlief MT, Fontijne WP. Rushpin
36. Wick M, Muller EJ, Kollig E, Muhr G. Midshaft fixation for midshaft clavicular nonunions: good
fractures of the clavicle with a shortening of more results in 13/14 cases. Acta Orthop Scand. 1999;70
than 2 cm predispose to nonunion. Arch Orthop (5):514–6.
Trauma Surg. 2001;121(4):207–11. 51. Sakellarides H. Pseudoarthrosis of the clavicle: a
37. Eskola A, Vainionpää S, Myllynen P, Pätiälä H, report of 20 cases. J Bone Joint Surg Am.
Rokkanen P. Outcome of clavicular fracture in 89 1961;43A:130–8.
patients. Arch Orthop Trauma Surg. 1986;105 52. Taylor AR. Nonunion of fractures of the clavicle: a
(6):337–8. review of 31 cases. J Bone Joint Surg Br.
38. Kulshrestha V, Roy T, Audige L. Operative versus 1969;51-B:568–9.
nonoperative management of displaced midshaft 53. Rockwood CA, Wirth MA. Don’t throw away the
clavicle fractures: a prospective cohort study. clavicle. Orthop Trans. 1992;16:763.
J Orthop Trauma. 2011;25(1):31–8. 54. Ballmer FT, Lambert SM, Hertel R. Decortication
39. Murray IR, Foster CJ, Eros A, Robinson CM. Risk and plate osteosynthesis for nonunion of the
factors for nonunion after nonoperative treatment of clavicle. J Shoulder Elbow Surg. 1998;7(6):581–5.
displaced midshaft fractures of the clavicle. J Bone 55. Bradbury N, Hutchinson J, Hahn D, Colton CL.
Joint Surg Am. 2013;95(13):1153–8. Clavicular nonunion. 31/32 healed after plate fixa-
40. Kabak S, Halici M, Tuncel M, Avsarogullari L, tion and bone grafting. Acta Orthop Scand. 1996;67
Karaoglu S. Treatment of midclavicular nonunion: (4):367–70.
comparison of dynamic compression plating and 56. Davids PH, Luitse JS, Strating RP, van der
low-contact dynamic compression plating techniques. Hart CP. Operative treatment for delayed union
J Shoulder Elbow Surg. 2004;13(4):396–403. and nonunion of midshaft clavicular fractures: AO
41. Moghaddam A, Zimmermann G, Hammer K, reconstruction plate fixation and early mobilization.
Bruckner T, Grutzner PA, von Recum J. Cigarette J Trauma. 1996;40(6):985–6.
smoking influences the clinical and occupational 57. Ebraheim NA, Mekhail AO, Darwich M. Open
outcome of patients with tibial shaft fractures. reduction and internal fixation with bone grafting of
Injury. 2011;42(12):1435–42. clavicular nonunion. J Trauma. 1997;42(4):701–4.
42. Gullihorn L, Karpman R, Lippiello L. Differential 58. Endrizzi DP, White RR, Babikian GM, Old AB.
effects of nicotine and smoke condensate on bone Nonunion of the clavicle treated with plate fixation:
a review of forty-seven consecutive cases. J Shoul- autogenous bone grafting for clavicular non-union.
der Elbow Surg. 2008;17(6):951–3. Injury. 2004;35(6):575–9.
59. Jupiter JB, Leffert RD. Non-union of the clavicle. 76. Robinson CM, Goudie EB, Murray IR, Jenkins PJ,
Associated complications and surgical management. Ahktar MA, Read EO, et al. Open reduction and
J Bone Joint Surg Am. 1987;69(5):753–60. plate fixation versus nonoperative treatment for
60. Laursen MB, Dossing KV. Clavicular nonunions displaced midshaft clavicular fractures: a multicen-
treated with compression plate fixation and cancel- ter, randomized, controlled trial. J Bone Joint Surg
lous bone grafting: the functional outcome. J Shoul- Am. 2013;95(17):1576–84.
der Elbow Surg. 1999;8(5):410–3. 77. Baker JF, Mullett H. Clavicle non-union: autolo-
61. Leupin S, Jupiter JB. LC-DC plating with bone gous bone graft is not a necessary augment to
graft in posttraumatic nonunions in the middle third internal fixation. Acta Orthop Belg. 2010;76
of the clavicle. Swiss Surg. 1998;4(2):89–94. (6):725–9.
62. Manske DJ, Szabo RM. The operative treatment of 78. Ramoutar DN, Rodrigues J, Quah C, Boulton C,
mid-shaft clavicular non-unions. J Bone Joint Surg Moran CG. Judet decortication and compression
Am. 1985;67(9):1367–71. plate fixation of long bone non-union: is bone graft
63. Wu CC, Shih CH, Chen WJ, Tai CL. Treatment of necessary? Injury. 2011;42(12):1430–4.
clavicular aseptic nonunion: comparison of plating 79. Ahlmann E, Patzakis M, Roidis N, Shepherd L,
and intramedullary nailing techniques. J Trauma. Holtom P. Comparison of anterior and posterior
1998;45(3):512–6. iliac crest bone grafts in terms of harvest-site
64. Jupiter JB, Ring D. Fractures of the clavicle. In: morbidity and functional outcomes. J Bone Joint
Iannotti JP, Williams GR, editors. Disorders of the Surg Am. 2002;84-a(5):716–720.
shoulder: diagnosis and management. Philadelphia: 80. Arrington ED, Smith WJ, Chambers HG, Buck-
Lippincott Williams & Wilkins; 1999. nell AL, Davino NA. Complications of iliac crest
65. Sadiq S, Waseem M, Peravalli B, Doyle J, Dun- bone graft harvesting. Clin Orthop Relat Res.
ningham T, Muddu BN. Single or double plating for 1996;329:300–9.
nonunion of the clavicle. Acta Orthop Belg. 81. Fowler BL, Dall BE, Rowe DE. Complications
2001;67(4):354–60. associated with harvesting autogenous iliac bone
66. Rosenberg N, Neumann L, Wallace AW. Functional graft. Am J Orthop (Belle Mead, NJ). 1995;24
outcome of surgical treatment of symptomatic (12):895–903.
nonunion and malunion of midshaft clavicle frac- 82. Friedlaender GE, Perry CR, Cole JD, Cook SD,
tures. J Shoulder Elbow Surg. 2007;16(5):510–3. Cierny G, Muschler GF, et al. Osteogenic protein-1
67. Simpson NS, Jupiter JB. Clavicular nonunion and (bone morphogenetic protein-7) in the treatment of
malunion: evaluation and surgical management. tibial nonunions. J Bone Joint Surg Am. 2001;83-A
J Amer Acad Orthop Surg. 1996;4(1):1–8. Suppl 1(Pt 2):S151–8.
68. Boyer MI, Axelrod TS. Atrophic nonunion of the 83. Goulet JA, Senunas LE, DeSilva GL, Green-
clavicle: treatment by compression plate, lag-screw field ML. Autogenous iliac crest bone graft. Com-
fixation and bone graft. J Bone Joint Surg Br. plications and functional assessment. Clin Orthop
1997;79(2):301–3. Relat Res. 1997;339:76–81.
69. Olsen BS, Vaesel MT, Sojbjerg JO. Treatment of 84. Hierholzer C, Sama D, Toro JB, Peterson M,
midshaft clavicular nonunion with plate fixation and Helfet DL. Plate fixation of ununited humeral shaft
autologous bone grafting. J Shoulder Elbow Surg. fractures: effect of type of bone graft on healing.
1995;4(5):337–44. J Bone Joint Surg Am. 2006;88(7):1442–7.
70. Stufkens SA, Kloen P. Treatment of midshaft 85. Keller EE, Triplett WW. Iliac bone grafting: review
clavicular delayed and non-unions with anteroinfe- of 160 consecutive cases. J Oral Maxillofac Surg.
rior locking compression plating. Arch Orthop 1987;45(1):11–4.
Trauma Surg. 2010;130(2):159–64. 86. Kurz LT, Garfin SR, Booth RE Jr. Harvesting
71. Pyper JB. Non-union of fractures of the clavicle. autogenous iliac bone grafts. A review of complica-
Injury. 1978;9(4):268–70. tions and techniques. Spine. 1989;14(12):1324–31.
72. Mullaji AB, Jupiter JB. Low-contact dynamic 87. Momberger NG, Smith J, Coleman DA. Vascularized
compression plating of the clavicle. Injury. fibular grafts for salvage reconstruction of clavicle
1994;25(1):41–5. nonunion. J Shoulder Elbow Surg. 2000;9(5):389–94.
73. Pedersen M, Poulsen KA, Thomsen F, Kris- 88. Boehme D, Curtis RJ Jr, DeHaan JT, Kay SP,
tiansen B. Operative treatment of clavicular non- Young DC, Rockwood CA Jr. Non-union of
union. Acta Orthop Belg. 1994;60(3):303–6. fractures of the mid-shaft of the clavicle. Treatment
74. Khan SA, Shamshery P, Gupta V, Trikha V, with a modified Hagie intramedullary pin and
Varshney MK, Kumar A. Locking compression autogenous bone-grafting. J Bone Joint Surg Am.
plate in long standing clavicular nonunions with 1991;73(8):1219–26.
poor bone stock. J Trauma. 2008;64(2):439–41. 89. Lyons FA, Rockwood CA Jr. Migration of pins
75. O’Connor D, Kutty S, McCabe JP. Long-term used in operations on the shoulder. J Bone Joint
functional outcome assessment of plate fixation and Surg Am. 1990;72(8):1262–7.
90. Naidoo P. Migration of a Kirschner wire from the 96. Gurd FB. The treatment of complete dislocation of
clavicle into the abdominal aorta. Arch Emerg Med. the outer end of the clavicle: an hitherto undescribed
1991;8(4):292–5. operation. Ann Surg. 1941;113(6):1094–8.
91. Nordback I, Markkula H. Migration of Kirschner 97. Sonnabend D. The origin of the shoulder: a fairytale
pin from clavicle into ascending aorta. Acta Chir based on fact. 9th International Congress on
Scand. 1985;151(2):177–9. Surgery of the Shoulder Meeting; 2004 May 2–5;
92. Abbott LC, Lucas DB. The function of the clavicle; Washington, DC.
its surgical significance. Ann Surg. 1954;140 98. Spar I. Total claviculectomy for pathological frac-
(4):583–99. tures. Clin Orthop Relat Res. 1977;129:236–7.
93. Codman EA. Rupture of the supraspinatus tendon 99. Wood VE. The results of total claviculectomy. Clin
and other lesions in or about the subacromial bursa. Orthop Relat Res. 1986;207:186–90.
The shoulder. Boston: Thomas Todd; 1934. 100. Krishnan SG, Schiffern SC, Pennington SD, Rim-
94. Elkin DC, Cooper FW Jr. Resection of the clavicle lawi M, Burkhead WZ Jr. Functional outcomes after
in vascular surgery. J Bone Joint Surg Am. total claviculectomy as a salvage procedure. A series
1946;28:117–9. of six cases. J Bone Joint Surg Am. 2007;89
95. Green RM, Waldman D, Ouriel K, Riggs P, (6):1215–9.
Deweese JA. Claviculectomy for subclavian venous 101. Wessel RN, Schaap GR. Outcome of total clav-
repair: long-term functional results. J Vasc Surg. iculectomy in six cases. J Shoulder Elbow Surg.
2000;32(2):315–21. 2007;16(3):312–5.
Proximal Humerus Nonunions
4
Ethan S. Lea, MD, MSc and Philip R. Wolinsky, MD
4.1 Introduction 4.2 Epidemiology
Fractures of the proximal humerus are common The incidence of nonunions of the surgical neck
injuries, accounting for about 5% of all fractures of the proximal humerus has been cited to be as
of the appendicular skeleton [1, 2]. The incidence high as 20% [9]. The challenge with these figures
increases as patients age [3] since this injury is that they are generated from tertiary care
occurs more frequently in individuals who have centers where patients have been referred for
decreased bone mineral density [4, 5]. As the treatment of their nonunions rather than from
population ages, the prevalence of proximal centers that provide the initial care for all of these
humerus fractures will rise with a projected injuries, which inflates the perceived incidence of
3-fold increase over the next three decades [6]. nonunions.
The majority of these fractures (approximately A large single-center clinical study found the
85% [7]) occurs as the result of a low energy rate of nonunion of fractures of the proximal
injury and are minimally displaced, have a stable humerus to be only 1.1% [10]. Court-Brown and
configurations, and are amenable to McQueen prospectively followed 1027 consec-
non-operative treatment. utive proximal humerus fractures treated at their
Some of these fractures will go on to develop institution, which provides care for all the prox-
a nonunion. Cadet et al. stated that the biological imal humerus fractures in a population of
factors that caused or contributed to the devel- approximately 650,000 individuals. The patients
opment of the nonunion needed to be treated in were treated at the discretion of the admitting
order for treatment of the nonunion to succeed surgeons: 89.1% were treated non-operatively,
[8]. As the number of these fractures grows, it 7.1% were treated with internal fixation, and
seems logical that the subset that fails to unite 3.8% with a hemiarthroplasty. Only 11 of the
will increase as well. 995 patients who were followed for one year
developed a nonunion. The fractures were clas-
sified using the AO/OTA and Neer classifications
and comminuted bifocal fractures (AO/OTA
B2.3) had the highest incidence of nonunion
with two of six patients failing to unite. The
E.S. Lea P.R. Wolinsky (&) authors stressed that while the nonunion rate of
Department of Orthopedic Surgery, University of 33.3% for this specific subset of patients was
California at Davis Medical Center, 4860 Y St, Suite high, these fractures were quite rare and rate
3800, Sacramento, CA 95608, USA
e-mail: philipwolinsky@hotmail.com calculations based on such small patient numbers

96 E.S. Lea, MD, MSc and P.R. Wolinsky, MD
may be unreliable. Displaced articular fractures of non-smokers for proximal humerus fractures
(AO/OTA C2.3) and Neer 4-part fractures had deemed appropriate for non-operative manage-
nonunion rates of 8.3 and 5.9% respectively, but ment [14]. Medical comorbidities including heart
the treatment of these groups were heterogeneous disease, hypertension, and chronic obstructive
and included some treated non-operatively and pulmonary disease have high rates of prevalence
some that underwent hemiarthroplasty. There- among patients with proximal humerus non-
fore, the true rates of nonunion for these subsets unions, which suggests an association between
are unknown. chronic medical comorbidities and nonunions,
Iyengar et al. performed a meta-analysis of 12 but the study designs demonstrating this associ-
studies with a total of 650 patients who under- ation could not infer causality. Factors such as
went non-operative treatment of their proximal alcohol abuse, other substance abuse, or psychi-
humerus fractures and found a 2% incidence of atric illnesses that affect patients’ ability to
nonunion (range 0–7%). These studies found that appropriately participate in their care have been
the incidence of nonunion varied little when described as a barrier to achieving union [13].
fractures were categorized using the Neer clas- Fracture characteristics including translation
sification: four studies (233 patients) found that and metaphyseal comminution have been sug-
Neer 1- and 2-part fractures had a 100% union gested as potentially increasing the risk of non-
rate, four studies (66 patients) found that Neer 3- union. Several studies have identified 2-part
and 4-part fractures had a 98% union rate, and fractures of the surgical neck as having a higher
four studies (351 patients) that included all risk of developing nonunions [13, 15, 16].
fracture types had a 96% union rate [11]. The Court-Brown and McQueen noted that translated
same group conducted a similar systematic surgical neck fractures (AO/OTA A3.2) had a
review of 12 studies of 514 patients with Neer 2-, nonunion rate of 3.8% that increased to 4.3%
3-, or 4-part fractures surgical stabilized with when adolescents and patients who had been lost
locked plating constructs and found that 16 to follow-up were excluded [10]. A subset anal-
patients failed to unite for a nonunion rate of ysis was performed based on the degree of
3.1% [12]. translation on initial lateral radiographs: the rate
of nonunion for fractures translated less than
33% was 2.6%, increasing to 10% for those
4.3 Risk Factors translated 33–66%, and 8.1% for those translated
66–100%. This suggests an association between
Risk factors for developing a nonunion include translational displacement and likelihood of
factors related to patient demographics and nonunion, but there does not seem to be a simple
fracture characteristics. Advancing age with its linear relationship. This is emphasized by the fact
corresponding loss of bone mineral density has that all the fractures that were translated 100%
been implicated as increasing the likelihood of achieved union, which is difficult to explain if
nonunion [13], but the largest prospective cohort increasing translation leads directly to higher risk
study that compared patients who developed of nonunions. It should be noted that a small
nonunions with those who did not, found no percentage of patients within these groups were
significant difference in the average ages between treated surgically but they represent a minority
the two groups [10]. Nutritional deficiencies and within each of the subsets described. Another
metabolic bone disease are recognized as con- subset analysis found that increasing fracture site
tributors to delayed unions and nonunions and angulation did not correlate with the risk of
should be screened for with appropriate labora- nonunion.
tory markers either at the time of initial fracture The effect of metaphyseal comminution on
treatment or at the earliest signs of delayed union was evaluated by combining the AO/OTA
fracture healing [13]. Smoking increases the odds subgroups A3.3 and B2.3. This group comprised
ratio of developing a nonunion to 5.5 times that only 52 of the 1027 fractures (5.1%), yet had a
4 Proximal Humerus Nonunions 97
nonunion rate of 7.7% compared to 0.7% for all blocked healing in 8 of 17 (47%) cases. The long
the other fractures [10]. It is unclear if greater head of the biceps was interposed in six non-
amounts of translation and comminution increase unions and the deltoid was the offending struc-
the risk of nonunion due to decreased cortical ture in the remaining two. Duralde et al. [17]
contact or if they serve as markers for disruption reviewed 20 patients who underwent surgical
of the medial soft tissues and damage to the intervention for proximal humerus nonunions
blood supply [8]. The authors of the largest and found soft tissue interposition in 8 of 12
prospective review of proximal humerus non- patients (67%) who had initially been treated
unions remained unable to define predictive cri- non-operatively. The contribution of soft tissue
teria for the development of nonunions because interposition as a risk factor for nonunion is
the incidence of nonunions is so low that an controversial. Court-Brown and McQueen felt it
adequate prospective study would require the was rare, especially for low energy fractures [10].
inclusion of an overwhelmingly and unrealisti- Another potential risk factor for the develop-
cally large number of patients [10]. ment of a nonunion is the presence of gleno-
The anatomy and biomechanics of the humeral joint arthrosis. Rooney and Cockshott
glenohumeral joint may contribute to the likeli- presented a case series of nonunion patients who
hood of developing a nonunion [8, 13]. Muscle had glenohumeral stiffness from rheumatoid
forces produce deformities, which may be larger arthritis or a surgical arthrodesis [20]. These
for more unstable fracture patterns. The insertion patients presumably had increased torsional and
of the subscapularis muscle can translate the bending forces transferred to the fracture site due
lesser tuberosity medially when it exists as a to the immobility of the adjacent joint, which
separate fragment, or may internally rotate the overcame the process of establishing bridging
head fragment depending on the remaining callus. This pathophysiology may be generalized
muscular forces acting on the proximal piece. to other restrictive phenomenon affecting gleno-
The rotator cuff muscles attached to the greater humeral motion such as osteoarthritis or adhesive
tuberosity will abduct and externally rotate the capsulitis, as well.
proximal fragment when the greater tuberosity Inadequate initial immobilization of the frac-
remains attached to the articular segment, or will ture may permit increased motion that exceeds
pull the greater tuberosity superiorly and poste- the ability to achieve union. This may result from
riorly when it is displaced. The humeral diaph- the lack of use of a sling, cuff and collar, or other
ysis is displaced anteriorly and medially by the forms of shoulder immobilization. Alternatively,
insertion of the pectoralis major and pulled cra- overly aggressive or premature rehabilitation has
nially by the deltoid muscle. The peri-articular been implicated as a potential cause of proximal
location of these fractures may contribute to the humerus nonunions. Unstable fractures treated
risk of nonunion if intra-capsular extension non-operatively must be given the opportunity to
allows synovial fluid to bathe the fracture frag- consolidate prior to beginning therapy in order to
ments and limit the formation and organization prevent excessive micro-motion or frank dis-
of hematoma that leads to callus formation [17, placement at the fracture site [13]. Inadequate
18]. immobilization may also occur after surgical
Interposed soft tissues between fracture frag- interventions that do not adequately stabilize the
ments may also be a predisposing factor for the osseous fragments or malunions reduce the
development of nonunions. Soft tissues struc- fracture into an alignment that risks further dis-
tures around the proximal humerus include: the placement postoperatively, such as varus angu-
tendon of the long head of the biceps, deltoid lation of the surgical neck.
muscle fibers, and the rotator cuff [13]. Nayak Distraction across a proximal humerus frac-
et al. [19] reviewed their experience with the ture has also been identified as a risk factor for
operative treatment of proximal humerus non- nonunion. Neer described the weight of the arm
unions and found that interposed structures itself as a sufficient distracting force to prevent
the healing processes needed to achieve union. inability to perform simple activities of daily
He reviewed a subset of nine patients with 3- and living due to pain at the nonunion site. The pain
4-part fractures that had failed to achieve union may be moderate at rest with the arm in a
after a trial of non-operative management. Six of dependent position but commonly increases with
the nine had been treated in a hanging arm cast attempted motion. Patients are often unable to lift
and two with overhead traction. He hypothesized even light loads. Depending on the chronicity of
that the distracting forces were the major cause of the nonunion, atrophy of the deltoid and
the patients’ nonunions [21]. para-scapular muscles may be noted. Evaluation
of axillary nerve function must be performed if
surgical intervention is being considered, as it
4.4 Classification will influence the success of certain procedures,
and may require electromyography (EMG) if the
Checchia et al. recognized the lack of a physical examination is equivocal. Integrity of
descriptive classification system for nonunions of the rotator cuff musculature should also be
the proximal humerus and published the fol- evaluated, but disuse and pain may make this
lowing four-group schema based on their review difficult and MRI may be needed for an adequate
of 21 cases. High 2-part nonunions describe evaluation.
nonunions of the anatomic neck with a small Patients with proximal humerus nonunions
articular segment and include 3-part fractures in may have soft tissue contractures that restrict
which the tuberosity segment consolidated with shoulder range of motion and the magnitude of
less than 5 mm of displacement. Low 2-part this limitation is related to the length of time
nonunions are differentiated by a primary non- from the initial injury. Court-Brown and
union of the surgical neck between the level of McQueen measured the shoulder range of motion
the lesser tuberosity and the insertion of the of patients following proximal humerus fractures
pectoralis major tendon and also include 3-part and compared the motion of patients who
fractures where the tuberosity united with less developed nonunions at 6, 13, 26, and 52 weeks
than 5 mm of displacement. Complex nonunions post injury to the motion of patients whose
include 3- and 4-part fractures or head-splitting fractures united. At all time points, flexion,
fractures in which the surgical neck nonunion is extension, abduction, internal and external rota-
accompanied by a tuberosity nonunion that is tion was significantly reduced in the individuals
displaced greater than 5 mm. Lost fragment with nonunions. At the final follow-up at
nonunions describe those with a large degree of 52 weeks, patients with union had regained 72%
bone loss associated with previous open injury or of their flexion, while those with nonunions
osteomyelitis [22]. This classification scheme, flexed to only 26% of their uninjured side.
though helpful in communicating nonunion Abduction showed an even more striking vari-
characteristics, has not widely been utilized in ance—58% of motion was regained with united
subsequent studies. Therefore, evaluation of fractures, but just 6% with a nonunion. Patients
treatment algorithms based on it has not been with nonunions recovered only 61% of extension
performed and prognostic implications of this (vs. 90% for patients with union), 62% of
classification are unknown. external rotation (vs. 79%), and 42% of internal
rotation (vs. 72%) at 52 weeks. Most signifi-
cantly, they found that as early as six weeks after
4.5 Patient Evaluation injury, patients who were developing nonunions
had less motion, and that any functional gains
4.5.1 Clinical Examination after 13 weeks post injury were minimal. Motion
increased in a relatively linear pattern for patients
Patients with proximal humerus nonunions fre- who achieved union, but patients with nonunions
quently present with a chief complaint of had not only less mobility, but lost motion in all
directions except external rotation after week 26. nonunion should be high among patients who are
This was attributed to contractures of the soft slow to regain function after a proximal humerus
tissues surrounding the shoulder combined with fracture.
transfer of motion to the fracture site [10]. Studies that used scales to quantify the pain of
Based on these findings, patients who are nonunions of proximal humerus fractures are
developing nonunions will have smaller arcs of rare. Several small case series of patients who
motion than expected as early as 6 weeks post underwent surgical intervention for symptomatic
injury. Established nonunions will likely begin to nonunions used different pain assessments tools.
lose mobility after 26 weeks, as opposed to Antuña et al. evaluated pain scores preopera-
patients with healed fractures, who will continue tively in 25 patients using a 5-point scale
to improve their ranges. Patients with proximal described by Cofield [25], where a score of four
humerus fractures who are failing to achieve indicated moderate pain and five meant severe
expected mobility gains should be worked up for pain. Their average score of 4.6 reflects the
nonunions as early as 13 weeks after injury. debilitating pain experienced by these individuals
Other published case series have reviewed [15]. Duralde et al. [17] screened 20 patients who
patients who underwent surgical interventions for went on to pursue surgery for their nonunions
their symptomatic proximal humerus nonunions and found that 16 (80%) classified their pain as
and measured the preoperative range of motion, totally disabling, requiring medications and
finding shoulder flexion consistently averaged interrupting sleep. These series illustrate the
between 35° and 46° [16, 19, 23, 24], shoulder severe, disabling pain that proximal humerus
abduction averaged between 35° and 41° [15, nonunions can cause, but only include patients
24], external rotation averaged 15°–26° [15, 16, that chose intervention for their symptomatic
23, 24], and internal rotation averaged 25° [24] nonunions and therefore have a selection bias.
or the ability to reach the sacrum with the Several authors note that there are patients with
involved extremity [15, 19]. These studies have a proximal humerus nonunions that are less
selection bias since patients chose surgical symptomatic and may not require further treat-
intervention for their disabling nonunions and ment [8, 26].
may have been more restricted than the average
non-united shoulder range of motion.
Court-Brown and McQueen did not publish 4.5.2 Radiographic Imaging
dedicated pain scores but used the Neer and
Constant scores that contain pain assessment. Radiographic evaluation of a proximal humerus
They used the time it took to resume specific nonunion begins with a standard shoulder series:
activities of daily living as a surrogate for a true AP in the scapular plane (Grashey view),
recovery from the pain and disability caused by scapular Y, and axillary views. Internal and
proximal humerus fractures. Patients with non- external rotational AP views may be of use but
unions took slightly longer on average than those have largely been supplanted by computed
who achieved union to regain the ability to per- tomography (CT). Historically, in cases when
form personal hygiene (5.4 vs. 5.1 weeks), standard roentgenographs were equivocal, stress
approximately one week longer to be able to radiographs were used to detect instability [27],
dress independently (5.7 vs. 4.6 weeks), and but CT has also replaced these. Comparison
greater than twice as long to resume the ability to views of the contralateral shoulder may be of
perform housework and shop (16.8 weeks vs. benefit for evaluating loss of length or associated
7.5 weeks and 17.4 weeks vs. 8.2 weeks, malunions involving the tuberosities.
respectively) [10]. Therefore, the degree of dis- Nonunions may be classified as hypertrophic
ability of proximal humerus nonunions, as well or atrophic. Hypertrophic nonunions have abun-
as the length of time it takes for that disability to dant callus that failed to bridge the fracture site,
resolve should be assessed, and suspicion for a and have been described as having the appearance
of an “elephant’s foot”. These nonunions possess Nuclear imaging modalities may offer additional
preserved vascularity along with the biologic information in the evaluation of proximal humerus
constituents to permit bone healing. Atrophic nonunions by evaluating callus vascularity and
nonunions appear osteopenic at the fracture mar- metabolic activity, identifying the presence of
gins with minimal callus formation. They are synovial pseudarthroses, or supporting the diag-
biologically inert with vascular and biologic nosis of acute or chronic infection at nonunion sites.
compromise and will require biologic augmenta- Brinker and O’Connor [28] outlined the use of
tion. Lytic or mixed lytic and blastic lesions can various nuclear imaging studies to assess non-
be signs of underlying pathologic or metastatic unions. Technetium-99m-pyrophosphate studies
processes. Signs of a sequestrum or involucrum will have increased uptake at the site of viable
are pathognomonic for infection. Increased den- nonunions as the complexes are incorporated into
sity and sclerosis throughout the humerus may hydroxyapatite crystals, a process which requires
indicate Paget’s disease or osteopetrosis. and indicates active osteoblastic activity and ade-
Radiographs should be examined for other quate vascularity. When a synovial pseudoarthro-
characteristics in addition to the nonunion. Evi- sis, or a fluid-filled pseudocapsule forming a mobile
dence of avascular necrosis of the humeral head false joint is present at a nonunion site, a
with subchondral sclerosis or humeral head col- technetium-99m-pyrophosphate scan will show a
lapse should be evaluated. The glenohumeral cold cleft between metabolically active, hot ends of
joint should also be inspected for signs of the nonunion. This corresponds with a lack of
degeneration including: joint space narrowing, vascular ingrowth and osteogenesis within the
subchondral sclerosis, osteophytes, and bone loss pseudocapsule. Radiolabeled white blood cell
involving either the humeral head or the glenoid. scans (such as indium-111 or technetium-99m-
The presence of avascular necrosis or advanced hexamethylpropylene amine oxime) can help sup-
osteoarthritis may shift treatment of a nonunion port the diagnosis of an acute infection since
towards an arthroplasty. increased activity at the nonunion site occurs when
Advanced imaging modalities may provide labeled polymorphomuclear leukocytes accumu-
additional information to assist in diagnosis and late. Alternatively, gallium scans use gallium-67
surgical planning. Computed tomography offers citrate (often with a technetium-99m sulfa colloid)
several benefits. The multiplanar imaging pro- that localizes to the site of a chronically infected
vided by computed tomography with 2- and nonunion.
3-dimensional reconstructions allows evaluation
of tuberosity malunions, head cavitation,
intra-articular extension, and glenohumeral 4.5.3 Laboratory Tests
arthritic changes. Although CT scanning offers
high sensitivity in evaluating for the presence of Laboratory analysis in patients with proximal
a nonunion, the specificity has been found to be humerus nonunions can help determine the cause
somewhat lower due to the high number of false of failure to unite and identify factors that should
positives. Magnetic resonance imaging can be corrected to allow healing. If the possibility of
demonstrate the presence of osteonecrosis and an infection at the nonunion site exists, preoper-
quantify the area involved. MRI may also ative laboratory evaluation should include an
address questions regarding the continuity and erythrocyte sedimentation rate and C-reactive
health of the rotator cuff that would influence protein level. These are non-specific markers of
arthroplasty options, though some centers are systemic inflammatory response and can be ele-
moving towards ultrasound for diagnosing rota- vated by other processes. A white blood cell count
tor cuff pathology. The latter remains highly may show leukocytosis, which further raises the
dependent on the capabilities of experienced suspicion of infection and the differential may
technologists, which may not be available in all demonstrate a left shift, with increased percent-
facilities. ages of polymorphonuclear cells. Ultimately, the
gold standard for diagnosing infection is cultures [10, 14], and Norris et al. consider fractures
taken from the nonunion site at the time of non-united if not clinically healed by 3 months
surgery. [18, 31]. Court-Brown and McQueen noted that
Other laboratory values may provide evidence the range of motion, Neer, and Constant scores
of nutritional or metabolic derangements that improved for all patients over 6 months follow-
may be preventing or slowing union. Nutritional ing injury but, in patients who developed non-
deficiencies may be assessed with an albumin unions, all of these were less favorable as early as
level or total lymphocyte count. Values 3.0 6 weeks. Nonunion patients improved from
g/dL and >1500 cells/mm3 are preferred to allow 6 weeks to 3 months, but showed only minimal
appropriate healing [28]. Other endocrine mark- gains between 3 and 6 months, which suggests
ers have been found to be associated with non- that nonunion patients can be identified early.
unions that were felt to not be the result of either After 6 months, nonunion patients had a decline
fracture-related or technical factors. Brinker et al. in motion of their glenohumeral joint and the
identified 37 patients with nonunions that were Neer and Constant scores [10]. This was attrib-
not related to infection or technical factors and uted to the development of soft tissue contrac-
referred those patients to an endocrinologist. tures at the shoulder after 6 months.
68% were found to be vitamin D deficient, 35% Based on these findings, proximal humerus
had abnormal calcium excretion, and 24% had nonunions may be identifiable as early as
functional thyroid abnormalities. Less frequent 3 months after injury, with a lack of callus for-
endocrinopathies were found by measuring mation on imaging, poorer glenohumeral range
levels of reproductive hormones, alkaline phos- of motion on physical exam than expected, and
phatase, parathyroid hormone, prolactin, and lower shoulder outcome scores than expected.
growth hormone. All 31 patients diagnosed with Once diagnosed, a proximal humerus nonunion
metabolic or endocrine abnormalities underwent should be addressed prior to 6 months after
treatment of the abnormality and 30 achieved injury, before soft tissue contractures at the
union. Four patients united without any surgical glenohumeral joint develop and present a barrier
intervention [29]. This study was not limited to to regaining optimal function.
patients with nonunions of the proximal humerus Beredjiklian et al. reviewed the results of 39
but demonstrates that metabolic and endocrine patients who underwent late surgical intervention
abnormalities may be associated with nonunions. for proximal humerus malunions. Intra-operatively
Preoperative assessment of a patient’s thyroid they found that 31 patients (79%) had concomitant
function, vitamin D and calcium levels may soft tissue pathology consisting of 25 capsular
detect modifiable factors that contributed to the contractures, 15 torn rotator cuffs, and two cases of
development of the nonunion. If these tests are subacromial impingement. These were treated via a
normal but the suspicion for metabolic abnor- circumferential capsular release, subscapularis
malities remains high due to lack of other obvi- tendon lengthening, rotator cuff repair, and
ous cause for nonunion, a referral to an acromioplasty. The authors noted a significant
endocrinologist may be advisable. difference in outcomes among patients who
underwent surgical intervention within one year of
initial injury (84% satisfactory results) versus those
4.6 Surgical Timing who were treated after one year (55% satisfactory
results) [32]. This was attributed to more soft tissue
The time course of healing of a proximal scarring and disuse atrophy in those patients for
humerus fracture is variable and can be affected whom operative intervention was delayed and
by fracture characteristics, systemic diseases, and emphasizes the benefit of identifying nonunions
metabolic abnormalities. Bridging callus is typi- and intervening early—prior to the development of
cally seen on radiographs at a mean of 6 weeks contractures. This study also stresses the need to
[30]. The median time to union is 13–14 weeks address soft tissue contractures during the surgical
procedure after they have developed to optimize the extensive and the postoperative course requires
outcome. compliance and assistance from family and
Studies that analyzed the results from small friends. For some patients, medical
series of patients with proximal humerus non- co-morbidities may make the risks of surgical
unions treated many months after their initial intervention unacceptably high. Alternatively,
injury emphasize the need to address the bony patients may be unable or unwilling to comply
nonunion as well as the soft tissues contractures with postoperative protocols. Some authors
during surgery in order to maximize functional consider a nonfunctional deltoid muscle to be a
improvement. Preoperative assessments demon- contraindication for operative treatment [22]. In
strated significant reductions in glenohumeral such cases, efforts to increase the patient’s heal-
motion in these patients associated with their ing potential by addressing nutritional and
prolonged disuse of the shoulder and highlight metabolic disorders should be undertaken and the
the need to address these contractures during focus of treatment should be on patient comfort
surgery to prevent a united fracture with a stiff (Fig. 4.1a–d).
shoulder joint [13]. The surgical indications for nonunions of the
proximal humerus include debilitating pain and
functional deficits. Some smaller studies found
4.7 Treatment Options that up to 50% of patients with proximal humerus
nonunions are minimally symptomatic and quite
4.7.1 Non-operative functional [19, 33]. A case report study that
reviewed the complications of locking plates
The treatment of proximal humerus nonunions used for the treatment of proximal humerus
must overcome numerous challenges including fractures found that although the Constant scores
the biologic insults from the initial injury and any were significantly lower for patients who devel-
previous surgeries, bone loss, humeral head oped nonunions (45 vs. 68), only two of the four
cavitation, osteopenia, soft tissue contractures, patients with a nonunion opted for revision
and infection. Patients are commonly elderly surgeries [34]. A thorough conversation should
with medical comorbidities. Surgery may be be undertaken with patients in order to assess
Fig. 4.1 An 82-year-old female two years status-post debridement, and placement of antibiotic cement spacer,
open reduction and internal fixation with standard T-plate seen one month postoperatively on AP (c) and lateral
with wound drainage, hardware failure, and surgical neck (d) views. Patient is now six months post-surgery,
nonunion on anteroposterior (AP) (a) and lateral comfortable and able to perform activities of daily living
(b) views. Underwent stage one of a planned two-stage with motion up to shoulder-level. She has declined her
procedure with removal of hardware, irrigation, second stage in favor of no further surgery
their pain and functional deficits and educate A review of proximal humerus fractures treated
them about the risks of revision surgery so the non-operatively found a 2% rate of AVN (12
patient may make an informed decision. studies and 650 patients) [11]. Similar reviews of
fractures treated using locked plating constructs
found rates of AVN ranging from 7.9 to 10.8%
4.7.2 Surgical Treatment [12, 35], although only four of the 51 patients who
developed AVN chose to undergo further surgical
Many techniques have been described for the treatment [35]. Patients who develop AVN after
repair of proximal humeral surgical neck nontreatment of acute proximal humerus fractures
unions. Plating techniques using conventional have lower Constant scores (average 46) [36] and
T-plates, fixed angle devices such as blade plates a decreased likelihood of achieving good or
or anatomically designed locking plates have excellent results when compared to those who do
been described. Intramedullary implants have not develop AVN [37].
also been utilized, ranging from older implants Gerber et al. performed a subgroup analysis of
such as Rush rods or Enders nails in conjunction patients with AVN. They found even with AVN and
with tension banding, to modern rigid inter- collapse, Constant scores of patients with an
locked intramedullary nails. In the setting of anatomical reduction were similar to patients who
bone loss or atrophic nonunions augmentation had undergone a primary hemiarthroplasty for a
with allograft or autograft bone is necessary and proximal humerus fracture. This suggests that
structural reinforcement with fibular strut grafts patients treated with osteosynthesis after an appro-
has been found to be biomechanically advanta- priate reduction who develop AVN may function as
geous. A successful nonunion repair requires well as patients who undergo shoulder hemiarthro-
adequate bone quality and a proximal fragment plasty [36]. Although the development of AVN
that is large enough to obtain purchase in. adversely affects Constant scores, the functional
Although fixed angle devices and locking screw limitations that result do not drive patients to
technology improve biomechanical stability in undergo further surgical intervention at high rates,
the setting of osteoporotic bone, the size and and they function at levels similar to patients whose
quality of the fragments should be carefully fractures were treated using a hemiarthroplasty
assessed before osteosynthesis is performed. acutely. These studies are of patients with acute
Significant medial calcar comminution has also proximal humerus fractures, rather than nonunions.
been described as a harbinger of difficulty Unfortunately, the current nonunion literature does
establishing a stable configuration that must be not allow calculation of the rate of AVN or esti-
addressed when undertaking fixation. Several mation of its clinical significance.
techniques have been described, including
impaling the head onto the shaft, reconstructing
the calcar, using calcar screws to substitute for 4.7.4 Osteosynthesis
the medial cortex, or using an intramedullary
fibular strut graft [8].
4.7.4.1 ORIF with Standard Plates

4.7.3 Avascular Necrosis Several studies reported good results when
osteosynthesis was used for the treatment of sur-
Humeral head viability is a consideration when gical neck nonunions of the proximal humerus.
deciding on the operative treatment of acute frac- Healy et al. retrospectively reviewed their expe-
tures since osteonecrosis and the resultant humeral rience and found better functional results after
head collapse may lead to poor results. However, open reduction and internal fixation than after
the incidence of avascular necrosis and its func- hemiarthroplasty, unreamed intramedullary
tional implications are not fully understood. implants, or non-operative management. Fixation
techniques varied- the majority of their patients 4.5 mm DC plate. Patients treated with blade
were treated with a 4.5 mm T-plate and a tension plates in this study had an increase in Constant
band through the rotator cuff, but others were scores from 30.5 to 85.5 and had improvements in
stabilized using semitubular plates, dynamic active shoulder flexion and abduction. Compli-
compression plates, or Cobra plates. 12 of 14 cations included two patients (4.4%) with surgical
(86%) nonunions achieved union at an average of site infections and two patients (4.4%) who
4.8 months postoperatively. Both nonunions that developed AVN [41]. Galatz and Iannotti treated
failed to unite were performed on the same patient 13 patients with nonunions of the proximal
who had significant medical comorbidities that humerus: ten with blade plates and three with
may have contributed to her lack of healing. T-plates, combined with auto- or allograft. Eleven
Autogenous bone grafting was performed in 12 of patients achieved excellent results and only one
14 cases. 11 of 12 grafted cases and 1 of 2 pro- patient failed to achieve union. The persistent
cedures performed without grafting achieved nonunion occurred in a wheelchair-dependent
union. There were nine good results, four fair patient who returned to functional use of the
results, and one poor result. Shoulder range of involved upper extremity early and ultimately
motion averaged 110° of elevation, 33° of external broke her T-plate but went on to successful union
rotation, and internal rotation to the thoracolumbar and excellent results following a revision surgery.
junction postoperatively [38]. Pain scores improved from 4.2 to 1.2 on a 5-point
scale and the average shoulder flexion improved
4.7.4.2 ORIF with Blade Plates from 23.8° prior to surgery to 143.8°. All patients
Osteosynthesis using a blade plate has been shown in this study achieved overhead elevation and
to be successful for achieving union for nonunions were able to perform activities of daily living
of the surgical neck (Fig. 4.2a–i). Ring et al. independently [13].
reviewed 25 patients who underwent blade plating
with autogenous iliac crest bone grafting, 23 of
which (92%) united their fractures. Eighty percent 4.7.5 Use of Augmentation
of the patients obtained good or excellent func-
tional results [39]. Allende and Allende presented
their results of 7 patients with atrophic proximal
humerus nonunions who underwent surgical 4.7.5.1 Structural Graft Augmentation
treatment using a locking blade plate and all seven The use of an intramedullary peg graft was first
cases achieved union. It took an average of described by Walch et al. who treated 20 patients
5.9 months for patients to have radiographic evi- with proximal humerus nonunions using
dence of union and patients who had allograft cortico-cancellous auto-graft harvested from the
bone grafting took longer to achieve union com- iliac crest, tibial crest, or fibula, in addition to
pared with those who received autogenous bone osteosynthesis using a T-plate. Nineteen of 20
graft (average of 7 vs. 5 months). Postoperative patients went on to union at an average of
scores using the Disabilities of the Arm, Shoulder, 4 months and demonstrated significant pain relief
and Hand (DASH) averaged 25 points and Con- and improvement in shoulder flexion from 60° to
stant scores averaged 72.7 [40]. Tauber et al. used 131°. The patients had six excellent and six good
a blade plate to treat 45 of 55 patients with prox- results, but 3 patients sustained tibial fractures
imal humerus nonunions who had sufficient bone following graft harvest and authors recom-
stock in the humeral head. A blade plate was mended against future use of this graft site [42].
fashioned by bending a one third tubular plate and The technique of using a fibular strut allograft
patients were not bone grafted. 41 of 45 patients as an intramedullary implant provides some
(91%) achieved union. The four failures were inherent stability when the graft is impacted into
attributed to varus collapse with screw pullout and the humeral head, and improves proximal screw
underwent revision with a blade plate made from a purchase as described by Badman and Mighell,
Fig. 4.2 A 37-year-old male sustained 3-part proximal (e) views. Patient underwent hardware removal and
humerus fracture (surgical neck and lesser tuberosity) revision fixation using a blade plate, fibular strut allograft
during an motor vehicle accident. a Underwent open (outlined), and bone morphogenic protein, seen immedi-
reduction and internal fixation with locked plating and ately post-operatively in AP with internal rotation (f) and
allograft, seen on anteroposterior (AP) (b) and axillary external rotation (g) views. Union was achieved at seven
(c) views. Went on to surgical neck nonunion with varies months on AP (h) and lateral (i) views
collapse and screw back out on AP (d) and axillary
but without the graft donor site morbidity noted assessment of shoulder motion showed average
above. Badman and Mighell published the results flexion to 115°, external rotation to 37°, and
of 18 patients who underwent intramedullary internal rotation to the 10th thoracic vertebra.
fibular strut allografting and stabilization of ASES scores improved from 40 preoperatively to
proximal humerus nonunions using locked plat- 81 and analog pain scores improved from 6.7 to
ing. 17 patients (94%) achieved union at an 1.5. Complications involved two posterior cord
average of 5.4 months. The remaining patient brachial plexus palsies that improved within
was a heavy smoker who had failed two previous 3 months and two cases of adhesive capsulitis,
attempts at surgical stabilization and ultimately both requiring arthroscopic capsular release
required a hemiarthroplasty. Postoperative [26].
4.7.5.2 Biologic Augmentation and 10 with hemiarthroplasty based on an intra-

Augmentation with the recombinant human bone operative assessment of the suitability of the
morphogenic protein rhBMP-2 can be used for fracture fragments for fixation. The fixation
nonunions when biological activity is felt to be construct utilized Enders rods with a nonab-
lacking. The current literature supports the use of sorbable suture or wire tension band and iliac
BMP for acute open tibia fractures [43], tibial crest bone grafting of the nonunion site with an
nonunions [44], and recalcitrant long bone non- onlay of cortical bone graft surrounding the
unions [45], but no studies have reported on its use nonunion site. Five of the 10 patients achieved
for nonunions of the proximal humerus. union at an average of seven months. Of the
A Cochrane review of BMP use for fracture remaining five, two were converted to a hemi-
healing in adults concluded that there is a paucity arthroplasty, one underwent revision ORIF with
of data currently available and its role in treating a free fibular graft, one developed a deep infec-
nonunions remains unclear. Furthermore, they tion and underwent removal of hardware and
highlighted the high risk of bias in these studies humeral head resection, and the final patient
due to industry involvement [46]. The U.S. Food refused further surgical intervention. Two of the
and Drug Administration has granted rhBMP-7 five that achieved union had an excellent result
(OP-1) a humanitarian device exemption (HDE) and three had a satisfactory result. Three later
for treating recalcitrant long bone nonunions when underwent reoperation for removal of painful
autograft is unfeasible and alternative treatments prominent hardware. All five that failed to
have failed [47], but rhBMP-7 cannot be used “off achieve union had unsatisfactory results [17].
label.” The FDA has approved rhBMP-2 for use in Nayak et al. reviewed 17 patients who
acute open tibia fractures, but its use for proximal underwent operative intervention for nonunions
humerus nonunions must be recognized as “off of the surgical neck: 10 were treated with open
label.” Therefore, the use of biologic augments reduction and internal fixation using Rush rods,
such as rhBMP-2 should be carefully weighed in tension banding, and bone graft, and seven had a
light of the additional cost and unsubstantiated hemiarthroplasty. Two of the 10 patients who
efficacy for treating proximal humerus nonunions had Rush rods had persistent lucencies on
compared to local or iliac crest autograft [8]. imaging indicating a failure to unite. The Rush
rod group had a better average postoperative
range of motion than the hemiarthroplasty group
4.7.6 Unreamed Intramedullary Rods with more elevation (140° vs. 110°), but com-
plications were common in this group. Intraop-
The use of unreamed intramedullary implants erative complications included one case of
(Rush rods) for the treatment of proximal circumflex artery laceration, one permanent
humerus nonunions led to less favorable results. axillary nerve injury, and one cortical perforation
Five patients in a series presented by Healy et al. with a Rush rod. Postoperatively, two patients
were treated using intramedullary implants: two had radiographic evidence of avascular necrosis
with Rush rods alone, one using Rush rods with a of the humeral head and all had symptoms con-
tension band, one with an Ender nail, and one sistent with impingement of the Rush rods. Eight
with a Lottes nail. Only one patient united their patients had hardware removal after achieving
nonunion and all had poor functional results. The union [19]. No study has been able to reproduce
average shoulder motion was 40° of flexion and the union rate of 92% obtained by Neer, who
10° of external rotation [38]. Duralde et al. pre- treated 13 patients using unlocked intramedullary
sented a retrospective review of 20 patients with implants with rotator cuff tension banding. All
nonunions of the surgical neck, including 10 patients required a second surgery to remove
treated with open reduction and internal fixation their prominent, symptomatic hardware [21].
4.7.7 Interlocking Intramedullary surgical stabilization using an interlocking

Nails intramedullary nail with bone grafting. The
intraoperative technique involved locking the
The availability of more modern interlocked nail with two proximal and two distal screws and
intramedullary nails has increased the union rates emphasized seating the nail below the subchon-
and decreased the rate of symptomatic hardware dral surface of the humeral head to avoid
when intramedullary implants are used for the impingement. Patients with large bone voids
treatment of nonunions (Fig. 4.3a–h). Yamane underwent cancellous iliac crest autografting to
et al. reviewed 13 patients who underwent fill the void, while smaller defects were filled
Fig. 4.3 A 60-year-old female sustained a surgical neck and immobility with varus collapse and hardware pullout
fracture after a ground-level fall, seen on anteroposterior on AP (e) and axillary (f) views. Patient underwent
(AP) (a) and coronal computed tomography (b). Under- hardware removal and revision fixation, using an inter-
went open reduction and internal fixation with locked locked intramedullary nail, allograft, and bone mor-
plating and allograft on AP (c) and lateral (d) views. phogenic protein, with union at three months on AP
Presented two years postoperatively with continued pain (g) and axillary (h) views
with tricalcium phosphate cement. All patients shortening between the articular surface and the
achieved union without evidence of malunion or deltoid insertion. Varus alignment is often a
avascular necrosis. Japanese Orthopedic Associ- progressiDeepika Uve deformity and should be
ation shoulder scores averaged 85 points post- carefully avoided. Newer interlocking intrame-
operatively, with four excellent, seven good, and dullary nails increase the stability at the non-
two fair results. Postoperative shoulder range of union site compared to Rush and Ender’s rods
motion demonstrated flexion to 122° and external and may be inserted with decreased soft tissue
rotation to 35°. The only complication was the stripping compared to plating constructs. These
backing out of proximal interlocking screws, advances have served to increase the number of
which required removal in 2 patients. It should proximal humerus nonunions that can be treated
be noted that 11 of 13 patients treated in this with internal fixation rather than an arthroplasty.
study had not previously undergone surgical
treatment and the two who had been treated
operatively underwent percutaneous pinning or 4.8 Arthroplasty
intramedullary nailing, so it is unclear if this
protocol would achieve similar results in patients Proximal humeral nonunions with bone stock
who had previously undergone fixation with that is insufficient to achieve adequate internal
plating constructs [16]. fixation or in which the humeral head is avas-
cular are felt to be more amenable to arthroplasty
than surgical stabilization. There are fracture
4.7.8 Summary of Internal Fixation characteristics that dictate which arthroplasty
Devices option is most suitable. A hemiarthroplasty
requires tuberosity integrity and a functional
Technological advances in implant design offer rotator cuff to achieve optimal results. When
the potential for improved union rates and significant glenoid arthritis is present patients
decreased postoperative hardware prominence may be better treated with a total shoulder
for nonunion patients treated with osteosynthesis. arthroplasty. Patients who need an arthroplasty
The need for adequate bone stock and a viable but do not have functional or repairable rotator
humeral head without significant glenohumeral cuffs or tuberosities may be candidates for a
arthritis has not been obviated by the availability reverse total shoulder arthroplasty, but a com-
of modern implants. Regardless of the implant petent deltoid and axillary nerve is a requirement.
used, the preparation of the nonunion site with
resection of fibrous tissue and avascular bone is
critical. Bone loss at the nonunion site and 4.8.1 Unconstrained Arthroplasty
humeral head cavitation are commonly encoun-
tered challenges and autograft bone grafting is Several older series reviewed the results of
widely utilized to address these issues, although patients who underwent treatment for proximal
some authors have achieved union of proximal humerus nonunions using a variety of modalities,
humerus nonunions using allograft, tricalcium including hemiarthroplasty. These studies inclu-
phosphate cements, or without augmentation. ded patients who underwent an assortment of
Locking plate technology and intramedullary fixation strategies in addition to hemiarthroplasty
fibular strut allografts have lessened, but cer- and were not randomized. Healy et al. presented
tainly not eliminated the difficulties associated a series of 25 patients that included six patients
with the osteopenia commonly encountered in who underwent a hemiarthroplasty with rotator
surgical neck nonunions. A careful balance cuff repair if necessary due to either inadequate
should be struck between the improved stability bone stock for internal fixation or significant
achieved through compression at the fracture site articular involvement. These patients had good
and the deltoid weakness associated with over pain relief but postoperative shoulder motion
only averaged 72° of flexion and 30° of external motion that is regained is unpredictable and less
rotation (compared to averages of 110° of flexion than that regained following successful
and 33° of external rotation among patients osteosynthesis.
undergoing fixation with plating constructs). One study reviewed the results of uncon-
Despite these limitations in motion, 50% of these strained shoulder arthroplasties (either hemi-
patients were rated as having good functional arthroplasty or total shoulder arthroplasty) for the
results [38]. sequelae of previous proximal humerus fractures.
Duralde et al. presented a series that included Surgical neck nonunions represented a small
10 patients who underwent hemiarthroplasty cohort (six patients) of the 71 patients in this
because of head cavitation, severe osteoporosis, study who presented with AVN and head col-
or glenohumeral arthritis. Comparison of these lapse, chronic dislocations, or tuberosity malu-
patients with 10 patients who underwent ORIF of nions. All the nonunions were treated with
their proximal humerus nonunions showed no hemiarthroplasties used greater tuberosity osteo-
difference in motion, function, or pain relief. tomies. These patients had unsatisfactory results
Overall functional results showed three excellent, with Constant score pain ratings of 7 (out of 15)
three satisfactory, and four unsatisfactory results, and disappointing improvements in shoulder
but among the patients who achieved tuberosity flexion from 50° preoperatively to 63° postop-
union, only one patient had an unsatisfactory eratively. Combining these patients with others
outcome. Hemiarthroplasty patients had a 70% involved in the study, the most significant pre-
rate of tuberosity union and two patients needed dictor of a poor outcome was the need for a
further surgery to address tuberosity nonunions. greater tuberosity osteotomy, which was ubiqui-
Other complications within the hemiarthroplasty tous among nonunions, and caused the authors to
group included two dislocations and one patient recommend osteosynthesis rather than arthro-
with asymmetric glenoid wear due to a humeral plasty in treating these injuries [48]. The same
component that was cemented too proudly [17]. authors later published a larger review including
Nayak et al. presented seven patients treated 22 surgical neck nonunions showing similarly
with hemiarthroplasties for nonunions of the disappointing results with Constant scores
surgical neck and compared them to 10 patients improving only from 21 to 36. They reiterated
undergoing ORIF using Rush rods, tension that they considered a surgical neck nonunion a
banding, and bone grafting. Patients treated with relative contraindication for an unconstrained
hemiarthroplasties had one grade more effective prosthesis. They cited the need for a tuberosity
pain relief but their shoulder flexion averaged osteotomy, the difficulties achieving tuberosity
110°, compared to 140° for the internal fixation union, and the poor functional results that fol-
patients. Functional improvement and UCLA lowed as reasons to avoid arthroplasty. Alterna-
rating scores were similar between the two tively, they recommended intramedullary peg
groups. All patients were able to perform activ- grafting and osteosynthesis if sufficient humeral
ities of daily living independently but no patient head bone stock existed or utilizing a low-profile
in either group returned to their pre-injury level prosthesis with large amounts of autografting to
of function. Complications included one axillary improve the likelihood of tuberosity union if
nerve palsy, two asymptomatic inferior sublux- osteosynthesis was not possible [49].
ations, and two cases of impingement syndrome Two more recent studies reviewed the use of
that improved with corticosteroid injections. unconstrained shoulder arthroplasty for the treat-
Tuberosity union was not quantified but the ment of nonunions of the proximal humerus and
authors attributed several patients with the have focused on identifying the risk factors for
inability to elevate beyond 90° to suspected unsatisfactory outcomes. Antuña et al. reviewed
tuberosity nonunion [17]. These studies demon- 25 patients treated with 21 hemiarthroplasties and
strate the effectiveness of a hemiarthroplasty for four total shoulder arthroplasties for nonunions of
pain relief but the amount of shoulder range of the proximal humerus. An unconstrained
arthroplasty was selected for each of these patients identified between hemiarthroplasties and TSA
due to: failure of previous fixation constructs or press-fit and cemented humeral prostheses.
leaving insufficient bone stock, head cavitation, Tuberosity healing was reviewed: 17 (25%)
significant osteoporosis, or advanced gleno- healed anatomically, 18 (27%) malunited with
humeral arthritis. Pain consistently improved for 5 mm displacement, 18 (27%) developed a
all patients treated with a hemiarthroplasty or a nonunion, and 14 (21%) were resorbed or
TSA. Range of motion averaged 88° of abduction, resected. Active elevation was significantly
38° of external rotation, and internal rotation to L3 decreased for patients with tuberosity nonunions
post operatively. Twenty of 25 patients consid- but the same patients did not have more pain or
ered themselves either much better or better than worse Neer functional scores. Bone grafting did
preoperatively. Neer functional results were: one not prevent tuberosity nonunions. Other compli-
excellent, 11 satisfactory, and 13 unsatisfactory cations included 11 severe subluxations or dis-
results, for an unsatisfactory rate greater than locations, two deep infections, and one late
50%. No subgroup analysis was performed to periprosthetic fracture [24].
compare hemiarthroplasty to total shoulder
arthroplasty patients, but, two of the patients with
unsatisfactory results who felt they were worse 4.8.2 Reverse Total Shoulder
after surgery were found to have advanced gle- Arthroplasty
noid arthritis and yet were treated with hemi-
arthroplasties. This points out the importance of The pain relieving benefit of unconstrained
glenoid inspection when choosing hemiarthro- arthroplasty when treating proximal humerus
plasty over TSA. The authors found that issues nonunions has been tempered by the unpre-
with tuberosity healing were the most common dictable results of glenohumeral motion. The
complications: nine patients had tuberosity strong association between postoperative range
resorption, two had nonunions, and one had a of motion and tuberosity healing has lead some
malunion, for an overall tuberosity complication to suggest reverse total shoulder arthroplasty as a
rate of 48%. The authors emphasized using heavy viable alternative to hemiarthroplasty and stan-
nonabsorbable sutures, bone graft to fill gaps dard total shoulder arthroplasty, due to its
between the tuberosities and the shaft, and decreased reliance on greater tuberosity healing.
restricting shoulder range of motion post opera- Martinez et al. reviewed 18 patients who under-
tively in an effort to minimize the risk of this went reverse total shoulder arthroplasty for the
complication. Additional complications included treatment of proximal humerus nonunions. Can-
one periprosthetic fracture and one dislocation, cellous autograft from the resected humeral head
both of which required surgical intervention. or, in cases of significant bone loss, humeral
There was also a significant correlation between cortical allograft was used to augment resorbed
increasing number of fracture parts using the Neer bone. Constant scores and subjective shoulder
classification and less pain relief, as well as less scores increased postoperatively to 55 and 50%
subjective satisfaction with the procedure [15]. of the contralateral uninvolved side, respectively.
Duquin and colleagues reviewed the Mayo Range of motion averaged 90° of flexion, 85° of
Clinic experience treating 67 proximal humerus abduction, 30° of external rotation, and 55° of
nonunions with unconstrained arthroplasties. internal rotation. Patient satisfaction was reported
Their results were similar to those published by with eight very satisfied, six satisfied, and four
Antuña – pain was reliably decreased but motion unsatisfied patients. Complications included one
was less predictable, with average elevation of transient axillary nerve palsy, two deep infec-
104° and external rotation of 50°. Neer functions, and two dislocations. The authors stressed
tional ratings showed 11 excellent and 21 satis- that in the setting of a proximal humerus non-
factory results, but greater than 50% union in an elderly patient, the rotator cuff is
unsatisfactory results. There were no differences often functionally insufficient due to the
combination of trauma and chronic disuse, increased fracture translation and metaphyseal
making a reverse TSA an excellent option comminution, but the relationship between these
because it relies on the deltoid rather than a parameters and the rate of nonunion is not sim-
competent rotator cuff to achieve elevation and ple. Patient-related risk factors include nutritional
abduction. Conversely, a functioning deltoid is a or metabolic deficiencies, smoking, and medical
prerequisite for reverse TSA and in the same comorbidities.
patient population disuse can make assessment Patients developing nonunions can generally
difficult, so the authors recommend electromyo- be identified as early as three months after injury
graphy if concerns related to deltoid function based on their restricted range of motion, lower
exist [23]. Constant scores, and greater difficulty with
returning to performing activities of daily living,
when compared to their healing counterparts.
4.8.3 Summary of Arthroplasty Once a nonunion has been identified, every effort
Options should be made to address the problem before six
months after the initial injury and prior to
Arthroplasty options serve as a last resort for glenohumeral soft tissue contractures develop
patients whose proximal humerus nonunions which may further limit the patient’s function. If
have left them with advanced glenohumeral this window is missed, any proposed surgical
arthritis, severe osteopenia, and insufficient bone plan should include an intraoperative assessment
stock. These procedures offer reliable pain relief, of soft tissue constraints to glenohumeral motion,
but are associated with high complication rates which should be addressed.
and, for unconstrained arthroplasties, less pre- Treatment options vary widely and range
dictable shoulder range of motion that is tied, at from nonsurgical management for minimally
least in part, to tuberosity healing. Reverse total symptomatic patients to surgical options includ-
shoulder arthroplasties avoid reliance on ing osteosynthesis with standard, fixed angle, or
tuberosity healing but require both axillary nerve locked plates, unreamed or interlocked intrame-
and deltoid function and have shorter longevity, dullary implants, and arthroplasty using hemi-
allowing them only to be recommended for arthroplasties, total shoulders, or reverse total
physiologically elderly patients. shoulders. Surgery may include augments such
as cancellous autograft and allograft or structural
grafts. From a functional perspective, postoper-
4.9 Conclusion ative Constant scores and range of motion are
higher if successful union can be achieved with
Fractures of the proximal humerus are common osteosynthesis compared to arthroplasty options.
and the majority of them unite uneventfully, Therefore, fracture characteristics that lend
many without surgical intervention. A small themselves to fixation are felt to be better prog-
percentage develop into nonunions but the cited nostic indicators, and include simpler patterns,
rates are often inflated by statistics yielded from better bone stock, and maintained vascularity.
tertiary referral centers. The small study sizes Patient factors that allow more aggressive reha-
available in the literature also cause difficulty in bilitation protocols postoperatively, such as
determining the true rate of nonunion due to the younger age and less medical comorbidity, may
infrequency of its occurrence. Nonunions of the also predict better functional outcomes, but the
proximal humerus present unique challenges due small studies available have not examined these
to biologic insults from the initial injury and variables specifically. Technological advances in
previous surgeries, bone loss, humeral head locking plate and interlocking nail design have
cavitation, osteopenia, soft tissue contractures, expanded the nonunions that may be amenable to
and infection. Risk factors for developing a osteosynthesis and every effort should be made
nonunion include fracture characteristics, such as to stabilize them. When these options are not
appropriate, arthroplasty offers favorable results 12. Sproul RC, Iyengar JJ, Devcic Z, Feeley BT. A sys-
for pain control but less predictable range of temic review of locking plate fixation of proximal
humerus fractures. Injury. 2011;42(4):408–13.
motion, which seems to be at least partially 13. Galatz LM, Iannotti JP. Management of surgical neck
dependent on achieving tuberosity union. Rev- nonunions. Orthop Clin North Am. 2000;31(1):51–61.
erse total shoulder arthroplasties offer the theo- 14. Hanson B, Neidenbach P, de Boer P, Stengel D.
retical advantage of decreased dependence on Functional outcomes after nonoperative management
of fractures of the proximal humerus. J Shoulder
tuberosity union but only one small study has Elbow Surg. 2009;18(4):612–21.
reviewed results from their use with proximal 15. Antuña SA, Sperling JW, Sánchez-Sotelo J,
humeral nonunions and more research is needed Cofield RH. Shoulder arthroplasty for proximal
to better elucidate the role for reverse TSA in humerus nonunions. J Shoulder Elbow Surg.
2002;11:114–21.
treating the challenges of proximal humeral 16. Yamane S, Suenaga N, Oizumi N, Minami A.
nonunions. Interlocking intramedullary nailing for nonunion of
the proximal humerus with Straight Nail System.
J Shoulder Elbow Surg. 2008;17(5):755–9.
17. Duralde XA, Flatow EL, Pollock RG, Nicholson GP,
Self EB, Bigliani LU. Operative treatment of
References nonunions of the surgical neck of the humerus.
1. Court-Brown CM, Caesar B. Epidemiology of adult 18. Cheung EV, Sperling JW. Management of proximal
fractures: a review. Injury. 2006;37(8):691–7. humeral nonunions and malunions. Orthop Clin N
2. Lind T, Krøner K, Jensen J. The epidemiology of Am. 2008;39(4):475–82.
fractures of the proximal humerus. Arch Orthop 19. Nayak NK, Schickendantz MS, Regan WD,
Trauma Surg. 1989;108(5):285–7. Hawkins RJ. Operative treatment of nonunion of
3. Court-Brown CM, Garg A, McQueen MM. The surgical neck fractures of the humerus. Clin
epidemiology of proximal humerus fractures. Acta Orthop. 1995;313:200–5.
Orthop Scand. 2001;72(4):365–71. 20. Rooney PJ, Cockshott WP. Pseudathrosis following
4. Lee CK, Hansen HR. Post-traumatic avascular proximal humeral fractures: a possible mechanism.
necrosis of the humeral head in displaced proximal Skeletal Radiol. 1986;15(1):21–4.
humerus fractures. J Trauma. 1981;21(9):788–91. 21. Neer CS II. Nonunion of the surgical neck of the
5. Jensen GF, Christiansen C, Boesen J, Hegedüs V, humerus. Orthop Trans. 1983;7:389.
Transbøl I. Relationship between bone mineral 22. Checchia SL, Doneux P, Miyazaki AN, Spir IA,
content and frequency of postmenopausal fractures. Bringel R, Ramos CH. Classification of non-unions
Acta Med Scand. 1983;213(1):61–3. of the proximal humerus. Int Orthop. 2000;24
6. Palvenen M, Kannus P, Niemi S, Parkkari J. Update (4):217–20.
in the epidemiology of proximal humeral fractures. 23. Martinez AA, Bejarano C, Carbonel I, Iglesias D,
Clin Orthop Relat Res. 2006;442:87–92. Gil-Albarova J, Herrera A. The treatment of proximal
7. Robinson CM. Proximal humerus fractures. In: humerus nonunions in older patients with reverse
Buchholz RW, Court-Brown CM, Heckman JD, shoulder arthroplasty. Injury. 2012;43(Suppl 2):S3–
Tornetta III P, editors. Rockwood and Green’s 6.
fractures in adults. 7th ed. Philadelphia: Lippincott 24. Duquin TR, Jacobson JA, Sanchez-Sotelo J, Sper-
Williams & Wilkins; 2010. p. 1039–105. ling JW, Cofield RH. Unconstrained shoulder arhtro-
8. Cadet ER, Yin B, Schulz B, Ahmad CS, Rosen- plasty for treatment of proximal humeral nonunions.
wasser MP. Proximal humerus and humeral shaft J Bone Joint Surg Am. 2012;94(17):1610–7.
nonunions. J Am Acad Orthop Surg. 2013;21 25. Cofield RH. Total shoulder arthroplasty with the
(9):538–47. Neer prosthesis. J Bone Joint Surg Am. 1984;66
9. Warner JP, Costouros JG, Gerber C. Fractures of the (6):899–906.
proximal humerus. In: Buchholz RW, Heckman JD, 26. Badman B, Mighell M, Drake G. Proximal humeral
Court-Brown C, editors. Rockwood and Green’s nonunions: surgical technique with fibular strut
fractures in adults. 6th ed. Philadelphia: Lippincott, allograft and fixed-angle locked plating. Tech Shoul-
Williams & Wilkins; 2006. p. 1039–105. der Elbow Surg. 2006;7(2):95–101.
10. Court-Brown CM, McQueen MM. Nonunions of the 27. Scheck M. Surgical treatment of nonunions of the
proximal humerus: Their prevalence and functional surgical neck of the humerus. Clin
outcome. J Trauma. 2008;64(6):1517–21. Orthop. 1982;167:255–9.
11. Iyengar JJ, Devcic Z, Sproul RC, Feeley BT. Non- 28. Brinker MR, O’Connor DP. Nonunions: evaluation
operative treatment of proximal humerus fractures: a and treatment. In: Browner BD, Jupiter JB,
systematic review. J Orthop Trauma. 2011;25 Levine AM, Trafton PG, Krettek C, editors. Skeletal
(10):612–7. trauma: basic science, management, and
reconstruction. 4th ed. Philadelphia: Saunders Else- proximal humerus nonunions. Int Orthop. 2009;33
vier; 2009. p. 615–708. (6):1649–54.
29. Brinker MR, O’Connor DP, Monla YT, Earth- 41. Tauber M, Brugger A, Povacz P, Resch H. Recon-
man TP. Metabolic and endocrine abnormalities in structive surgical treatment without bone grafting in
patients with nonunions. J Orthop Trama. 2007;21 nonunions of humeral surgical neck fractures.
(8):557–70. J Orthop Trauma. 2011;25(7):392–8.
30. Kazakos K, Lyras DN, Galanis V, Verettas D, 42. Walch G, Badet R, Nove-Josserand L, Levigne C.
Psillakis I, Chatzipappas C, Xarchas K. Internal Nonunions of the surgical neck of the humerus:
fixation of proximal humerus fractures using the surgical treatment with an intramedullary bone peg,
polarus intramedullary nail. Arch Orthop Trauma internal fixation, and cancellous bone grafting.
Surg. 2007;127(7):503–8. J Shoulder Elbow Surg. 1996;5(3):161–8.
31. Norris TR, Turner JA, Bovil D. Nonunion of the 43. Govender S, Csimma C, Genant HK, Valentin-Opran
upper humerus: an analysis of the etiology and A, Amit Y, Arbel R, et al. BMP-1 evaluation in
treatment in 28 cases. In: Post M, Morrey BF, surgery for tibial trauma (BESTT) study
Hawkins RJ, editors. Surgery of the shoulder. group. Recombinant human bone morphogenic
Chicago: Mosby Year Book; 1990. p. 63–7. protein-2 for treatment of open tibial fractures: a
32. Beredjiklian PK, Iannotti JP, Norris TR, Wil- prospective, controlled, randomized study of four
liams GR. Operative treatment of malunion of a hundred and fifty patients. J Bone Joint Surg Am.
fracture of the proximal aspect of the humerus. 2002;84(12):2123–34.
J Bone Joint Surg Am. 1998;80(10):1484–97. 44. Friedlander GE, Perry CR, Cole JD, Cook SD,
33. Rockwood CA, Pearce JC. Management of proximal Cierny G, Muchler GF, et al. Osteogenic protein-1
humerus nonunions. Orthop Trans. 1989;13:644. (bone morphogenic protein-7) in the treatment of
34. Clavert P, Adam P, Bevort A, Bonnomet F, tibial nonunions. J Bone Joint Surg Am. 2001;83-A
Kempf JF. Pitfalls and complication with locking Suppl 1(Pt 2):S151–8.
plate for proximal humerus fracture. J Shoulder 45. Calori GM, Tagliabue L, Gala L, d’Imporzano M,
Elbow Surg. 2010;19(4):489–94. Peretti G, Albisetti W. Application of rhBMP-7 and
35. Thanasas C, Kontakis G, Angoules A, Limb D, platelet-rich plasma in the treatment of long bone
Giannoudis P. Treatment of proximal humerus non-unions: a prospective randomized clinical study
fractures with locking plates: a systematic review. on 120 patients. Injury. 2008;39(12):1391–402.
J Shoulder Elbow Surg. 2009;18(6):837–44. 46. Garrison KR, Shemilt I, Donell S, Ryder JJ, Mug-
36. Gerber C, Hersche O, Berberat C. The clinical ford M, Harvey I, et al. Bone morphogenic protein
relevance of posttraumatic avascular necrosis of the (BMP) for fracture healing in adults. Cochrane
humeral head. J Shoulder Elbow Surg. 1998;7 Database Syst Rev. 2010;16(6):CD006950.
(6):586–90. 47. Schultz DG, U.S. Food and Drug Administration.
37. Wijgman AJ, Roolker W, Patt TW, Raaymakers El, Public health notifications (medical devices). FDA
Marti RK. Open reduction and internal fixation of public health notification: life-threatening complica-
three and four-part fractures of the proximal part of tions associated with recombinant human bone
the humerus. J Bone Joint Surg Am. 2002;84-A morphogenetic protein in cervical spine fusion.
(11):1919–25. 1 Jul 2008. Accessed 30 Sept 2015.
38. Healy WL, Jupiter JB, Kristiansen TK, White RR. 48. Boileau P, Trojani C, Walch G, Krishnan SG,
Nonunion of the proximal humerus. J Orthop Romeo A, Sinnerton R. Shoulder arthroplasty for
Trauma. 1990;4(4):424–31. the treatment of the sequelae of fractures of the
39. Ring D, McKee MD, Perey BH, Jupiter JB. The use proximal humerus. J Shoulder Elbow Surg. 2001;10
of a blade plate and autogenous cancellous bone graft (4):299–308.
in the treatment of ununited fractures of the proximal 49. Boileau P, Chuinard C, Le Huec JC, Walch G,
humerus. J Shoulder Elbow Surg. 2001;10(6):501–7. Trojani C. Proximal humerus fracture sequelae:
40. Allende C, Allende BT. The use of a new locking 90 impact of a new radiographic classification on arthro-
degree blade plate in the treatment of atrophic plasty. Clin Orthop Relat Res. 2006;442:121–30.
Supracondylar Humeral Nonunions
5
Joseph Borrelli Jr., MD, MBA
infection as a cause of their nonunion. Screening

5.1 Introduction
laboratory tests to assess the potential presence of
an infection includes a complete blood count,
Nonunions of the supracondylar area of the
erythrocyte sedimentation rate, and C-reactive
humerus following fracture are relatively
protein. The use of a combined white blood
uncommon but present the surgeon and the
cell/sulfur colloid scan has been found to be the
patient with a very challenging problem. Non-
least predictive method of revealing infection and
unions of the supracondylar area of the humerus
is not cost effective [5]. Of course, care must be
often result in limited use of the upper extremity
taken in interpreting each of these tests, as false
as well as pain and instability at the site and
positives are possible in the setting of hyper-
elbow stiffness [1, 2]. Risk factors for their
trophic nonunions, particularly those with loose
development include, soft tissue interposition at
or broken metallic implants.
the time of the initial fracture, infection,
In general, a thorough history and physical
mechanical instability, and/or poor blood supply
examination combined with orthogonal plain
at the fracture site, often resulting from excess
radiographs centered on the supracondylar and
soft tissue stripping at the time of surgery.
elbow area in question is usually sufficient to
Additionally, systemic vitamin D deficiency and
make the diagnosis of a nonunion. Reviewing
overall poor nutrition and smoking have been
serial radiographs over time will often show
associated with the development of long bone
widening of the fracture gap, loosening or
nonunions [3, 4].
breakage of the implants, and the development of
irritation-type callus, all of which are consistent
with the development of a nonunion. Recently,
5.2 Assessment
computer tomography (CT) scanning technology
has improved to the point where, when the
The assessment of patients with a supracondylar
diagnosis of a nonunion is still in doubt, CT
nonunion begins with a detailed history and
scans with sagittal and coronal reconstructions,
physical examination. All patients, but particu-
and even 3D reconstructions, may be helpful in
larly those who have previously undergone sur-
confirming the presence of a nonunion. Although
gery, should be assessed for the presence of
studies of the use of CT scans for the diagnosis
of distal humeral nonunions have not been pub-
lished, there are studies supporting their use in
J. Borrelli Jr. (&) the assessment of nonunions of the feet and
Orthopedic Surgery, BayCare Medical Group,
following spinal fusions to assess healing [6, 7].
St. Joseph’s Hospital-North, 4211 Van Dyke Rd,
Ste 200, Lutz, FL 33558, USA Magnetic resonance imaging (MRI) has not
e-mail: joseph.borrelli@BayCare.org shown to be helpful in the diagnosis of long bone

116 J. Borrelli Jr., MD, MBA
nonunions and should therefore be used spar- parallel with both plates placed posteriorly, or one
ingly in this setting. plate placed medially and one plate placed later-
ally. For nonunions of the supracondylar or
intracondylar areas deemed non-reconstructible,
5.3 Surgical Tactic particularly in the low-demand elderly patients,
cemented total elbow arthroplasty offers a good
Once the diagnosis of a supracondylar humerus solution to this difficult problem [14–16]. Occa-
nonunion has been made, the most predictable sionally, supracondylar nonunions with loss of
method to restore stability and elbow function is distal bone stock require the use of a customized
generally to remove the previously placed prosthesis, commonly used in the setting of a
implants and to perform revision open reduction tumor resection [17].
and internal fixation (ORIF) [1, 2, 8–11]. When The use of thin wire external fixation in the
treating nonunions, efforts should be made to treatment of infected distal humeral nonunions
obtain good fragment apposition and interfrag- has also been reported [18, 19]. However, the use
mentary compression between the fracture frag- of these complex external fixation devices should
ments while restoring absolute stability. This be kept in the hands of the most experienced
is often difficult because of the commonly surgeons who perform these procedures routinely
small size of the distal fragment, which is and who fully understand the potential and lim-
further compromised by the presence of disuse itations of these devices.
osteopenia. In oligotrophic and atrophic non- In the setting of an infected supracondylar
unions, efforts should be made not only to obtain humeral nonunion, a thorough irrigation and
good apposition and interfragmentary compres- debridement to include the removal of all avas-
sion between the fracture fragments but also in cular and infected bone, all metallic implants,
the effort to improve the local biological envi- and any devitalized soft tissues must be under-
ronment. The addition of autologous bone grafts, taken. Temporary stability of the fracture frag-
allograft demineralized bone matrix (DBX), or ments is advantageous during the antibiotic
other biologic stimulants to the nonunion site phase and can sometimes (ideally) be provided
during revision ORIF has been shown to be with the application of an external fixator inser-
helpful in obtaining union [11–13]. ted into the proximal and distal fragments, or
Generally, too distal to be effectively treated spanning the elbow. The use of circular frames
with intramedullary nails, nonunions of the utilizing thin wires also has a role here, although
supracondylar area of the humerus are commonly great care must be taken to avoid neurovascular
stabilized with a combination of plate(s) and structures when placing the transfixation wires in
screws [2, 11]. In the author’s experience, non- each of the fragments [18, 19]. The surrounding
unions of the distal humeral shaft (upper regions of soft tissue envelope must also be addressed, as
the supracondylar area of the humerus) can be inadequate soft tissue coverage of the bone
effectively treated with a single large fragment fragments and elbow joint will compromise the
compression plate, as long as the plate can be chances for infection eradication and ultimate
placed distal enough to allow at least two to three healing of the nonunion. In addition to all of
bicortical screws (4–6 cortices) to be placed within these considerations, the patient will also need a
the distal fragment without compromise to the minimum of six weeks of bacterial-specific sys-
olecranon fossa. Nonunions of the true supra- temic antibiotics. Generally, administration of
condylar region of the distal humerus are generally these powerful antibiotics should be performed
treated with double plating after restoring articular in close consultation with an expert in infectious
congruity, if necessary, of the distal humerus. disease. Additionally, to bring additional antibi-
Anatomically specific plates are available with otics into the infected nonunion site, a poly-
locking screw capability and are positioned either methylmethacrylate spacer impregnated with a
90° to each other (posterolateral and medial) or broad-spectrum antibiotic can be placed in the
5 Supracondylar Humeral Nonunions 117
nonunion space and removed during the defini- preoperatively, or if there is obvious impinge-
tive fixation of the nonunion. ment of the ulnar nerve after revision of the
nonunion or replacement of the elbow, ulnar
nerve transposition may certainly be warranted.
5.4 Surgical Approaches At times, augmentation of the nonunion fixa-
tion construct may be desirable to support heal-
In most cases where primary or revision ORIF of ing. In these cases, nonunion augmentation could
a supracondylar nonunion is being undertaken, include autologous bone graft, DBX or
the distal humerus is approached via a posterior bone-stimulating proteins, or a combination of
incision, with the patient positioned either in the these [12]. If an autologous iliac crest bone graft
lateral decubitus or prone position. To gain is to be utilized, the iliac crest can be easily
access to the distal humerus, the exposure can be exposed with the patient in the lateral or prone
performed by developing planes medial and lat- position. Preferentially, the outer iliac table and
eral to the distal triceps and its tendonitis inser- the intra-table trabecular bone are harvested with
tion. Of course, this should only be done once the an acetabular reamer, as previously described
ulnar nerve has been identified and isolated so [22–25]. This technique provides finely minced
that it can be protected throughout the procedure. pieces of cortical and cancellus bone to facilitate
The distal humerus can also be approached nonunion healing. Corticocancellous strips and
through a triceps splitting approach or a triceps cancellous bone fragments can also be harvested
tendon turndown. If the exposure is needed from the iliac crest, and at times from the proxi-
proximal to the distal flare of the humerus, care mal ulna, and placed at the nonunion site. Dem-
must be taken to locate and protect the radial ineralized bone matrix (DBX) is available in
nerve and the profunda brachial artery as they several different forms and has been shown to
pass from proximal medial to distal lateral and augment healing of humeral nonunions [12].
around the lateral aspect of the distal humerus Bone morphogenic proteins (BMP) have also
[20]. If there is an intra-articular component of been used to augment healing in certain non-
the distal humeral nonunion, then consideration unions, but in general their use in the treatment of
must be given to performing an olecranon humeral nonunions has not been approved [26–
osteotomy to allow exposure and anatomic 29]. However, BMP-7 has been approved for use
reduction of the articular component. If a total in the setting of a persistent fracture nonunion.
elbow arthroplasty is to be performed, then the Use of these proteins in patients of childbearing
elbow can generally be replaced through inter- age must be made with great caution.
vals medial and lateral to the distal triceps
without elevating the triceps off the olecranon
and proximal ulna and without an olecranon 5.5 Postoperative Rehabilitation
osteotomy [14–16]. Again, in each case, the
ulnar nerve must be identified and carefully Generally, once the surgical incision has healed
mobilized from around the medial epicondyle and the sutures have been removed, rehabilitation
and protected. During each of these approaches of the elbow, shoulder, and forearm is begun in
for treatment of distal humeral nonunions, earnest. Rehabilitation should include active and
transposition of the ulnar nerve is not necessarily active assisted flexion and extension of the
transposed and, in fact, transposition is discour- elbow, pronation and supination of the forearm,
aged; a recent study has shown that transposition and range of motion of the shoulder. Particular
of the ulnar nerve following ORIF of acute distal attention should be paid to elbow extension and
humerus fractures has been associated with a forearm supination, as these are often the most
fourfold increase in postoperative ulnar neuritis difficult motions to recover. Occasionally,
[21]. If the patient had signs and symptoms of dynamic and passive splinting devices in con-
ulnar nerve compression at the elbow junction with therapist-supervised rehabilitation
are necessary to help the patient regain forearm residual instability at the nonunion site. There-
and elbow range of motion [30, 31]. These fore, all efforts must be made to obtain fragment
techniques can usually be started within the first apposition and interfragmentary compression at
six to eight weeks once there is radiographic the time of revision surgery. Atrophic nonunions
evidence of healing at the nonunion site. Also, at will benefit by the addition of biologically active
approximately 6 weeks postoperatively, physical materials (DBX, BMPs, iliac crest bone graft),
therapist directed range of motion exercises and while oligotrophic and hypertrophic nonunions
muscle-strengthening exercises can be initiated. can often be treated with reduction and inter-
The goal of the postoperative rehabilitation fragmentary compression alone. Of course, sel-
should be to achieve full flexion and extension of dom is the outcome after revision ORIF for a
the elbow, full pronation and supination of the nonunion typically as good as after primary
forearm, and the restoration of normal shoulder healing following the initial procedure.
motion.
5.7 Conclusion
5.6 Complications
Open reduction and internal fixation of distal
As with any surgical intervention the risk of humeral nonunions can be quite challenging. It is
complications is substantial and is increased as important to try and determine the cause of the
the number of surgeries performed on the distal nonunion so that these factors can be addressed
humerus increases. These risks include infection, and reversed. Typically, nonunions in this area are
nerve damage, vascular injury, persistent non- thought to occur as a result of soft tissue interpo-
union, loss of fixation, and implant failure. In an sition (non-operatively treated fractures), fracture
effort to avoid infection, preoperative antibiotics fragment instability, underlying infection, poor
should be administered immediately before local vascular environment, or underlying sys-
tourniquet inflation and incision and for 24 h temic metabolic conditions. Before undertaking
postoperatively. Also, meticulous soft tissue revision surgery, the surgeon should examine the
handling techniques and timely surgery while patient thoroughly, the patient’s elbow and fore-
maintaining soft tissue attachments to the non- arm range of motion should be documented, and
union fragments should be adhered to in order to necessary images, to best define the nonunion,
help decrease the incidence of deep infection. To should be obtained. The presence of infection
avoid nerve injury the ulnar nerve and, on needs to be determined and a thorough preopera-
occasion, the radial nerve should be identified tive plan to address each of the contributing factors
and isolated sufficiently to allow gentle retraction developed. Prior to surgery, the surgeon must
out of harm’s way during reduction and fixation discuss the potential risks and benefits, as well as
of the nonunion. Knowing preoperatively whe- the limitations, of the proposed surgery with the
ther or not the ulnar nerve was transposed patient. Surgery should be carried out with the
will assist greatly in its identification and goals to restore length, alignment, and rotation of
protection during the procedure. Intra-operative the extremity, contracture release of the elbow,
use of a tourniquet or meticulous intra-operative and to obtain fragment reduction and interfrag-
hemostasis will minimize the risk of intra- mentary compression and stable fixation to allow
operative blood loss and postoperative hema- early elbow and forearm range of motion and
toma formation, which can delay soft tissue osseous healing. Bone graft or bone graft substi-
healing and put the patient at increased risk of tutes should be used liberally in the setting of
infection. Persistence of the nonunion is not oligotrophic or atrophic nonunions. Release and
uncommon and is most likely the result of poor mobilization of the elbow and proximal radius and
vascularity of the nonunion fragments, scarring, ulnar joints are essential to facilitate healing of the
and hypovascular surrounding soft tissues and/or nonunion and recovery of forearm range of
motion. Patients must be made aware that a fair removal of the implants, debridement of the
percentage of these nonunions persist and addi- nonunion site, revision ORIF with 90–90 spe-
tional surgery may be necessary. cialty plates. An olecranon osteotomy was per-
formed to maximize exposure of the distal
fragment and to aid in the mobilization of the
5.8 Case Discussions elbow joint. The nonunion site was also aug-
mented with demineralized bone matrix and
Case 1 allograft cancellous bone chips and the nonunion
went onto heal uneventfully.
A 45-year-old male who sustained a gunshot
wound of his distal arm and humerus resulting in Case 2
a supracondylar humerus fracture. The patient
was initially treated with irrigation and debride- A 53-year-old male fell at home sustaining a
ment and ORIF of his distal humerus through a closed supracondylar intercondylar humeral
posterior approach with parallel plates. fracture. He was originally treated with a closed
This fracture was repaired with parallel plates reduction and splinting of the elbow and was
(Fig. 5.1), including a one-third tubular plate subsequently treated with ORIF with an olecra-
laterally, and a locking specialty plate medially. non osteotomy. Early motion was begun at
A nonunion developed as indicated by the loose approximately 10 days postoperatively, and his
and failing fixation, the appearance of a gap at the incision healed uneventfully. Although his pain
previous fracture site, and considerable irritation decreased during the first several weeks, at six
callus along the medial and lateral humerus. weeks post-op, he felt a snap in the elbow and
The nonunion (Fig. 5.2) was first assessed for experienced increased pain and swelling of the
the presence of infection and then treated with elbow (Figs. 5.3, 5.4, 5.5 and 5.6).
Fig. 5.1 Anteroposterior

(a) and lateral (b) radiographs
of a supracondylar humerus
nonunion. Loss of fixation
and alignment of the fracture
fragments can be seen, as well
as a gap at the previous
fracture site
Fig. 5.2 Anteroposterior (a) and lateral (b) radiographs of the nonunion site, which has been reduced and stabilized
with a medial and lateral plate and bone grafted with an iliac crest bone graft assisted by an olecranon osteotomy
Fig. 5.3 Anteroposterior (a) and lateral (b) radiographs demonstrating a displaced supracondylar intercondylar
humerus fracture
Fig. 5.4 Anteroposterior (a) and lateral (b) radiographs following open reduction and internal fixation of the displaced
supracondylar intercondylar humerus fracture shown in Fig. 5.3
Fig. 5.5 Anteroposterior (a), oblique (b), and lateral (c) radiographs show failure of medial plate, an indication that a
nonunion is developing
Fig. 5.6 Anteroposterior (a), oblique (b), and lateral (c) radiographs after removal of medial implants, reduction of the
medial condyle, and revision open reduction and internal fixation with an iliac crest bone graft
associated with unsuccessful humeral condylar frac-

References ture repair with autologous iliac bone reconstruction.
1. Ring D, Gullotta L, Jupiter JB. Unstable nonunions 10. Helfet DL, Kloen P, Anand N, Rosen HS. Open
of the distal part of the humerus. J Bone Joint Surg reduction and internal fixation of delayed unions and
Am. 2003;85-A(6):1040–5. nonunions of fractures of the distal part of the
2. Jupiter JB. The management of nonunion and humerus. J Bone Joint Surg Am. 2003;85-A(1):
malunion of distal humerus—30 year experience. 33–40.
J Ortho Trauma. 2008;22(10):742–50. 11. Helfet DL, Kloen A, Anand N, Rosen HS. ORIF of
3. Hak DJ, Fitzpatrick D, Bishop JA, Marsh JL, Tilp S, delayed unions and nonunions of distal humeral
Schnettler R, et al. Delayed union and nonunions: fractures. Surgical technique. J Bone Joint Surg Am.
epidemiology, clinical issues, and financial aspects. 2004;86-A (Suppl 1):18–29.
Injury. 2014;45(Suppl 2):S3–7. 12. Hierholzer C, Sama D, Toro JB, Peterson M,
4. Brinker MA, O’Connor DP, Monla YT. Metabolic Helfet DL. Plate fixation of ununited humeral shaft
and endocrine abnormalities in patients with non- fracture: effect of type of bone graft on healing.
unions. J Ortho Trauma. 2007;21(8):557–70. J Bone Joint Surg Am. 2006;88(7):1442–7.
5. Stucken C, Olszewski DC, Creevy WR, 13. Crawford CH 3rd, Seligson D. Atrophic nonunion of
Murakami AM, Tornetta P. Preoperative diagnosis humeal diaphysis treated with locking plate and
of infection in patients with nonunions. J Bone Joint recombinant bone morphogenetic protein: nine cases.
Surg Am. 2013;95(15):1409–12. Am J Orthop. 2009;38(11):567–70.
6. Carreon LY, Djurasovic M, Glassman SD, 14. Mansat P, Bonnevialle N, Rongieres M, Mansat M,
Sailer P. Diagnostic accuracy and reliability of Bonnevialle P. Experience with the Coonrad-Morrey
fine-cut CT scans with reconstructions to determine total elbow arthroplasty: 78 consecutive total elbow
the status of an instrumented posterolateral fusion arthroplasties reviewed with an average 5 years of
with surgical exploration as reference standard. follow-up. J Shoulder Elbow Surg. 2013;22(11):
Spine. 2007;32(8):892–5. 1461–8.
7. Jones CP, Coughlin MJ, Shurnas PS. Prospective CT 15. Espiga X, Antuna SA, Ferreres A. Linked total elbow
scan evaluation of hindfoot nonunions treated with arthroplasty as treatment of distal humerus nonunions
revision surgery and low-intensity ultrasound stimu- in patient older that 70 years. Acta Ortho Belg.
lation. Foot Ankle Int. 2006;27(4):229–35. 2011;77(3):304–10.
8. Allende C, Allende BT. Post-traumatic distal 16. Baksi DP, Pal AK, Baksi D. Prosthetic replacement
humerus non-union: Open reduction and internal of elbow for intercondylar fractures (recent of
fixation: long-term results. Int Ortho. 2009;33(5): ununited) of humerus in elderly. Int Ortho. 2011;
1289–94. 35(8):1171–7.
9. Niu Y, Bai Y, Xu S, Wu D, Liu X, Wang P, et al. 17. Anmin A, Suresh S, Sanghrajka A, Cannon SR,
Treatment of bone nonunion and bone defects Briggs TW, Unwin P. Custom-made endoprosthetic
reconstruction of the distal humerus for non- 25. Brawley SC, Simpson RB. Results of an alternative
tumorous pathology. Acta Orthop Belg. 2008;74(4): autologous iliac crest bone graft method. Orthope-
446–50. dics. 2006;29(4):342–5.
18. Brinker MR, O’Connor DP, Crouch CC, 26. Murena L, Canton G, Vulcano E, Surace MF,
Mehlhoff TL, Bennett JB. Ilizarov treatment of Cherubino P. Treatment of humeral shaft aseptic
infected nonunions of the distal humerus after failure nonunions in elderly patients with opposite structural
of internal fixation: an outcomes study. J Orthop allograft, BMP-7, and mesenchymal stem cells.
Trauma. 2007;21(3):178–84. Orthopedics. 2014;37(2):201–5.
19. Ring D, Jupiter JB, Toh S. Salvage of contaminated 27. Starman JS, Bosse MJ, Cates CA, Norton HJ. Recom-
fractures of the distal humerus with thin wire binant human bone morphogenetic protein-2 use in
external fixation. Clin Orthop Relat Res. 1999;359: the off-label treatment of nonunions and acute
203–8. fractures: a retrospective review. J Trauma Acute
20. Gerwin M, Hotchkiss RN, Weiland AJ. Alterative Care Surg. 2012;72(3):676–81.
operative exposures of the posterior aspect of the 28. Giannoudis PV, Kanakaris NK, Dimitriou R, Gill I,
humeral diaphysis with reference to the radial nerve. Kolimarala V, Montgomery RJ. The synergistic
J Bone Joint Surg Am. 1996;78(11):1690–5. effect of autograft and BMP-7 in the treatment of
21. Chen RC, Harris DL, Leduc S, Borrelli J, Tornetta P, atrophic nonunions. Clin Orthop Relat Res. 2009;467
Ricci WM. Is ulnar nerve transposition beneficial (12):3239–48.
during open reduction internal fixation of distal 29. Bong MR, Capla EL, Egol KA, Sorkin AT, Diste-
humerus fractures? J Ortho Trauma. 2010;24(7): fano M, Buckle R, et al. Osteogenic protein-1 (bone
391–4. morphogenic protein-7) combined with various
22. Sanders R, DiPasquale T. A technique for obtaining adjuncts in the treatment of humeral diaphyseal
bone graft. J Ortho Trauma. 1989;3(4):287–9. nonunions. Bull Hosp Jt Dis. 2005;63(1–2):20–3.
23. Westrich GH, Geller DS, O’Malley MJ, Deland JT, 30. Lindenhovius AL, Doornberg JN, Brouwer KM,
Helfet DL. Anterior iliac crest bone graft harvesting Jupiter JB, Mudgal CS, Ring D. A prospective
using the corticocancellous reamer system. J Ortho randomized controlled trial of dynamic versus static
Trauma. 2001;15(7):500–5. progressive elbow splinting for posttraumatic elbow
24. Gossman DG, Rosenblum W, Arosarena O, stiffness. J Bone Joint Surg Am. 2012;94(8):694–700.
Valentino J. The acetabular reamer: a unique tool 31. Bash DS, Spur ME. An alternative to turnbuckle
for anterior iliac crest bone graft harvesting. Laryn- splinting for elbow flexion. J Hand Ther. 2000;13(3):
goscope. 2005;115(3):557–9. 237–40.
Nonunions of the Forearm
6
Fred G. Corley, MD and Ben S. Francisco, MD
tion appropriately. The forearm axis of rotation

6.1 Introduction
passes from the center of the radial head through
the fovea of the distal ulna [5], thereby allowing
Although nonunion of forearm fractures is an
the mobile radius to rotate around the fixed ulna
uncommon outcome, it does occur and proper
150–180° as the forearm musculature moves the
treatment of the nonunion is essential to preserve
hand from pronation to supination. Working
function. The forearm itself is considered by
simultaneously with the carpus, this degree of
many to be a functional unit, or even a joint, and
freedom provides the forearm with the ability to
like articular fractures, forearm nonunions should
position the hand intricately in space.
be restored to anatomic alignment.
Various authors have determined forearm
range of motion with 68–70° pronation and 75–
85° supination being average [6–8]. However,
6.2 Anatomy
Morrey et al. noted that only 55° of supination
6.2.1 Overview and 50° of pronation are needed for most activ-
ities of daily living, indicating that supination is
The forearm is composed of the radius and ulna, needed more than pronation, one of the reasons
and the “forearm joint” is the result of the for that being, if one abducts the shoulder, it
interplay between these two bones at their two essentially pronates the forearm.
distinct articulations proximally and distally, the
proximal radioulnar joint (PRUJ) and the distal
radioulnar joint (DRUJ). Interposed between
6.2.2 Bony Anatomy
these two joints and the two bones is the in-
The radius is a long bone with proximal and
terosseous membrane (IOM), which provides
distal epiphyses. At the proximal end, the radial
stability to the forearm and transfers forces from
head is nearly round and articulates with the ulna
the distal radius and ulna proximally, as well as
in its radial fossa and also with the capitellum of
stabilizing the PRUJ and DRUJ [1–4]. The
the humerus. As the radius moves distal, it nar-
coordinated action and stability provided by
rows to form the radial neck. On its ulnar side,
these three structures allow the forearm to func-
there is a bony prominence called the radial or
bicpital tuberosity, where the biceps tendon
inserts. In its midsubstance, the radius is trian-
F.G. Corley (&) B.S. Francisco, gular in shape. The ulnar border is the apex of the
Department of Orthopedics, University of Texas triangle and is the origin of the IOM.
Health Science Center at San Antonio, 7703 Floyd
Curl Dr., MC 7774, San Antonio, TX 78229, USA In addition, the radius has three bows. At the
e-mail: corley@uthscsa.edu distal one-fifth, the radius has a convex dorsal

126 F.G. Corley, MD and B.S. Francisco, MD
bow, and at the proximal one-fifth, the radius has This apex receives the insertions of the IOM. At
a convex ventral bow. The middle three-fifths its distal aspect, the ulna flares to form the ulnar
contain the most prominent radial bow. This bow head and ulnar styloid. On its radial side, the ulna
corresponds to the insertion sites of the pronator articulates with the stigmoid notch of the radius
teres and supinator muscles, and hence, the radial to form the DRUJ. The ulnar styloid forms the
bow is vital in ensuring correct forearm rotation. origin for the triangular fibrocartilage complex
A fracture in this area leads to the most disability (TFCC).
if the fracture is not anatomically aligned, as the
lever arms of these respective muscles are short-
ened [9]. Interestingly, in Burwell and Charnley’s 6.2.3 Muscle Anatomy
series of 231 fractures in 150 adult patients that
93% of radius fractures occurred in the middle The volar forearm is comprised of fourteen
three-fifths of the radius [10] and Sage and Smith muscles, whereas the dorsal forearm has thirteen
reported that three-fifths of their radius fractures involved muscles. As one moves from proximal
were found in the middle third [11]. to distal along the shaft of the radius, the fol-
Therefore, in forearm fractures, the radial bow lowing muscles are encountered: insertion of
must be restored or loss of rotation will ensue. brachioradialis, origin of pronator quadratus,
Schemitsch et al. [12] demonstrate the method by origin of flexor pollicis longus, origin of flexor
which the apex of the radial bow can be found. digitorum superficialis, insertion of pronator
This is foundational in the treatment of forearm teres, insertion of supinator, and insertion of
fractures and the prevention of anatomic and biceps.
rotational malunion. As one moves distal to proximal along the
At its distal end, the radius flares to receive dorsal radius, the following muscles are
articulations with the ulna at the sigmoid notch, encountered: insertion of brachioradialis, origin
the lunate and the scaphoid at the lunate and of extensor pollicis brevis, insertion of pronator
scaphoid fossae, and also forms the radial styloid teres, insertion of abductor pollicis longus, and
on the radial side to receive the insertion of the insertion of supinator [14].
brachioradialis. The radius rotates about the ulna The musculature of the forearm is the driving
150–180°. This arc of rotation has a longitudinal force of forearm motion and plays vital roles in
axis that is centered in the radial head proximally hand and wrist function. Their coordinated
and at its distal end passes through the center of actions contribute to the needed pronation,
the ulnar head and consequently the index finger. supination, flexion, and extension needed to
As the radius rotates around the ulna, the ulna perform the many gross and fine movements
moves in a varus–valgus direction about 9° at the needed in the complex functions involved in
elbow, thus allowing the ulna to move out of the daily life.
way of the rotating radius distally [13]. It is also these muscles that can contribute to
The ulna is also a long bone. At its proximal the formation of anatomic and rotational malu-
end is the olecranon, which receives the insertion nions that are a result of forearm fractures. The
of the triceps. Anterior to this is the trochlear muscles responsible for pronation are mainly the
notch, which forms the stable articulation with pronator teres and to a lesser extent the pronator
the trochlea of the humerus. The trochlear notch quadratus, while the biceps brachii and supinator
terminates in the coronoid process. On its radial are supinators. It is these muscles that produce
side is the radial notch or fossa, which, together the majority of deformity in fractures of the
with the radial head, forms the PRUJ. On the forearm, causing the fracture ends to approach
ulnar side of the trochlear notch is a small each other centrally toward IOM. Furthermore,
tuberosity, which receives the insertion of the the proximal fragments tend to be flexed, the
brachialis. The diaphysis of the ulna is triangular ulna by the brachialis and the radius by the
in shape, with the apex pointed to the radius. biceps brachii [9].
6 Nonunions of the Forearm 127
The brachioradialis is also a major deforming roughly 3.5 cm in width and 0.94 mm at its
force. Its action is best demonstrated in a thickest point [16].
Galeazzi fracture, where the distal one-third of The distal membranous portion is composed
the radius is pulled into valgus, as there is no of the distal oblique bundle. This portion of the
opposing force. The pronator quadratus is also IOM is found under the pronator quadratus,
involved in this fracture, wherein it pulls the inserts on the inferior rim of the sigmoid notch,
distal fragment into pronation as a result of and blends with the DRUJ, TFCC, and dorsal
unopposed action. and palmar ligaments at its most distal aspect
[19]. Working in concert, these structures serve
to stabilize the DRUJ [2–4].
6.2.4 Distal Radial Ulnar Joint The proximal oblique cord and dorsal oblique
Anatomy accessory cord comprise the proximal membra-
nous portion [19]. The dorsal oblique accessory
Distally, the ulna articulates with the radius at the cord has also been called the proximal ascending
sigmoid notch to form the DRUJ. This joint is bundle [17], or the proximal interosseous band
stabilized primarily by the TFCC. The palmar [20]. The proximal oblique cord is found
radioulnar ligament, dorsal radioulnar ligament, between the origin of flexor digitorum profundus
articular disk, ulnocarpal ligaments, extensor and the supinator, originates from the anterolat-
carpi ulnaris subsheath, and meniscus homolog eral aspect of the coronoid process, and inserts
comprise the TFCC. The ligamentous complex is just distal to the radial tuberosity. The dorsal
the primary stabilizer of the DRUJ, whereas the oblique accessory cord is located below the ori-
fibrocartilage component transmits force across gin of the abductor pollicus longus muscle [19].
the ulnocarpal joint. The differences in curvatures The central ligamentous complex is composed
in the ulnar head and sigmoid notch allow for of several distinct bands: the stout central band,
DRUJ incongruity and thus the ability of these one to five accessory bands, membranous por-
two structures to rotate and translate relative to tions, and the proximal interosseous band [19,
one another, thereby providing a portion of the 20]. The central ligamentous complex is divided
rotation necessary for forearm movements [15]. into the central and accessory bands. The central
band is the most robust of the bands of IOM and
is always present; as such, it is considered to be
6.2.5 Interosseous Membrane of prime importance. Furthermore, it comprises
Anatomy 40–60% of the total IOM [21]. Hotchkiss et al.
reported that it provided 71% of the stiffness of
The IOM and its anatomy have been studied both the IOM [16]. The central band, which is 3.5 cm
anatomically and biomechanically by various in width or 2.6 cm if measured perpendicular to
authors [1, 16–21]. Biomechanically, the IOM its fibers, originates on the radius an average of
serves as an origin for forearm musculature, 7.7 cm distal to the articular surface of the radial
stabilizes the DRUJ [2–4] and the longitudinal head. As the central band moves distally toward
forearm, transmits loads from the radius to the its ulnar insertion at a 21 degree angle relative to
ulna, and allows for smooth forearm rotation [1, the longitudinal axis of the ulna, the fibers fan
3, 4, 8–11]. Anatomically, the IOM can be out and form an insertion 4.2 cm in length on the
divided into distal membranous, middle liga- ulna. The average insertion point of the central
mentous, and proximal membranous portions band is 13.7 cm distal to the tip of the olecranon
[19]. Together, these structures average roughly [20]. The accessory bands are distinct anatomic
22 cm in length, with the radial origin being an structures, separate from the central band, and
average of 10.6 cm in length and ulnar insertion vary in number. Furthermore, they are less robust
measuring 10.6 cm [20]. The width of the IOM is of structures [19, 20].
6.2.6 Proximal Radioulnar Joint robust soft tissue envelope and fracture charac-
Anatomy teristics are such that it does not allow the frag-
ments to be reduced and maintained adequately.
Proximally, the radius articulates with the ulna at We do not have a specific bone defect size to
the PRUJ, which is composed of the radial head, employ acute bone grafting. If the patient has a
the capitellum of the humerus, and the lesser contaminated open fracture and there is a seg-
sigmoid notch of the ulna. The PRUJ is con- mental bone loss that does not allow for reduc-
strained and stabilized by the annular ligament, tion and stabilization, one can consider using a
the lateral ulnar collateral ligament, the radial Masquelet technique, wherein an
collateral ligament, and the surrounding elbow antibiotic-impregnated cement spacer is placed in
joint capsule and musculature. The intrinsic bony the bone defect. The surgeon can then bring the
anatomy of the proximal ulna and its articulation patient back to the operating room after 4–
with the distal humerus allow the ulna to be a 6 weeks for removal of the antibiotic cement
fixed construct around which the radius can spacer and insertion of bone graft. The bone
rotate. graft, in both instances, can be harvested from
the anterior iliac crest. Another option that we
have often employed if there is an open fracture
6.3 Causes of Nonunion and/or significant bone loss is simply plating the
fracture at the appropriate length without insert-
Overall rates of nonunion of forearm fractures ing bone graft at the index procedure. The patient
involving the diaphysis are less than 2% for the can then be brought back to the operating room
radius and 4–6% for the ulna [22]. at 4–6 weeks, bone graft can then be placed, and
With the advent of AO (Arbeitsgemeinschaft the fracture replated.
für Osteosynthesefragen/Association for the There is a higher incidence of nonunions of
Study of Internal Fixation) principles of com- the ulna for several reasons. First, the ulna has a
pression plating and implants that resist poorer blood supply due to a less-robust soft
deforming forces, most diaphyseal fractures of tissue envelope. Second, it is more common for it
the radius and ulna will heal in a timely manner to be an open fracture. Third, it is fixed at both
when fixed surgically. ends, resulting in increased strain and stress on
Nonunions that do occur can be classified into the ulna as compared to the radius. However, a
hypertrophic, oligotrophic, or atrophic non- patient may function well with a nonunion of the
unions. Hypertrophic nonunions are rare in the ulna if that patient is able to perform their
forearm, whereas atrophic nonunions are far activities of daily living without incident or sig-
more common. The blood supply to atrophic nificant pain. If the patient is functioning well
nonunions is poor, and little callous is formed with a nonunion, we do not advocate for surgical
adding to the challenge of healing [23]. intervention.
Causes of nonunion have been enumerated, Maes et al. [24] reported on 133 fractures of
but bone loss; comminution; infection; poor soft the forearm initially treated nonoperatively. Of
tissue coverage or absence of soft tissue cover- that total, 81 involved both the radius and ulna,
age; inadequate stabilization; patient comorbidi- 22 were isolated to the ulna, and 30 to the radius.
ties, including smoking, diabetes, malnutrition, Interestingly, the nonunion incidence was 9/81
renal dysfunction, and poor surgical technique (11%) for combined radial and ulnar shaft frac-
with devascularization of the bone; and inade- tures, 6/22 (27%) for isolated ulnar shaft frac-
quate stabilization and lack of compression all tures, and 1/30 (3%) for isolated radial shaft
contribute to nonunion [23]. fractures. These types of results for nonoperative
Often, the question arises when one should treatment encouraged early authors to pursue
use bone graft acutely to prevent a nonunion. We surgical treatment to improve clinical outcomes
employ acute bone grafting when there is a [9–11, 25–30].
Subsequently, Anderson et al. [31] reported on Preoperative planning includes, if needed:

244 patients with 330 fractures of the forearm that perfect orthogonal anteroposterior and lateral
were treated with acute open reduction and internal radiographs of both forearms. Computed
fixation using ASIF (Association for the Study of tomography is also a valuable option to help the
Internal Fixation) compression techniques with or surgeon in determining the surgical approach, as
without some other form of fixation, such as a Sage well as in evaluating the nonunion.
nail and acute bone grafting if more than one-third If the patient was surgically treated at another
of the circumference of the shaft was comminuted. institution, operative reports and other pertinent
There were 112 fractures of both the radius and records should be sought and obtained.
ulna, 82 isolated fractures of the radial shaft, and 50 If the possibility of infection exists, complete
isolated fractures of the ulna. 97.1% of fractures blood count, erythrocyte sedimentation rate, and
healed. There were 7 patients who developed C-reactive protein tests should be ordered to help
nonunions, 4 nonunions of the radius and 5 of the determine the need for more advanced imaging
ulna. This report demonstrated much improved studies.
results with regard to nonunion for compression Thought should be given to the implants that
plating of forearm fractures, in particular fractures will be used. For example, will dynamic com-
involving both the radius and ulna. pression plate (DCP), limited-contact dynamic
When clinically assessing a patient with a compression plate (LC-DCP), and/or locking
diaphyseal forearm fracture, the surgeon must plates be applied? If infection is a concern, one
ask himself the following question: Have the should also consider the need for external fixa-
bones had adequate time to heal? Four to six tion, cement spacers, and the possibility of
months are the usual period given for forearm staging the surgical procedures.
bony and functional healing. Other questions that If the nonunion is the result of a prior surgical
should be asked: Is the fracture unstable? Is there failure, instruments for screw removal should be
adequate soft tissue coverage? Were there mul- on hand, as well as possible power tools for
tiple injuries? Has the patient had continual pain? cutting titanium and stainless steel.
Has there been any progression of healing from Thought should also be given to the need
month one to month six? for bone graft. If autograft is chosen, the
In considering the radiograph, the surgeon planned harvest site should be prepped. We
must ask certain questions: recommend not to take bone graft from prox-
imal ulna/distal radius because it creates a
1. Are there gaps on the X-ray? stress riser, and thus, the likelihood of iatro-
2. Has there been any softening of the fracture genic fracture increases. If bone graft is nee-
surfaces (a sign of healing)? ded, we either use anterior iliac crest if cortical
3. Is there any hardware loosening? is desired, or if cancellous bone is needed, a
4. Is there any hardware in the fracture site? significant amount can be obtained from the
5. Is there any sign of infection? proximal tibia by making a bone window at
6. Is the fracture site atrophic or hypertrophic? Gerdy’s tubercle and then curetting the can-
cellous bone. Care must be taken to not pen-
etrate the knee joint. Bone graft substitutes
should be on hand if autograft is not chosen.
6.4 Surgical Treatment If there is a possibility of infection, the frac-
ture site should be exposed first before graft
6.4.1 Preoperative Planning taken.
Appropriate medical evaluation, surgical risk
After determination that a nonunion exists, the stratification, and laboratory work should be
surgical alternatives are discussed with the done, and blood products, if needed, should be
patient and the surgery is planned. available.
6.4.2 Operating Room Setup the muscle and not the bone. Choose the interval
between the flexor carpi ulnaris and extensor
An adequate sized room should be obtained to carpi ulnaris. This is adequate for exposure of all
allow for an arm table, C-arm, and back tables for the ulnar diaphysis.
bone graft. If an arm table is used, the C-arm is For the radius, the distal two-thirds of the
positioned to enter from the foot to obtain adequate shaft can be approached through a volar-Henry
views. The monitor is placed at the location most approach.
convenient and freely viewed by the surgeon.
The arm should be draped out at the shoulder 1. The skin incision should follow a line
and a sterile tourniquet applied. The draping of drawn from the radial styloid distally to the
the patient should also allow the forearm to be biceps tendon proximally.
positioned across the chest. If needed, access to 2. Sharply incise the skin and subcutaneous
the opposite side of the table allows the arm to be tissue to the forearm fascia.
positioned over the chest. 3. Incise the fascia over the flexor carpi radi-
For the procedure, the surgeon should arrange alis tendon down to the tendon itself.
for the presence of capable assistants and the 4. Use a moist lap sponge to dissect the sub-
engagement of an anesthesiologist. cutaneous tissue off the fascia, exposing the
With regard to instruments that should be interval between the flexor carpi radialis and
available, bone tools, rongeurs, osteotomes, the radial artery.
burrs, curettes, screw removal and broken screw 5. The artery does not need to be dissected out
removal sets, and saws and/or burrs capable of completely unless the incision approaches
cutting stainless steel and titanium should all be the mid-forearm.
readily accessible in the operating room suite. 6. If the fracture requires exposure proximal to
There should also be a discussion with the the mid-forearm, the radial artery needs to
patient about the type of anesthesia to be used. be dissected so that it can be mobilized and
Will regional or general anesthesia be adminis- retracted either radially or ulnarly.
tered? If no autogenous iliac graft is needed, 7. The deep interval between the flexor polli-
regional anesthesia is preferred. cis longus and the brachioradialis is devel-
In addition, the surgeon should discuss post- oped proximally and distally.
operative pain, bleeding, swelling, and compart- 8. The pronator quadratus is dissected and
ment syndrome with patient and the appropriate subperiosteally cleared off the distal radius,
treatment and response to these postoperative from its radial styloid attachment.
conditions. 9. If proximal exposure is needed, the pronator
teres is sharply incised and its tendinous
insertion dissected off the bone.
6.4.3 Surgical Approach 10. Volar exposure can be extended proximal to
and Exposure the bicipital tuberosity by ligating the radial
recurrent vessels and subperiosteally dis-
If there is concern about the previous skin inci- secting the supinator off the radius, with
sion(s), a fresh incision can be made through protection of the posterior interosseous
virgin tissue. Potential reasons for doing this nerve.
would be that you did not perform the index 11. The volar approach to the radius allows
procedure, and based on the skin incision, there exposure of the entire diaphysis. One must
is potential for damage to important structures. be aware of and protect the radial artery, the
Otherwise, the previous incision and/or incisions superficial sensory branch of the radial
can be utilized. nerve, and the posterior interosseous nerve
For the ulna diaphysis, the approach should be proximally, as well as the brachial artery
just dorsal to the subcutaneous border and over and median nerve.
The posterior Thompson approach can be used 4. Comfortable seating and height for the
for those nonunions that require exposure of the surgeon and assistant along with loupe
entire route of the posterior interosseous nerve. magnification
5. Draw incisions with a marker.
1. The skin incision is made along a line with 6. If two incisions are needed, allow at least a
the forearm pronated, starting at the lateral six to eight centimeter interval between the
epicondyle of the elbow and ending over incisions.
Lister’s tubercle. 7. The secret to soft tissue dissection is ade-
2. The skin is incised, and a moist lap sponge is quate tension on the tissues in the correct
used to dissect the subcutaneous tissue off the vector.
fascia. 8. Dissect from normal to abnormal tissue.
3. The interval between the extensor digitorum Never seek to identify structures in scar
comminus and the extensor carpi radialis tissue.
brevis is more easily found distally. 9. Dissect with the tips of your scissors.
4. In large individuals, you can use the bovie to 10. Scissors work best in normal tissue, and a
stimulate the muscle bellies proximally and scalpel is needed in scar tissue.
easily separate the extensor digitorum com- 11. Hemostasis can be obtained with clips,
minus from the extensor carpi radialis brevis. sutures, or the bovie. To avoid any intimal
5. The dissection through the muscle bellies is damage to the artery, bovie at least one
more easily done from distal to proximal. centimeter away from the artery.
6. The glistening fascia over the supinator is 12. Most exposures in normal tissue can be
easily identified, and the distal border of the done by dividing fascia and mesentery,
muscle is the anatomical point where the avoiding proximity to major nerves and
posterior interosseous nerve arborizes. vessels.
7. Prior to branching, the posterior interosseous 13. Place retractors appropriately and remember
nerve lies between the two muscle layers of “the bone and periosteum are your friends.”
the supinator accompanied by its artery and 14. Keep tissues moist.
vein. It can be easily found and freed up to 15. If, after adequate exposure, there is concern
the radial head. about bleeding, let your tourniquet down to
8. The supinator then can be easily elevated off control it. It is often easier to ligate vessels
the proximal radius. that may be difficult to reach if the fracture
is not stabilized with a plate. After
Pitfalls occur when the proper interval is not hemostasis is obtained, you can reinflate the
recognized, and denervation of the extensor tourniquet and place the fixation.
digitorum comminus can occur. Vigorous
retraction of the posterior interosseous nerve can
result in a posterior interosseous nerve palsy.
6.4.5 Bone Preparation
6.4.4 Essentials of Exposure The surgeon should avoid extensive subperiostal

stripping. The periosteal elevator should be used
1. Adequate draping to allow full exposure of against the acute angle of the muscle attach-
the limb ments. Once the fracture site is exposed and the
2. Appropriate functioning tourniquet and plate(s), if any, removed, use bone hooks to bring
equipment the bone to you. Meticulous attention should be
3. Functioning C-arm and easily available placed upon preserving the soft tissues and blood
screens supply.
Technique for bone preparation: 6.5 Patient Variables
1. Curette the screw holes. There are many factors that are patient-dependent
2. Using a rongeur, debride the bone ends back and they include:
to bleeding bone at fracture site.
3. Reconstitute the medulary canal in both 1. Smoking
fragments to allow ingress of pluripotent 2. Adequate control of comorbidities: diabetes,
cells. hypertension, cardiac disease, liver disease,
4. Use an osteotome to “rose petal” the cortical chronic obstructive pulmonary disease, and
bone for one inch on both sides of nonunion alcoholism.
site [32]. 3. Preoperative assessment of nutritional status
5. Select a plate that has at least six cortices in should be included in the postoperative
nonviolated bone on both sides of the non- course.
union site. Do not use previous drill holes. 4. Insomnia, depression, anxiety, bipolar, and
6. Use the plate to reestablish length, and if a fibromyalgia all have some influence on the
gap exists at the nonunion site, it should be fracture healing.
filled with cancellous or corticocancellous
graft. It is preferable to use autologous graft
and insert it immediately after harvesting it.
Do not let it sit on the back table. Defects up 6.6 Anticipated
to four to six centimeters should be treated Outcomes/Recurrent
with corticocancellous grafts that give struc- Nonunions
tural support. Defects larger than six cen-
timeters should be considered for Successful healing in uncomplicated nonunions
vascularized grafts [33–38]. treated appropriately should be within the 85–
7. Perform a routine closure without drains and 100% range [23, 34–39]. Failure of a nonunion
apply a bulky sterile dressing with a sugar- following appropriate surgical intervention is
tong splint. usually related to infection, hardware failure,
8. Depending upon the stability of the fixation, smoking, and chronic illness. If the nonunion is
active range of motion can be initiated at treated appropriately, these nonunions are
10 days with a removable orthosis. beyond the treating surgeon’s control. Bone
stimulators can be of some benefit, but our
experience with them is limited.
These recurrent nonunions must be evalu-
6.4.6 Technical Points ated as to the cause and can be treated with
vascularized grafts, possible plate replacement,
1. Preserve soft tissues as much as possible. and autologous grafts. Difficult nonunions,
2. Begin dissection in normal tissue infected nonunions, and nonunions with miss-
3. Keep tissues moist. ing bone can be treated with one-bone
4. Achieve meticulous hemostasis. forearm.
5. The tourniquet should be released at 60 min
and possibly reinflated if needed.
6. Antibiotics should be administered preopera- 6.7 Case Discussions
tively and postoperatively for 24 h.
7. Leave sutures in place for two weeks. Case 1 (Fig. 6.1)
8. Obtain radiographs at two months unless
otherwise indicated. A 45-year-old, right-hand dominant, day laborer,
9. The expected healing time is 6–12 months. who was a 2-pack-per-day smoker, sustained a
Fig. 6.1 a, b Anteroposterior (AP) and lateral of nonunion after failed closed treatment. c, d AP and lateral of after
repair of nonunion
direct blow to his ulnar forearm, resulting in a Case 3 (Fig. 6.3)

fracture of the shaft of the ulna. It was decided, in
conjunction with the patient, to treat his fracture A 63-year-old, left-hand dominant male had
with nonoperative methods. He subsequently end-stage renal disease as a result of poorly
went on to develop an atrophic nonunion. Once controlled diabetes mellitus. He sustained multi-
nonoperative treatment was exhausted, the ple orthopedic and general surgery injuries,
patient underwent nonunion takedown, including a proximal radius fracture that was
rose-petaling of both fracture ends, reestablishing identified 4 months after his initial accident. The
of the canals, application of a corticocancellous patient was treated by closed methods for medi-
autograft, open reduction and internal fixation cal reasons and also because the patient refused
with an interfragmentary compression screw, and surgical intervention. He was definitively man-
then the fracture was neutralized with a 3.5 mm aged in an orthosis which eventually healed at
locking compression plate The patient stopped 12 months after his accident. He was able to
smoking as well and went on to heal his non- obtain 15° of pronation and 40° of supination and
union uneventfully. had function appropriate for his daily needs.
Case 2 (Fig. 6.2) Case 4 (Fig. 6.4)
A 37-year-old, right-hand dominant male sus- A 26-year-old, right-hand dominant male sus-
tained a gunshot wound to his ulna. He refused tained a handgun gunshot wound to the forearm,
surgical intervention and was treated with closed resulting in a 3-cm bone defect of the radius. Ini-
methods. At his last follow-up, he had 55° of tially, this was plated out to length and not grafted.
supination and 35° of pronation. He continued to The fixation ultimately failed, and a nonunion
refuse surgical interventions, stating that he had ensued. He also had a significant decline in his
adequate function and eventually was lost to forearm rotation, resulting in 25° of supination and
follow-up. 35° of pronation. The patient was taken back to the
Fig. 6.2 a, b Anteroposterior (AP) and lateral after gunshot wound. c, d AP and lateral with established
“asymptomatic” nonunion
operating room, a corticocancellous graft was resulting 4-cm bone defect. The patient was
harvested from the patient’s ipsilateral iliac crest, treated initially with irrigation and debridement
and a repeat osteosynthesis was performed. The of devitalized bone and soft tissue, application of
patient went on to heal, and his supination an antibiotic-impregnated cement spacer, and
improved to 55° and pronation to 50°. plate fixation. The patient was lost to follow-up
and returned a year later. The construct had
Case 5 (Fig. 6.5) ultimately failed and the patient had decreased
function with 45° of supination and 55° of
A 34-year-old, left-hand dominant male sus- pronation. He desired to have repeated surgical
tained a gunshot wound to the radius with a treatment and returned to the operating room for
Fig. 6.3 a, b Anteroposterior (AP) and lateral of delayed presentation of right proximal radius fracture. c, d AP and
lateral at 1 year with healed proximal radius fracture despite a delayed presentation
a joint intervention with plastic surgery. A vas- extremity amputations and open fractures of his
cularized fibula was harvested and then insetted. right radius and ulna and concomitant massive
He subsequently went on to heal successfully soft tissue injury to the right forearm. The
with the vascularized fibular graft. His supination patient’s forearm fractures were initially treated
improved to 60°, and pronation stayed the same with a temporary external fixator. Once the
at 55°. patient was stable medically, he returned to the
operating room, where the external fixator was
Case 6 (Fig. 6.6) removed, the radius was plated out to length, and
an antibiotic-impregnated cement spacer was
A 39-year-old, right-hand dominant sustained inserted. The definitive plan was to take the
multiple traumatic injuries, including lower patient back to the operating room to remove the
Fig. 6.4 a, b Anteroposterior (AP) and lateral after f AP and lateral after repeat osteosynthesis and iliac crest
initial plating. Length was re-established. c, d AP and bone graft. g, h AP and lateral after healing of nonunion
lateral with nonunion of radius and failure of hardware. e,
cement spacer and insert a corticocancellous procedure. At that point, an intra-operative

graft; however, a radioulnar synostosis began to fluoroscopic stress test was performed, which
form. The bridging wrist plate was removed. The confirmed that the radioulnar synostosis had
patient continued with symptomatic ulna hard- matured and the decision was made to leave the
ware and ulnar-sided wrist pain, necessitating cement spacer in place as the patient had essen-
hardware removal from the ulna and a Darrach tially developed a one-bone forearm.
Fig. 6.5 a, b Anteroposterior (AP) and lateral of initial post-op AP and lateral showing complete healing of c
gunshot wound to right radius. c, d Immediate post-op AP nonunion with complete incorporation of vascularized
and lateral of revision open reduction internal fixation fibula
with vascularized fibular strut and plate. e, f 11-month
b Fig. 6.6 a, b Anteroposterior (AP) and lateral at initial in radial defect. g Follow-up radiograph showing devel-
injury showing obvious soft tissue defect and significant opment of radioulnar synostosis. (h, i) AP and lateral of
bony injury. c, d AP and lateral showing temporizing subsequent removal of hardware from ulna and Darrach
external fixation and the significant bone loss from the procedure. j, k Final follow-up AP and lateral with healed
injury and after debridement. e, f AP and lateral after ulna, stable radius with retained cement and establish-
“bridging” internal fixation of wrist and open reduction ment of radioulnar synostosis
internal fixation of radius and ulna. Cement spacer placed
Fig. 6.7 a Clinical picture of left forearm-infected one-bone forearm with proximal ulna plated to distal
nonunion. b Clinical picture with splint application radial shaft. d Healed one-bone forearm after plate
showing correction of deformity. c Creation of a healed removal secondary to hardware irritation
Case 7 (Fig. 6.7) institution for evaluation. He was taken to the

operating room for evaluation and operative
A 43-year-old, left-hand dominant male day treatment. His nonunion site was taken down,
laborer sustained open fractures of his left radius and the distal radius was inserted into the prox-
and ulna in a work-related accident. This was imal ulna, and then, the fracture was plated. In
treated nonoperatively and went on to become an addition, he was treated with appropriate antibi-
infected nonunion. His treating physicians at that otic therapy and his fracture subsequently healed.
time continued to treat him nonoperatively, and He returned 2 years later with complaints of
the patient was able to work with his homemade hardware irritation, and the plate was then
splint. He subsequently presented to our removed (see Fig. 6.7d).
References interosseous membrane of the forearm: pathome-

chanics of proximal migration of the radius. J Hand
Surg Am. 1989;14(2 Pt 1):256–61.
1. McGinley JC, Kozin SH. Interosseous membrane 17. Poitevin LA. Anatomy and biomechanics of the
anatomy and functional mechanics. Clin interosseous membrane: its importance in the longi-
Orthop. 2001;383:108–22. tudinal stability of the forearm. Hand Clin. 2001;17
2. Ward LD, Ambrose CG, Masson MV, Levaro F. The (1):97–110, vii.
role of the distal radioulnar ligaments, interosseous 18. Nakamura T, Yabe Y, Horiuchi Y. Functional
membrane, and joint capsule in distal radioulnar joint anatomy of the interosseous membrane of the
stability. J Hand Surg Am. 2000;25(2):341–51. forearm—dynamic changes during rotation. Hand
3. Watanabe H, Berger RA, Berglund LJ, Zobitz ME, Surg. 1999;4(1):67–73.
An KN. Contribution of the interosseous membrane 19. Noda K, Goto A, Murase T, Sugamoto K,
to distal radioulnar joint constraint. J Hand Surg Am. Yoshikawa H, Moritomo H. Interosseous membrane
2005;30(6):1164–71. of the forearm: an anatomical study of ligament
4. Kihara H, Short WH, Werner FW, Fortino MD, attachment locations. J Hand Surg Am. 2009;34
Palmer AK. The stabilizing mechanism of the distal (3):415–22.
radioulnar joint during pronation and supination. 20. Skahen JR 3rd, Palmer AK, Werner FW,
J Hand Surg Am. 1995;20(6):930–6. Fortino MD. The interosseous membrane of the
5. Soubeyrand M, Wassermann V, Hirsch C, Oberlin C, forearm: anatomy and function. J Hand Surg Am.
Gagey O, Dumontier C. The middle radioulnar joint 1997;22(6):981–5.
and triarticular forearm complex. J Hand Surg Eur. 21. Schneiderman G, Meldrum RD, Bloebaum RD,
2011;36(6):447–54. Tarr R, Sarmiento A. The interosseous membrane
6. Morrey BF, Askew LJ, Chao EY. A biomechanical of the forearm: structure and its role in Galeazzi
study of normal functional elbow motion. J Bone fractures. J Trauma. 1993;35(6):879–85.
Joint Surg Am. 1981;63(6):872–7. 22. Chapman MW, Gordon JE, Zissimos AG.
7. Tarr RR, Garfinkel AI, Sarmiento A. The effects of Compression-plate fixation of acute fractures of the
angular and rotational deformities of both bones of diaphyses of the radius and ulna. J Bone Joint Surg
the forearm. An in vitro study. J Bone Joint Surg Am. Am. 1989;71(2):159–69.
1984;66(1):65–70. 23. Stewart RL. Operative treatment of radius and ulna
8. Youm Y, Dryer RF, Thambyrajah K, Flatt AE, diaphyseal nonunions. In: Wiesel SW, editor. Oper-
Sprague BL. Biomechanical analyses of forearm ative techniques in orthopaedic surgery. 1st ed.
pronation-supination and elbow flexion-extension. Philadelphia, PA: Lippincott Williams & Wilkins;
J Biomech. 1979;12(4):245–55. 2011. p. 2156–61.
9. Jinkins WJ Jr, Lockhart LD, Eggers GW. Frac- 24. Maes JM, Van Velthoven V. Clinical and statistical
tures of the forearm in adults. South research on the occurrence of pseudarthrosis of ulna
Med J. 1960;53:669–79. and radius in forearm fractures. Acta Orthop Belg.
10. Burwell HN, Charnley AD. Treatment of forearm 1977;43(6):767–80.
fractures in adults with particular reference to plate 25. Knight RA, Purvis GD. Fractures of both bones of
fixation. J Bone Joint Surg Br. 1964;46:404–25. the forearm in adults. J Bone Joint Surg Am.
11. Sage FP, Smith H. Medullary fixation of forearm 1949;31A(4):755–64.
fractures. J Bone Joint Surg Am. 1957;39-A(1):91–8. 26. Sage FP. Medullary fixation of fractures of the
12. Schemitsch EH, Richards RR. The effect of malunion forearm. A study of the medullary canal of the radius
on functional outcome after plate fixation of fractures and a report of fifty fractures of the radius treated
of both bones of the forearm in adults. J Bone Joint with a prebent triangular nail. J Bone Joint Surg Am.
Surg Am. 1992;74(7):1068–78. 1959;41-A:1489–516.
13. Boland, MR. Open reduction and internal fixation of 27. Hughston JC. Fracture of the distal radial shaft;
diaphyseal forearm fractures. In: Wiesel, SW, editor. mistakes in management. J Bone Joint Surg Am.
Operative techniques in orthopaedic surgery. 1st ed. 1957;39-A(2):249–264; passim.
Philadelphia, PA: Lippincott Williams and Wilkins; 28. Caden JG. Internal fixation of fractures of the forearm.
2011. p. 2127–9. J Bone Joint Surg Am. 1961;43(8):1115–21.
14. Llusa Perez M, Meri Vived A, Ruano Gil D. Surgical 29. Hicks JH. Fractures of the forearm treated by rigid
atlas of the musculoskelatal system. 1st ed. Rose- fixation. J Bone Joint Surg Br. 1961;43-B:680–7.
mont, IL: American Academy of Orthopaedic Sur- 30. Sargent JP, Teipner WA. Treatment of forearm shaft
geons; 2008. p. 61–5. fractures by double-plating; a preliminary report.
15. Huang JI, Hanel DP. Anatomy and biomechanics of J Bone Joint Surg Am. 1965;47(8):1475–90.
the distal radioulnar joint. Hand Clin. 2012;28 31. Anderson LD, Sisk D, Tooms RE, Park WI 3rd.
(2):157–63. Compression-plate fixation in acute diaphyseal frac-
16. Hotchkiss RN, An KN, Sowa DT, Basta S, Wei- tures of the radius and ulna. J Bone Joint Surg Am.
land AJ. An anatomic and mechanical study of the 1975;57(3):287–97.
32. Richard MJ, Ruch DS, Aldridge JM 3rd. Malunions 36. Wood MB. Upper extremity reconstruction by vas-
and nonunions of the forearm. Hand Clin. 2007;23 cularized bone transfers: results and complications.
(2):235–243, vii. J Hand Surg Am. 1987;12(3):422–7.
33. Shelton WR, Sage FP. Modified Nicoll-graft treat- 37. Jupiter JB, Gerhard HJ, Guerrero J, Nunley JA,
ment of gap non-unions in the upper extremity. Levin LS. Treatment of segmental defects of the
J Bone Joint Surg Am. 1981;63(2):226–31. radius with use of the vascularized osteoseptocuta-
34. Moroni A, Rollo G, Guzzardella M, Zinghi G. Surgical neous fibular autogenous graft. J Bone Joint Surg
treatment of isolated forearm non-union with segmen- Am. 1997;79(4):542–50.
tal bone loss. Injury. 1997;28(8):497–504. 38. Dell PC, Sheppard JE. Vascularized bone grafts in
35. Ring D, Allende C, Jafarnia K, Allende BT, the treatment of infected forearm nonunions. J Hand
Jupiter JB. Ununited diaphyseal forearm fractures Surg Am. 1984;9(5):653–8.
with segmental defects: plate fixation and autogenous 39. Davey PA, Simonis RB. Modification of the Nicoll
cancellous bone-grafting. J Bone Joint Surg Am. bone-grafting technique for nonunion of the radius
2004;86-A(11):2440–5. and/or ulna. J Bone Joint Surg Br. 2002;84(1):30–3.
Nonunions of the Wrist and Hand
7
Matthew Lyons, MD, Ahmad Fashandi, MD
and Aaron M. Freilich, MD
7.1 Introduction 7.2 Distal Radius Fracture

Nonunion
Nonunions involving the hand and wrist can
occur after injury and result in continued pain, 7.2.1 Background
weakness, and delay in recovery of function. If
unrecognized and undertreated, they can be a Distal radius fractures are the most common
source of frustration for both the patient and the injury of the upper extremity. In the literature,
treating clinician. Injury to any of the bones in management of these fractures has commonly
the hand and wrist can be result in nonunion; focused on restoring key parameters of skeletal
however, some, such as scaphoid and hamate, as alignment, rather than obtaining consolidation of
a consequence of their unique vascularity, are the fracture itself [1–3]. This is due to the fact
more likely to be involved than others. Depend- that treatment of these injuries rarely results in
ing on which bone is involved, different strate- fracture nonunion. The low incidence of non-
gies for treatment need to be adopted. In all union is often attributed to the excellent
cases, the underlying cause of the nonunion intraosseous blood supply of the distal radius,
should be assessed. This can be a problem with healing potential of cancellous metaphyseal
soft tissue coverage, vascularity, motion at the bone, impaction of fracture fragments, and min-
fracture, infection, or metabolic causes. In the imal soft tissue disruption caused by the typical
carpal bones of the wrist, osteonecrosis becomes mechanism of a low-energy fall [1]. Nonunion is
an important complicating factor that needs to be more common in the setting of attempted surgi-
addressed. In many patients, obtaining healing of cal fixation than closed management and appears
the fracture will be the goal; however, salvage to be increasing over the past two decades [1, 2,
procedures such as arthrodesis or even amputa- 4]. A potential causative factor for this increase is
tion should be considered depending on the advent of fixed angle volar plating, which has
circumstances. allowed surgeons to restore and maintain radial
length in the setting of high-energy injuries with
extensive comminution, resulting in gap creation
in the metaphyseal bone [1, 5]. Other risk factors
include unstable fracture patterns with inade-
M. Lyons A. Fashandi A.M. Freilich (&) quate fixation, concomitant fracture of the distal
Department of Orthopedics, University of Virginia, ulna, open comminuted fractures, infections,
P.O. Box 800159, Charlottesville, VA 22908-0159,
USA pathologic lesions, and inadequate period of
e-mail: amf7z@virginia.edu immobilization [1, 2, 4]. Whatever the cause,

144 M. Lyons, MD et al.
nonunions of the distal radius often result in radiograph by drawing two lines tangential to the
significant pain, stiffness, and dysfunction, which radial styloid and ulnar head articular surface and
necessitate surgical stabilization. Treatment of perpendicular to their shaft axes. The distance
these injuries is aimed at obtaining bony stability between the lines is measured. The average nor-
and restoring alignment and can be difficult due mal value is 11 mm with an acceptable limit of
to small, osteopenic bone fragments, bony reab- 4 mm. Radial inclination, also measured on a PA
sorption at the fracture site, and associated soft radiograph, is the angle formed by one line per-
tissue contracture [1]. pendicular to the longitudinal axis of the radial
shaft and a second line along the distal radius
articular surface. The normal value for this angle
7.2.2 Anatomy is 22° with an acceptable change of 15° in either
direction. Measured on the lateral radiograph,
The failure to recognize key fracture components radial tilt is the angle between the distal radius
that predict fracture instability is a risk factor for articular surface and a line perpendicular to the
nonunion with both conservative and operative longitudinal axis of the radial shaft. It has a nor-
management of these injuries. Therefore, an mal value of 11° of volar tilt with an acceptable
understanding of the anatomic alignment and limit of 15° of dorsal or 20° of volar tilt. Ulnar
relationships between the distal radius and variance is calculated on the PA radiograph by the
radiocarpal articulation is paramount to the axial difference in length from lines drawn par-
treatment of distal radius fractures. The distal allel to the ulnar head articular surface and ulnar
radius forms an articular platform on which the most edge of articular distal radius. It averages
carpus rests. The stability provided by the neutral to 1 mm of ulnar negative variance with
radiocarpal articulation and its surrounding liga- an acceptable limit of ±4 mm [6, 7]. Articular
ments allows the complex functions performed incongruity is determined by breaks in the dense
by the carpus and hand. The distal radius has subchondral bone on the PA radiograph, less than
three concave surfaces, which form the founda- 2 mm of articular gap or step-off is acceptable. If
tions of this articulation, the scaphoid fossa, these parameters are met and able to be main-
lunate fossa, and sigmoid notch. The scaphoid tained by either closed or open means, the out-
and lunate fossas are divided by a sagittal plane come of treatment will be determined by the
ridge. The radiocarpal articulation is further sta- degree of soft tissue injury rather than skeletal
bilized by the strong volar radial ligamentous deformity.
structures, including the radioscaphocapitate
(RSC), radiolunotriquetral (RLT), radioscaphol-
unate (RSL), and dorsal radiotriquetral 7.2.3 Classification Systems
(RT) ligaments. The sigmoid notch acts an
articulation for the distal ulna, allowing forearm Multiple classification systems have been pro-
motion through rotation of the radius around the posed to describe distal radius injuries patterns,
ulna. It has well-defined dorsal, volar, and distal each with its own inherent strengths and weak-
walls, with further stability provided by the nesses. The goal of any surgeon taking care of
components of the triangular fibrocartilage these injuries is to become familiar enough with
complex (TFCC), including deep and superficial common injury patterns as to recognize the per-
volar and dorsal radioulnar ligaments [5]. sonality of the individual fracture [3]. This is a
Key radiographic parameters predict stability combination of the energy, deformity, com-
and dictate treatment of the distal radius. These minution, soft tissue disruption, and stability
include radial height, radial inclination, radial tilt, inherent in the fracture pattern that are predictive
ulnar variance, articular congruity, and distal of outcome and help determine treatment.
radioulnar joint (DRUJ) stability. Radial height is Given that the majority of distal radius frac-
measured on the posterioranterior (PA) ture nonunions involve high-energy injuries and
7 Nonunions of the Wrist and Hand 145
intra-articular fracture patterns, we find the latent nonunion period between 3 and 6 months
columnar classification proposed by Rikli and after injury may complain of only stiffness, as
Regazzoni and the fragment-specific system of they have experienced insignificant time or
Medoff most useful in approaching these injuries activity level to develop other symptoms.
[8–10]. Rikli and Regazzoni’s classification Patients should be questioned about potential risk
system divides the wrist into three columns. The factors for nonunion, including advanced age,
radial column is composed of the radial styloid medical comorbidities, smoking, nonsteroidal
and scaphoid facet of the distal radius. Realign- anti-inflammatory use, metabolic disease, and
ment of this column effectively restores radial nutritional deficiency. A complete medical his-
height and inclination. The intermediate column tory should be elicited, with particular attention
contains the primary load bearing portion of the given to the patient’s occupational demands,
distal radius and is encompassed by the lunate recreational activities, and goals for treatment.
facet of the distal radius. Fractures of this column The preoperative evaluation of nonunion should
may also disrupt the sigmoid notch and distal include a characterization of biologic healing
radioulnar articulation. Finally, the medial col- capacity, deformity, presence or absence of
umn is composed of the distal ulna, TFCC, and infection, and host status. Correction of modifi-
radioulnar ligaments. Injuries to the medial col- able risk factors, such as tobacco cessation and
umn may result in DRUJ instability. The integ- nutritional status, should be attempted prior to
rity and stability of the DRUJ should be assessed surgical intervention [12]. As signs and symp-
in the setting of every distal radius fracture, and toms of infection can be subtle, we recommend
the treatment of DRUJ instability will be covered obtaining screening laboratories, including com-
later in this chapter. plete blood count (CBC), erythrocyte sedimen-
The fragment-specific classification system tation rate (ESR), and C reactive protein (CRP).
proposed by Medoff is a direct extension of the The physical examination may be difficult in
Melone Classification. Originally proposed in the setting of a nonunion. It should attempt to
1984, Melone’s system divides the distal radius focus on the strength, range of motion, and sta-
into four components: the shaft, radial styloid, bility of the upper extremity from the shoulder to
and volar medial and dorsal medial facets [11]. the digits utilizing the unaffected side for com-
Medoff expanded this system to include five parison whenever possible. Wrist range of
major fracture fragments, including the radial motion, including flexion, extension, pronation,
styloid, dorsal wall, impacted articular segment, and supination, should be tested. A complete
dorsal ulnar corner, and volar rim [9, 10]. The motor and sensory examination should be per-
system provides a treatment algorithm for formed. When combined with provocative tests
fragment-specific fixation that can be useful with for carpal tunnel syndrome, it can expose an
obtaining stability in the setting of a nonunion. underlying median nerve injury. Skin inspection
should be performed for any previous surgical
incisions, which may have an effect on the choice
7.2.4 Clinical Evaluation for later surgical approach. Also, given the
association of distal radius fractures with com-
As with any initial assessment, clinical evalua- plex regional pain syndrome, special attention
tion should begin with a detailed history of the should be given to disproportionate pain, finger
patient’s injury mechanism and any attempted stiffness, swelling, allodynia, or paresthesias.
treatment. Any previous attempt at surgical When the patient describes a history of instabil-
treatment should be elicited as it may play a role ity, stress tests should seek to localize it to the
in preoperative planning. Patients frequently radiocarpal, midcarpal, or DRUJs.
present with a combination of pain and dys- Although TFCC injuries commonly occur with
function, related to deformity and instability. It distal radius fractures, radiocarpal or midcarpal
should be noted that a number of patients in the instability is extremely rare. Increased anterior to
Fig. 7.1 Computed

tomography evaluation of
distal radius nonunion after
attempted fixation and
allograft bone grafting for
comminuted open fracture
posterior translation of the ulna on the radius comparison views of the contralateral wrist can
compared to the contralateral side can be aid with preoperative planning. While the
indicative of DRUJ instability. Finally, an Allen radiographic parameters serve as a useful
test should be performed to assess the specific benchmark in treatment, each patient’s symp-
vascular supply crossing the zone of injury [13]. toms and function should be taken in account
Radiographic examination constitutes the when determining a treatment course.
second core component of the clinical examina- The role of advanced imaging is difficult to
tion. All previous radiographs should be obtained define and should be considered on a
and reviewed. A current series of wrist radio- case-by-case basis. Rotational deformity and
graphs should be obtained, including PA, lateral, articular congruity are often difficult to assess on
and oblique views. High-quality, appropriately plain radiographs, and computed tomography
aligned radiographs allow measurement of the (CT) can be a useful adjunct. Three-dimensional
key parameters outlined earlier that portend reconstruction images provide information on
fracture stability and help determine surgical both articular displacement and axial plane
approach. These include radial height, radial deformities (Fig. 7.1). CT also aids in assessment
inclination, radial tilt, ulnar variance, articular of bone reabsorption and osteopenia, which may
congruity, and distal radial ulnar joint stability [6, influence fixation method.
7]. Radiographs may reveal persistent fracture
lines, scalloping, or bone reabsorption in the
distal segment, as well as loose or broken hard- 7.2.5 Treatment
ware. Wrist flexion and extension radiographs
may provide additional benefit by revealing Given the rarity of distal radius nonunions, it is
fracture site motion. Hypertrophic, oligotrophic, not possible to provide treatment recommenda-
and atrophic radiographic appearance allows the tions based on statistical analysis. Instead, each
clinician to make inferences about the degree of patient should be considered individually, with
fracture stability, the presence of infection, and the appropriate treatment determined by the
the biologic viability of the fracture fragments fracture personality, patient demands, and pres-
prior to developing a treatment plan. The presence or absence of infection. Nonoperative
ence of an atrophic nonunion should raise con- treatment is typically appropriate only in the
cern for an underlying infectious process. Also, setting of elderly patients with very low
functional demands. Historically, the most com- and limited options for fracture fixation. The
mon treatment of nonunion was with wrist majority of these injuries will present after a
arthrodesis. However, advances in implant tech- failed attempt at previous surgical fixation. In this
nology, such as fixed angle volar locking plates, setting, the orthogonal approach to plating as
have improved fracture union rates. Open outlined by Fernandez, Ring, and Jupiter can
reduction and internal fixation may have a sig- prove useful [1, 16]. The use of a dual plating
nificantly positive impact on upper extremity technique with fixed angle locking plates pro-
function through the resulting preservation of vides a greater number of fixation points in the
wrist motion [1, 14, 15]. Similar to malunited distal segment and greater stability in the setting
distal radius fractures, patients with nonunion of atrophic or osteopenic bone (Fig. 7.2). Using
often experience activity limiting pain and this technique, the distal radius is approached
instability, with the majority of problems arising through a volar or combined volar and dorsal
from fracture malalignment. In the setting of approach, depending on the type of deformity
acute distal radius fractures, the radiographic present. The hybrid volar approach proposed by
parameters of radial height, radial inclination, Chhabra et al. can also be utilized to release the
radial tilt, ulnar variance, and articular congruity carpal tunnel if median nerve symptoms are pre-
are used to correct malalignment. These same sent preoperatively or if substantial deformity
parameters are useful in the treatment of non- correction or soft tissue contracture may place the
union. The goal of treatment should be to provide median nerve at risk postoperatively [17]. The
acceptable, stable fracture alignment with a soft fracture ends are identified, fibrous tissue and
tissue envelope devoid of infection and a bio- synovial membrane are removed, and sclerotic or
logic environment that is capable of fracture necrotic bone is resected. The medullary canal of
healing. each end of the metaphyseal component of the
As discussed earlier in the chapter, failure of fracture is opened to facilitate intraosseous
healing of metaphyseal distal radius fractures is ingress of blood and growth factors for fracture
rare. The definition of delayed union and nonhealing. Release or z-lengthening of the bra-
union of the distal radius is not clearly defined in chioradialis and flexor carpi radialis tendons may
the literature. While the severity of injury plays be required to correct loss of radial height and
an obvious role in the rate of healing, one would inclination. An external fixator with 2.5-mm
expect to see radiographic evidence of progres- Schanz screws is often used to aid with reduc-
sive healing within 3 months after initial injury. tion and maintain alignment prior to internal fix-
Even in fractures with extensive comminution, ation. Care should be taken to not place Schanz
one would expect evidence of healing after pins where they may interfere with plate fixation.
4 months. When there is lack of progressive Kirschner wires are used to stabilize individual
healing in the latent nonunion period between 3 fracture fragments. Using the Rikli and Regaz-
and 6 months from injury or initial treatment, zoni columnar classification, one plate is placed
continued immobilization and limb disuse are on the radial styloid and lateral radius to control
likely to have a detrimental effect on function and the lateral column, while a second plate is placed
range of motion. Further conservative treatment on the volar or dorsal cortex to stabilize the
also fails to address the primary problem if intermediate column [8]. We recommend initial
malalignment or atrophic changes are present on fixation of the more stable column to the radial
radiographs. Therefore, we recommend a low shaft, as this may aid with determination of
threshold to surgical intervention during this time alignment for the more complex or comminuted
period. column. Autogenous cancellous bone graft is
Surgical treatment of nonunited distal radius used to pack the defect. Following fracture fixa-
fractures with multiple fracture fragments is tion, the distal radial ulnar joint should be asses-
challenging due the osteopenic quality of the sed for congruency and arthrosis. If uncorrectable
bone, potential presence of fracture reabsorption, malalignment or advance arthritic changes of the
Fig. 7.2 Volar and dorsal

dual plating of a distal radius
fracture
sigmoid notch–distal ulna articulation are present, Given that the majority of distal radius frac-
a salvage procedure such as a Darrach distal ulnar ture nonunions have undergone previous surgical
resection or Bowers hemiresection interposition intervention, failure to adequately address pat-
arthroplasty may be required. terns of fracture instability should be considered
Ring and Jupiter reported on a total of 23 in the preoperative evaluation. The most com-
patients treated with this technique over a mon example of this is the failure to recognize
10-year period [18]. At an average of 28-month and stabilize the volar ulnar corner fragment.
follow-up, all but one had gone on to a successful While the advance in fixed angle volar plating
union. The one persistent nonunion was eventu- techniques has overall improved fracture fixation
ally treated with a wrist fusion. Two patients and allowed earlier return of function, it can be
were treated with DRUJ salvage procedures difficult to maintain reduction in complex
(Darrach or Bowers). While patients demon- intra-articular fractures with a volar ulnar corner
strated significant improvement in range of or rim fracture of the distal radius. Stability of the
motion and function, they never regained normal volar ulnar corner is critical to providing struc-
motion and only 7 of 23 had good or excellent tural support to the carpus and failure to maintain
results, according to the rating system of Fer- reduction leads to volar carpal subluxation or
nandez. In another group of 10 patients, Fer- dislocation and catastrophic effects on wrist
nandez et al. reported similar results [1]. All 10 function [19–21]. The fragment-specific fixation
successfully healed their fractures. Distal ulnar method proposed by Medoff allows use of two or
salvage procedures were performed in 4 patients. more low-profile implants to strategically capture
Average wrist flexion was 50°, wrist extension specific fracture fragments [10]. This technique
was 55°, and pronation and supination were 70 may be useful in the setting of a latent nonunion
and 75°. According to the Fernandez functional before osteopenia and fracture reabsorption have
result system, there were 3 excellent, 4 good, 2 developed. Newer fixed angle, low-profile hook
fair, and 1 poor result. plates may allow more stable fixation of the volar
rim in osteoporotic bone, as long as significant contact, dynamic compression locking plate is
metaphyseal comminution is not present [22]. In used. Blunt dissection of the proximal incision
the setting of significant metadiaphyseal com- between the brachioradialis and second dorsal
minution or fracture reabsorption, the use of compartment tendons is performed until the
low-profile implants is contraindicated, but the dorsal distal radius is exposed. Care should be
principles of fracture-specific fixation remain the taken to avoid damage to the superficial sensory
same and control of the volar-ulnar fragment is nerve, which emerges from deep to the
critical for stabilization of the carpus. brachioradialis.
Dorsal distraction bridge plating is a useful If the second metacarpal is chosen, the plate
technique in nonunions with extensive com- will be placed in the second dorsal compartment,
minution and bone loss [23, 24]. Segelman and and if the third metacarpal is chosen, the plate
Clark [2] have suggested that union may not be will be placed deep to the tendons of the fourth
possible if less than 5 mm of subchondral bone is compartment. A Freer or Cobb elevator can be
present beneath the lunate facet, as there is used to create a path from the distal to the
inadequate space available for implant fixation. proximal incisions, and the plate is slide from
Dorsal distraction plating alleviates this problem distal to proximal, avoiding impingement of the
and allows for both correction of severe radial extensors. A cortical screw is placed in the center
shortening and bridging of osteopenic metaphy- hole of the plate distally to stabilize it to the
seal bone (Fig. 7.3). An initial 4-cm dorsal metacarpal. Fracture reduction is performed with
incision is made over the second or third meta- longitudinal traction to restore length, palmar
carpal, and the extensor tendon is retracted. translation of the hand relative to the forearm to
Choice of the second or third metacarpal is a correct radial tilt, and pronation of the hand rel-
matter of surgeon preference and may be influ- ative to the radius to counteract supination of the
enced by individual fracture characteristics. carpus. Once reduction is confirmed, a second
A 4-cm second dorsal incision is centered over cortical screw is placed in the middle hole of the
the distal radius, at least 4 cm proximal to the proximal end of the plate. The remaining holes
level of fracture comminution. Fluoroscopic are filled with locking screws. Fluoroscopy and
superimposition of the dorsal plate can aid with physical examination are used to rule out over
plate selection and positioning of the proximal distraction. The radiocarpal space should not be
incision. Typically, a 12 or 14 hole limited greater than 5 mm, and full passive finger flexion
Fig. 7.3 Dorsal distraction

(bridge) plating of a distal
radius nonunion
should be present. Supplemental fixation, motion results in a significant improvement in

including Kirschner wires and fragment-specific upper extremity function, it is now more com-
plates, can be used to attempt to improve align- monly employed as a salvage procedure. It is
ment and articular congruency. Iliac crest auto- effective a achieving bone stability when recon-
genous bone graft can be used to fill defects. struction is either not possible or not recom-
Following surgery, patients are placed in a short mended. Potential indications include advanced
arm splint and digit motion is initiated immedi- radiocarpal or midcarpal degenerative changes
ately. The plate is removed at the time of radio- and extensive bone loss or reabsorption.
graphic fracture consolidation, and wrist range Management of infection in the setting of
motion and progressive strengthening are started. nonunion adds an additional layer of complexity.
Mithani et al. [23] evaluated a total of 8 A low threshold should be maintained for the
patients treated with this technique, reporting suspicion of infection, especially with the
healing in all, with plate removal at an average of appearance of an atrophic malunion on radio-
148 days from surgery. Patients reported statis- graphs. Screening white blood cell count, CRP,
tically significant improvement in range of and ESR are recommended in all patients. If
motion and DASH scores from preoperative elevated, CRP and ESR are both independently
values. Despite successful healing, this technique predictive of infection, and the likelihood of
has some potential complications. Patients must infection increases with each additional positive
be instructed that plate immobilization for 5– test [25]. Radionucleotide bone scans have been
6 months results in a significant loss of wrist recommended as an additional screening tool for
range of motion, which requires committed and infection, but they are not cost-effective and do
prolonged therapy to recover. Additionally, over not increase the predictive value in the setting of
distraction of the wrist can lead to digit con- positive laboratory values [25]. Intra-operative
tracture, radial nerve neuritis, and complex tissue cultures from the nonunion site and any
regional pain syndrome. Finally, there is a risk of associated purulence can be obtained to provide a
tendon irritation and even tendon rupture if the definitive diagnosis and aid with antibiotic ther-
plate is placed in a position that causes tendon apy. There is no literature to guide whether
entrapment. infections of the distal radius with retained
Wrist arthrodesis was historically recom- hardware can be treated with a single operation
mended as first-line treatment for distal radius or require a staged approach. In the setting of
fracture nonunion (Fig. 7.4) [16, 17]. However, gross contamination, eradication of the infection
since preservation of even minimal wrist range of is advisable prior to proceeding with definitive
Fig. 7.4 a Chronic nonunion of distal radius fracture after volar plating. b Treated with hardware removal and wrist
arthrodesis
fixation. An external fixator can be useful in recovery of full range of motion and strength is
correcting and maintaining alignment, while unlikely.
allowing for repeat surgical debridement. Treat-
ment of an infected distal radius nonunion should
be viewed as a collaboration between the hand 7.3 Distal Radioulnar Joint
surgeon and an infectious disease team well Instability
experienced in the care of orthopedic infections.
Autogenous cancellous bone graft is used to 7.3.1 Background
improve biology and fill gaps created by fracture
realignment. The distal ulna, olecranon, and iliac The DRUJ is a diarthrodial articulation that acts as
crest can all be used a potential donor sites, the distal stabilizing structure between the radius
depending on the amount of graft needed. In the and ulna. It functions as a pivot point, allowing
setting of a large segmental defect, tricortical the radius to rotate round the ulna in supination
iliac crest graft can be harvested, but with and pronation. The ligamentous structures, which
increased donor site morbidity. Bone graft sub- confer its stability, can be injured in a mechanism
stitutes have also been developed, including bone similar to that, which produces distal radius
morphologic protein, demineralized bone matrix, fractures. This typically consists of an axial load
and synthetics, such as calcium phosphate, cal- with wrist pronation and extension, such as a fall
cium sulfate, and hydroxyapatite. While they on an outstretched wrist. As a result, DRUJ
remove the risk of donor site morbidity, all have instability can occur either in isolation or the
significantly increased cost and have not proven setting of a distal radius fracture [28].
superior to autograft bone in fracture treatment Chronic instability can occur from both a
[26, 27]. Calcium phosphate grafts have an nonunited fracture of the base of the ulnar styloid
osteoconductive potential and high compressive or a purely ligamentous injury to the TFCC and
strength and may be useful when combined with DRUJ joint capsule. In the latter setting, the
an osteoinductive substance to fill a large seg- instability can be viewed as a consequence of
mental defect. soft tissue “nonunion” of the DRUJ. Given the
frequency with which DRUJ instability accom-
panies a distal radius fracture, as high as 11% in
7.2.6 Postoperative Care one study, the management of distal radius
nonunion and malunion should include a clear
Given that many patients will have significant understanding of the evaluation and management
preoperative stiffness and dysfunction from of this injury [29].
deformity and prolonged immobilization, early
postoperative range of motion should be
emphasized, with initiation of digit and forearm 7.3.2 Anatomy and Biomechanics
range of motion on the morning after surgery.
Patients are immobilized in a postoperative short The articulation between the distal radius through
arm splint for a total of 10–14 days. After this, a the sigmoid notch and the ulnar head encom-
removable splint can be placed and rehabilitation passes the bony architecture of the DRUJ. The
is initiated, focusing on early progressive range relative asymmetry of this relationship results in
of motion. Strengthening exercises are restricted minimal conferred stability. The sigmoid notch is
until there is radiographic evidence of healing, significantly more shallow than the ulnar head,
usually 12 weeks from surgery. With more with a radius of curvature that is 50–100%
extensive bone loss, healing may be detailed. greater (15–19 mm vs. 10 mm) [28, 30]. Given
Patients should be counseled that this is a salvage the size mismatch with the ulnar head, the volar
procedure in the setting of significant preopera- and dorsal rims of the sigmoid notch must con-
tive stiffness, deformity, and dysfunction and that tribute to stability. Post-traumatic deficiencies of
either rim lead to decreased joint stability in [28]. The components of the TFCC are the volar
biomechanical testing [28, 31–34]. The dorsal and dorsal radioulnar ligaments, volar and dorsal
rim has an acute angulation, while the volar rim ulnocarpal ligaments, articular disk, meniscus
is rounded with a fibrocartilaginous lip [28]. homolog, and extensor carpi ulnar tendon sheath
There is significant variation between the coronal [28, 37]. The radioulnar ligaments contribute the
and axial alignment of the DRUJ articular sur- primary stabilizing force to the DRUJ and are
face. In the coronal plane, the joint assumes one necessary for normal stability with volar and
of three slopes relative to the long axis of the dorsal translation. Cadaveric studies have
radius and ulna: parallel, oblique, or reverse demonstrated that they maintain normal DRUJ
oblique [33]. While, at baseline, the shape has no joint kinematics after division of other soft tissue
effect on stability or function, changes in relative stabilizers [38]. Originating from the volar and
length of the radius or ulna may result in a dorsal aspects of the sigmoid notch, each liga-
mismatch between the articular surfaces and ment divides into two limbs in the coronal plane
increased contact pressure. For instance, an ulnar as they extend across the ulna. The deep
shortening osteotomy in a patient with a reverse radioulnar ligament is the more proximal of the
oblique configuration can result in loading and two as it attaches to the fovea, a shallow con-
increased contact pressure in the proximal ulnar cavity at the base of the ulnar styloid that is bare
head and sigmoid notch [35]. of cartilage and contains blood vessels that sup-
In the axial or transverse plane, there are 4 ply the TFCC. The superficial limb courses dis-
potential configurations of the sigmoid notch: flat tally, inserting on the mid-portion of the ulnar
face, ski slope, C type, and S type. The flat notch styloid [39]. The differing attachment points of
shape has the least inherent stability and is more the ligaments exert an effect on the stability of
prone to failure with soft tissue reconstruction ulnar styloid fractures. Fractures of the base of
alone in a cadaveric model [36]. The ulnar head the styloid often indicate instability of the DRUJ
is the stable unit of the DRUJ articulation, acting from detachment of the deep limbs from the
as a platform around which the radius rotates. fovea [28].
The contour of the ulnar head articular surface is The precise mechanism by which the
often slightly asymmetric when compared to that radioulnar ligaments provide stability remains in
of the sigmoid notch, which can lead to a CAM debate, although both ligaments must be present
effect with forearm rotation and further propagate to prevent translation in either direction [28, 32,
instability with ligamentous injury [28]. The end 40]. In one proposed mechanism, the dorsal
result of these anatomic factors is that only 20% radioulnar ligaments act as the primary restraint
of DRUJ constraint is provided by the ulnar head from dorsal dislocation of the ulna with prona-
and sigmoid notch articulation, with the majority tion, while the volar ligament opposed volar
of stability contributed by soft tissue attach- escape in supination. This is supported by a
ments, including the TFCC [32]. cadaveric study measuring ligament strain at the
Along with the TFCC, the DRUJ capsule, extremes of rotation [40–42]. The other theory is
pronator quadratus, extensor carpi ulnaris, and the exact opposite and was developed by
interosseus membrane all contribute to the sta- observing bone displacement with application of
bility of the DRUJ. The components of the TFCC a translation force. In this scenario, the volar
are the most important in providing stability and ligaments opposing dorsal displacement in
maintaining DRUJ kinematics [30, 37]. pronation and dorsal ligaments countering volar
The TFCC is comprised of several confluent displacement in supination via a tethering effect
structures that provide distinct functions. They [32, 43]. Both theories have little effect on clin-
include extending the articular surface of the ical management, other than reinforcing the idea
radius to cover the ulnar head, absorbing and that the ligaments exert a synergistic effect and
transmitting axial load forces, linking the distal that injury to both ligaments should be expected
radius and ulna, and supporting the ulnar carpus in the setting of bidirectional or severe
unidirectional instability that permits joint dis- activity limiting pain, weakness, and mechanical
location [28, 40]. symptoms. Patients with more severe instability
The articular disk of the TFCC is composed of may describe a palpable clunk with forearm
fibrocartilage, which extends from the articular rotation activities, such as turning a screwdriver
edge of the distal radius until in blends with the [28].
deep volar and dorsal radioulnar ligaments. Its Given that injuries to the radiocarpal, ulno-
primary function is to bear and transmit compres- carpal, lunotriquetral, and proximal radioulnar
sive loads across the DRUJ and provides minimal joints can cause similar symptoms, a compre-
stability to DRUJ translation [41, 44]. The deep hensive examination of the affected extremity
radioulnar ligaments reinforce the disk by pre- from the elbow distal is required. The examiner
venting splaying with compressive force [35]. should begin with inspection of the DRUJ, wrist,
The ECU tendon sheath runs from the dorsal and forearm, evaluating for swelling or promi-
ulnar head to the carpus. It not only stabilizes the nence of the ulnar head compared to the con-
ECU tendon, but acts to augment the dorsal DRUJ tralateral side. Tenderness to palpation of the
joint capsule. The volar portion of the TFCC also ulnar styloid can occur in the setting of fracture
contains three ulnocarpal ligaments named for the nonunion. Tenderness of the fovea, located at the
carpal bone to which they insert. These are the soft depression between the flexor carpi ulnaris,
ulnotriquetral, ulnolunate, and ulnocapitate liga- ulnar styloid, and triquetrum, can be indicative of
ments. The ulnotriquetral and ulnolunate ligaments a TFCC injury. Active and passive range of
originate from the volar radiolunate ligament, motion of both extremities, including wrist
while the ulnocapitate ligament runs more volar motion and pro-supination, should be measured
and originates from the fovea. They theoretically for comparison. Significant crepitus or decreased
provide a restraint to ulnocarpal translation and and painful motion should be noted, as this may
rotation, although their contribution to DRUJ sta- be indicative of DRUJ arthritis, which would be a
bility is unknown [28]. The meniscus homolog is contraindication to a soft tissue reconstructive
named for the loose connective tissue that occupies procedure [45]. The shuck test is performed to
the space between joint capsule, disk, and proximal assess for stability. The distal ulna is grasped with
surface of the triquetrum and provides an unknown one hand, while the distal radius is stabilized by
function [28]. the other hand. The ulna is then forcefully trans-
lated in a volar and dorsal direction. Pain or
increased translation compared to the contralat-
7.3.3 Clinical Evaluation eral side is indicative of DRUJ injury. Translation
can vary depending on forearm rotation, so the
Incompetency of the soft tissue stabilizers of the test should be repeated in supination, pronation,
DRUJ manifests clinically as complaints of pain, and neutral positioning [28, 45].
decreased grip strength and mechanical symp- Ulnocarpal joint stress testing should be per-
toms [39]. As with any initial assessment, it formed to assess for TFCC disk tears and
should begin with a detailed history of the symptomatic ulnocarpal impaction syndrome.
patient’s injury mechanism and attempted treat- The presence of pain with the press test, in which
ment. The location, quality, severity, and fre- a patient axially loads the wrist by using his or
quency of pain and instability should be her arms to push up from a seated to a standing
obtained, along with any factors that alleviate or position, is indicative of a TFCC tear [45]. The
aggravate symptoms. Patients will typically test may also be informative in the setting of
describe a history of a fall on an outstretched DRUJ instability. With attempted press up from
wrist or passive forceful wrist rotation, such as the chair, the ulnar head will appear more
with a jammed power tool [28]. Initial depressed on the affected side. Maneuvers in
ulnar-sided wrist pain that is aggravated by which the wrist is taken through a range of
forearm rotation, may, over time, evolve into motion (flexion–extension and/or pronation–
supination) while the wrist is axially loaded in allow visualization of the sigmoid notch and
ulnar deviation are also useful in eliciting TFCC ulnar head to assess for fracture, incongruence,
and ulnocarpal pathology [28]. and early degenerative changes and osteophyte
ECU subluxation and tendinitis and lunotri- formation [28]. Radiographs should also be used
quetral (LT) ligament tears can both result in to assess for a malunited distal radius fracture,
symptoms similar to DRUJ injury. Instability of which may be the source of DRUJ instability.
the ECU can be elicited by the “ice cream scoop” Computed tomography can be a useful tool as
test with rotation of the wrist from a position of it is able both to assess the congruency of the
pronation and radial deviation to supination and distal radioulnar articulation and to evaluate for
ulnar deviation. Lunotriquetral ligament injury advanced degenerative changes, which would
can be examined using the LT shear test. The contraindicate a soft tissue reconstruction. It is
lunate is stabilized between the examiner’s index most useful in the case of a symptomatic patient
finger and thumb, while the other hand provides a with subtle instability [28]. For complete com-
volar to dorsal translated force to the triquetrum. parison, both wrists should be imaged in identi-
A positive test is manifested by pain [28]. cal forearm positions, including neutral,
pronation, and supination [28]. Measurement
methods utilize the axial images. These include
7.3.4 Imaging the dorsal and volar radioulnar lines, congruency
method, epicenter method, and radioulnar ratio
Initial radiographic evaluation should consist of [49, 51, 52]. Numerous studies have demon-
standard PA and lateral views. Bilateral images strated variability in findings for all methods. So,
are useful for comparison, and care should be most authors recommend combing multiple
taken to insure proper positioning in order to measurements, while continuing to rely most
obtain symmetric views. Slight variations in heavily on the patient examination and clinical
rotation can exert a significant effect on mea- history [28].
surement of radiographic parameters. Ulnar Due to its ability to provide detailed images of
variance is best measured using the PA view, the soft tissue structures, MRI is the primary
although diastasis between the distal radius and advanced imaging modality utilized in evaluation
ulna, especially when not present on images of of TFCC tears in the acute setting. Arthroscopy
the contralateral limb, can be indicative of DRUJ remains the gold standard for diagnosis of TFCC
instability. On occasion, a small fleck of bone is injuries, but the sensitivity and specificity of MRI
avulsed from the fovea, indicating detachment of continues to improve [28, 53]. In the subacute or
the deep radioulnar ligaments [28]. In patients chronic setting, the use of MRI is less well
with symptoms of ulnocarpal impaction syn- defined. Although it has yet to be formally tested,
drome, the forearm pronated PA or clenched fist MRI may be useful for assessing attenuation of
view can be helpful in assessing for dynamic the TFCC and resulting inability to perform a
ulnar positive variance [46–48]. The goal of an primary repair in the patient who present weeks
accurate lateral view is for the pisiform to bisect or months from initial injury.
the volar surfaces of the distal pole of the sca-
phoid and the capitate (scaphopisocapitate lateral
view) [28]. While commonly used to assess for 7.4 Treatment
DRUJ instability, it is imprecise, as a subluxed
ulna can appear reduced and a dislocated ulna 7.4.1 Acute Distal Radioulnar Joint
can appear subluxed with only 10° of forearm Instability
rotation [49]. Suspected instability can be further
evaluated by lateral stress view with the patient The treatment of acute instability of the DRUJ
holding a 5 lb. weight in a position of pronation falls largely outside the scope of a chapter on
[50]. Semi-pronated and semi-supinated views nonunion of the wrist and hand, but bears
mentioning for sake of completeness. The most and result in DRUJ instability, especially in the
common cause of acute DRUJ instability is a setting of significant fracture displacement [28,
distal radius fracture. The majority of these 29, 57]. Ulnar styloid fracture nonunions are
injuries will be stable following accurate fracture often asymptomatic. If the DRUJ is stable,
reduction and stabilization. Following manage- painful tip fractures can be excised without
ment of the distal radius fracture, stability should affecting stability [29, 57]. In the setting of a
be reassessed over a full range of pronation and large fragment and stable DRUJ, excision of a
supination. If stability is maintained in only full symptomatic fragment should be performed with
pronation or supination, the DRUJ should be caution (Fig. 7.5). Stability should be
pinned in that position. Commonly, dorsal dis- re-assessed, and if the DRUJ becomes unstable,
locations are stable in supination, while volar the TFCC should be repaired to the fovea of the
dislocations are stable in pronation [28]. If styloid, using interosseus sutures [57]. In the
instability persists, then an open repair of the setting of DRUJ instability, fixation of a styloid
TFCC, which is discussed in detail later in this base fracture can be attempted. Numerous tech-
chapter, is indicated. niques have been described, including Kirschner
wires, compression screws, mini fragment plates,
tension band wiring, and sutures anchors [28].
7.4.2 Chronic Distal Radioulnar Joint The implant chosen depends on a combination of
Instability surgeon preference and the size of the fragment.
DRUJ stability should be re-assessed following
The goals of any treatment for DRUJ instability fragment fixation and, if instability persists,
should be restoration of stability and a full, requires a soft tissue reconstruction procedure.
pain-free range of motion. Although it is With regard to isolated soft tissue injury, tears
unknown if DRUJ instability predisposes a of the ulnar-sided attachments of the TFCC are
patient to arthritis, chronic instability symptoms most commonly associated with instability of the
will rarely improve without surgical management DRUJ. While tears of TFCC are common in the
[28]. Functional bracing has been proposed,
which showed effectiveness in decreasing sub-
luxation and improving range of motion [54].
This may be a consideration in lower demand
patients. Surgical treatment options include
operative fixation of instability resulting from an
ulnar styloid nonunion, direct repair of the
TFCC, or soft tissue reconstruction. In the setting
of chronic instability, the TFCC is frequently
irreparable and a soft tissue reconstruction tech-
nique is indicated [39]. Less straightforward is
the treatment of subacute injuries or subtle
instability, in which soft tissue attenuation is less
pronounced.
Ulnar styloid fractures are a common finding
in the setting of a distal radius fracture, occurring
61% of the time [55]. As discussed earlier,
fractures of the tip of the styloid often retain
stability of the DRUJ, as the deep volar and
dorsal radioulnar ligaments remain intact [28,
56]. Fractures of the base of the styloid may Fig. 7.5 Chronic ulnar styloid nonunion that remained
involve both the deep and superficial ligaments asymptomatic
setting of distal radius fractures, the majority will described, which can be divided into the cate-
not cause acute TFCC instability and do not gories of extra-articular linking of the radius and
progress if adequately addressed at the time of ulna via tenodesis or ulnocarpal sling and
injury [28]. Both arthroscopic and open repair intra-articular reconstruction of the radioulnar
techniques have been described for repair of ligaments. Indirect reconstruction techniques
TFCC injuries resulting in DRUJ instability. The have been studied in a cadaveric model by Adams
indications for arthroscopic TFCC repair in the and Petersen and failed to restore native DRUJ
setting of chronic DRUJ instability have not been stability or kinematics [58]. However, they may
completely described, and there is concern that be necessary in the setting of a previous ulnar
soft tissue repair alone may not confer adequate head resection, when an anatomic reconstruction
stability. Newer arthroscopic techniques, such as is no longer possible. These include proposed
pushlock anchors, which facilitate repair of the techniques by Boyes and Bunnell and by Hui and
TFCC directly to the fovea through a drill hole in Linscheid utilizing a strip of the flexor carpi
the ulna, may be of benefit, but have not been ulnaris (FCU) tendon to reconstruct the volar
adequately studied in this setting. Open repair is ulnocarpal ligaments [59, 60]. Both techniques
performed through a dorsal approach to the raise concern for the long-term stability of the
DRUJ between the fifth and sixth extensor DRUJ due to the unknown contribution to DRUJ
compartments, as described by Adams [28]. The stability from volar ulnocarpal ligaments. There is
extensor digiti mini tendon is mobilized and an additional risk of loss of motion from the
retracted ulnarly. An L-shaped capsulotomy is tethering effect of the tendon [39].
then made in the dorsal capsule, with the longi- Attempted anatomic reconstruction of one or
tudinal portion of the incision centered over the both radioulnar ligaments has been described in
radial aspect of the ulnar neck and the transverse techniques by Scheker et al. [50], by Johnston
limb beginning proximal to the dorsal radioulnar et al. [61], and by Adams and Berger [31]. In the
ligament. With retraction of the capsulotomy, the technique by Scheker et al., a tendon graft is used
TFCC can be visualized. to reconstruct only the dorsal radioulnar liga-
If amenable to repair, a second transverse ment. This raises concern for the long-term sta-
capsular incision is made distal to the dorsal bility of the construct, as cadaveric models have
radioulnar ligament to visualize the tear. Sutures demonstrated that both ligaments must be rup-
are passed in either a horizontal or vertical mat- tured for instability to occur [62]. Nonetheless,
tress configuration through the peripheral edge of they reported that all 14 patients treated with the
the tear and adjacent joint capsule. Holes are procedure were satisfied with their outcome, with
placed in the ulna using K wires or a small cal- no recurrent instability, improved grip strength,
iber drill, facilitating direct repair of the tear to and near complete resolution of pain at an aver-
the bone of the fovea. A suture-passing device is age of 1.5-year follow-up [50]. The techniques
valuable in passing the sutures through the bone proposed by Johnston et al. and by Adams and
tunnels. The sutures are then tied over the bone Berger seek to reconstruct both the volar and
with ulnar reduced in neutral forearm rotation. dorsal radioulnar ligaments with a palmaris
The dorsal capsule and extensor retinaculum are longus autograft [31, 61]. Both reported similar
closed in a single layer, excluding the extensor midterm results. Johnston et al. [61] reported
digiti minimi, which is left superficial to the satisfactory results in 13 of 14 patients, with
closure. Following completion of the case, DRUJ range of motion at least 90% of the unaffected
stability should be restored. If not, augmentation side in all patients. Adams and Berger reported
with a soft tissue reconstruction should be that patients recovered approximately 85% of the
considered. grip strength and wrist motion of the contralateral
Soft tissue reconstruction procedures are side [31]. In the both studies, 12 of 14 patients
indicated in the setting of an irreparable TFCC were able to return to their previous level of
injury. Numerous surgical techniques have been employment [31, 61].
Similar to open treatment of a TFCC injury, approach to the DRUJ. Osteotomes are used to
the technique described by Adams and Berger make 2 parallel transverse osteotomies in the
utilizes a dorsal approach to the DRUJ between dorsal ulnar corner of the distal radius, one just
the fifth and sixth extensor compartments [28, proximal to the subchondral surface and the
39]. The extensor retinaculum is divided longi- second at the proximal edge of the sigmoid
tudinally for later repair, and the EDM tendon is notch. A third, longitudinal cut is then performed
mobilized and retracted ulnarly. A dorsal connecting the parallel osteotomies 5 mm from
L-shaped capsulotomy is performed, and care the sigmoid notch. This produces a thin osteo-
should be taken to not violate the ECU sheath. cartilaginous flap, which can be backfilled with
The periosteum of the dorso-ulnar distal radius is cancellous bone graft from the distal radius. To
elevated, deep to the fourth extensor compart- this point, only a case report and a technique
ment. Depending on the size of the palmaris article have been published on this subject.
longus graft, a 3.2–4 mm cannulated drill bit Clinical trials would be beneficial to determine
system is used to place a tunnel in the distal long-term results.
radius from dorsal to volar at a position
approximately 5 mm proximal to the lunate fossa 7.4.2.1 Postoperative Management
and 5 mm radial to the sigmoid notch. The same Following soft tissue reconstruction, patients are
cannulated drill is used to place a second tunnel immobilized in a long are splint for 3 weeks in
in the ulnar beginning in the ulnar neck and neutral forearm rotation. At 3 weeks, the splint is
exiting at the fovea. C-arm fluoroscopy is valu- removed and they are converted to a short arm
able in confirming the position of the guide wires cast and limited forearm rotation is permitted for
prior to drilling. A whipstitch is placed in each an additional 3 weeks. The patient is then con-
end of the graft, and a suture passer is then used verted to a removable wrist brace to be used for
to weave the graft through the radius and ulna. an additional 2 months. Therapy is initiated with
The remaining limbs of the graft are passed active and gentle-only passive wrist flexion,
around the subcutaneous border of the ulnar neck extension, pronation, and supination. Strength-
and tied into place with the ulna reduced and the ening is started early, with care taken to avoid
forearm in neutral rotation. Care should be taken high forces with the arm in full pronation and
to insure that branches of the ulnar nerve are supination. More aggressive passive range of
entrapped in the construct. Additionally, if the motion and strengthening are delayed until the
graft is not long enough to pass around the ulna 4-month mark, with the goal of recovering 85%
after passing through the ulnar tunnel, a bio- of native forearm rotation by 6 months. Return to
composite interference screw can be utilized to activities and lifting greater than 5 lbs is delayed
stabilize the graft within the ulnar tunnel [28]. until at least 4 months post-surgery [28, 31].
The dorsal capsulotomy and extensor retinacu-
lum are closed in a single layer, and the EDM
tendon is left superficial to the capsular closure. 7.5 Carpal Bones
In patients with flat face alignment of the
sigmoid notch or who have sustained a fracture 7.5.1 Scaphoid
of the rim of the sigmoid notch, an osteoplasty
can be considered as an isolated or complimen- 7.5.1.1 Background
tary procedure to prevent dorsal subluxation of The scaphoid is the most commonly fractured
the distal ulna [28]. Axial computed tomography bone in the carpus, accounting for between 60
images can be useful in assessing notch align- and 70% of all carpal bone fractures [65]. Frac-
ment and deformity [28]. The procedure pro- tures most frequently affect a young, active, male
posed by Wallwork and Bain can be combined patient population. In a study of US military
with reconstruction of the TFCC ligaments [63, personal, the incidence was 121 per 100,000
64]. The distal radius is accessed through a dorsal person-years, with fractures most commonly
occurring in males ages 21–24 [66]. Achieving The bone of the proximal pole is the most dense, as
union is of paramount importance, as fractures it transmits axial load across the carpus from the
that fail to heal progress to a predictable pattern distal radius. In contrast, the bone of the waist has
of disability, carpal collapse, and eventual the lowest density, which may explain why a
radiocarpal arthrosis. Healing of these injuries is majority of fractures occur in this region [73, 74].
a complex process, influenced by fracture loca- Morphologic evaluation has demonstrated that
tion and orientation, displacement, and vascular male scaphoids are longer than those of females,
supply to the scaphoid. As a result of the liga- which may have an impact on screw length with
mentous connections between the bones of the surgical fixation [73].
carpus, a healing scaphoid fracture is subjected to As a result of the scaphoid being almost
significant shearing and bending forces [65]. entirely covered with articular cartilage, there is
Despite this, nondisplaced or minimally dis- minimal space for perforating blood vessels to
placed fractures that involve the body of the enter. This unique anatomy results in a complex
scaphoid and distal can be treated with rigid blood supply. A cadaveric study performed by
immobilization in a cast with expected union Gelberman et al. determined that approximately
rates reported at greater than 90% [67]. Delayed 80% of blood flow was provided via a branch of
diagnosis, fracture displacement greater than the radial artery entering the dorsal ridge at the
1 mm, angulation greater than 15°, proximal scaphoid waist and supplying the proximal pole
fracture location, and evidence of osteonecrosis in a retrograde fashion. The remaining 20% is
on radiographs represent risk factors for non- supplied by further volar radial branches entering
union and are commonly used as indications for the distal pole. The tenuous blood supply of the
surgical management. The rate of union after proximal pole results in a unique susceptibility to
acute surgical management has been reported in nonunion and avascular necrosis following frac-
several meta-analyses and approaches 100% [68– tures of the scaphoid waist or proximal [75].
72]. As a result, correct initial assessment and The precise mechanism by which a scaphoid
management of these fractures is of paramount fracture occurs has not been clearly elucidated,
importance. When fractures fail to heal as although it appears to be a combination of axial
expected or present in a delayed fashion care of load and either wrist hyperextension or, less
the nonunited scaphoid can prove challenging. commonly, wrist flexion [76, 77]. As described
The correct approach to a delayed union or earlier, healing potential is dependent on a
nonunion of the scaphoid is a topic of continuing number of factors including location, displace-
research and debate among hand specialists and ment, angulation, and vascularity [28].
will be discussed in this section. Intramembranous ossification is the mechanism
by which scaphoid fractures heal. The resulting
7.5.1.2 Anatomy and Biomechanics lack of protective callus formation renders the
The bones of the carpus are aligned in two scaphoid susceptible to mechanical forces
matching rows, supported by both intrinsic liga- throughout the healing process. This can lead to
ments and a complex system of volar and dorsal fracture displacement or angulation if proper
extrinsic ligaments. The scaphoid has a complex immobilization or stabilization is not provided.
three-dimensional anatomy, closely resembling a A scaphoid waist fracture is exposed to a com-
peanut, with articular cartilage covering 80% of its bination of bending, shearing, and translation
surface. It is the only carpal bone that bridges both forces. Axial load applied to the wrist prior to
carpal rows. There are three anatomic regions into fracture healing will result in progressive flexion
which the bone is divided: the proximal pole, waist and pronation of the distal pole. Over time, the
or body, and distal pole or tubercle. The proximal distal pole will continue to angulate as volar bone
pole articulates with scaphoid fossa of the distal is reabsorbed, leading to a “humpback defor-
radius and the lunate, while the distal pole forms mity.” This is further compounded by the limited
the scapho-trapezial-trapezoid articulation. blood supply of proximal fractures.
The combination of fracture displacement, onset of pain, or loss of motion. Clinical records
angulation, and absent blood supply all con- should be reviewed to determine any previous
tribute to the development of a nonunion and treatment. In the case of closed management, the
eventual avascular necrosis [28]. duration and type of immobilization use should be
If left untreated, scaphoid nonunion leads to a obtained. For patients treated with surgery, the
predictable pattern of degenerative arthritis approach and type of fixation used are beneficial
within the radiocarpal and midcarpal joints for later preoperative planning. Although the
known as scaphoid nonunion advanced collapse majority of these patients are young and active, a
(SNAC). The scaphoid is a vital link between the complete medical history should be obtained.
proximal and distal carpal rows. As a result, Particular attention should be applied to use of
scaphoid nonunion leads to a significant disrup- tobacco products. While not an absolute con-
tion of carpal mechanics. In a normal wrist, the traindication to surgery, their use is a risk factor for
scaphoid and lunate are connected by the nonunion and the patient should be counseled for
scapholunate interosseus ligament, which draws and offered help with tobacco cessation [79, 80].
the lunate into a flexed position with the scaphoid The physical examination should attempt to
with wrist radial deviation. The volar collapse of localize the source of pain in as gentle a manner
the distal pole in a humpback deformity results in as possible. Wrist range of motion should be
reduced carpal height. With radial deviation, the obtained and compared to the contralateral side.
lunate continues to extend along with the proxi- While pain is not always localized to the ana-
mal pole of the scaphoid, while the distal pole tomic snuff box area or either pole of the sca-
remains in a flexed position. Over time, the phoid, diffuse pain and significantly decreased
resulting dorsal intercalated segmental instability range of motion should alert the examiner to the
(DISI) pattern becomes fixed and the progression possibility of advanced degenerative changes.
of degenerative changes is similar to that The goal of the radiographic evaluation should
observed in a scapholunate ligament deficient be to determine the degree of healing, alignment,
wrist. Degenerative changes begin in the radial and vascularity of the fracture, as well as any
styloid articulation with the scaphoid before evidence of degenerative changes. Initial radio-
spreading to the entire radioscaphoid, radiolu- graphs should include standard posteroanterior,
nate, and scaphocapitate joints, and ending in lateral, 45 degree pronated oblique, and navicular
pancarpal arthritis. Patients may be initially (PA in wrist ulnar deviation) views. They may
asymptomatic, but will eventually develop pro- reveal sclerosis, cyst formation, bone reabsorp-
gressive activity-related pain. In one study, 97% tion, fracture displacement or angulation, or
of patients with at least a 5-year history of sca- hardware loosening or failure. The lateral radio-
phoid nonunion demonstrated degenerative graphs can also be used to evaluate for a DISI
changes on radiographs [78]. deformity, with a scapholunate angle >60° or a
radiolunate angle >30° [81, 82]. If present, it
7.5.1.3 Clinical Evaluation factors into preoperative planning, as correction of
The evaluation of patients with known or expected both the alignment of the scaphoid and the normal
scaphoid nonunion should begin with a detailed scapholunate relationship can be challenging [83].
history. Most patients will often present with pain, All previous radiographs, including initial injury
stiffness, or inability to resume normal activities films, should be reviewed to determine the pro-
beyond the period of time one would expect for gression of healing and any evidence of progres-
fracture healing, while a subset will remain sive fracture displacement or angulation.
asymptomatic. Although it can be difficult, an Computed tomography (CT) scans provide
onset of injury should try to be obtained. Occa- the most detailed images of the osseous anatomy
sionally, patients will not be able to recall a and can be useful in determining nonunion in the
specific event, but rather a decrease in function, setting of equivocal radiographs. CT has
demonstrated high intra-observer reliability in radio-opacity of the proximal pole and lack of
determining displacement and fracture union bridging trabeculae between fracture fragments
(Fig. 7.6) [84]. CT can also provide valuable [85, 89].
information regarding bone reabsorption follow- Due to the unreliability of radiographs in
ing collapse and early evidence of degenerative predicting osteonecrosis, MRI plays a key role in
changes. CT images are used to determine the preoperative evaluation of scaphoid non-
angulation of the scaphoid with the lateral union. Studies have established MRI as the most
intrascaphoid angle or height-to-length ratio on effective imaging modality in determining vas-
sagittal images [84]. For accuracy, the CT should cularity of the proximal pole (Fig. 7.7) [90–93].
be oriented perpendicular to the long axis of the It can also be effective in diagnosing occult
scaphoid, rather than the wrist [85]. The normal scaphoid fractures and determining the degree of
lateral intrascaphoid angle is 24°, while an angle devascularized bone in patients who have
greater than 45° is predictive of an increased risk undergone previous surgical treatment [83].
of arthritis, even in healed fractures [86]. Decreased or absent signal intensity on T1 and
Although the height-to-length ratio has demon- T2 weight images has been associated with
strated a greater intra-observer reliability than the compromised vascular supply [93]. Patients with
intrascaphoid angles, the clinical significance of this finding on preoperative MRI demonstrated
this is unknown. Scaphoid collapse is considered suboptimal healing rates when not treated with a
significant with a height-to-length ratio greater vascularized bone graft [93]. Additionally, a
than 0.65 [87, 88]. CT can also evaluate for clinical study has directly correlated absence of
technical errors, such as screw misplacement and T1 signal on MRI with the presence of
inadequate fracture reduction and compression. osteonecrosis, empty bone lacunae, and poor
Current CT protocols with metal suppression are uptake on bone scan within intra-operative bone
useful in minimizing hardware artifact. While not biopsy specimens [93]. While MRI is unneces-
as effective as MRI, CT can predict proximal sary if plain radiographs clearly demonstrate
pole osteonecrosis through increased osteonecrosis, it is recommended to completely
Fig. 7.6 Computed

tomography scan showing
nonunion of a scaphoid waist
fracture previously treated
with screw fixation
Fig. 7.7 Magnetic resonance

image of scaphoid nonunion
with signal changes on t1 and
t2 consistent with avascular
necrosis of the proximal pole
evaluate for the presence of osteonecrosis in any to elbow and wrist stiffness, exerting its own
waist or proximal pole fracture with an estab- negative impact on function and quality of life.
lished diagnosis of nonunion. Further, management of delayed union without
bone reabsorption, collapse, or osteonecrosis is
7.5.1.4 Treatment technically less demanding and may be accom-
Surgical treatment is indicated in the setting of plished with compression screw without the need
nonunion, as spontaneous healing is extremely for bone graft. Whether initial patient presenta-
rare [94]. The addition of cast immobilization tion is acute or delayed, fracture displacement
and/or pulsed electromagnetic field treatment with greater than 1 mm, fracture comminution, frac-
a bone stimulator does not result in predictable ture of the proximal pole, fracture angulation as
consolidation once the diagnosis of nonunion has manifested by an intrascaphoid angle greater
been established [95]. The healing rate was only than 45° or height-to-length ratio greater than
69% with the use of casting and bone stimulators 0.65, and poor patient compliance all represent
in nondisplaced nonunions without radiographic predictors of eventual non- or malunion and
evidence of osteonecrosis [95]. warrant immediate surgical management [86, 87,
Delayed union represents a category of frac- 98].
ture that merits individual discussion. Although In the setting of established nonunion, the
there is no clear definition of delayed union, it majority of authors recommend open reduction
should be considered when radiographs fail to and internal fixation of the fracture with bone
demonstrate expected progression of healing. graft (Fig. 7.8) [98–102]. Throughout the evo-
Initial conservative treatment of nondisplaced lution of surgical stabilization, a number of
fractures of the waist or distal pole is a reason- implants have been used, including Kirschner
able approach. However, surgical intervention wires, staples, plates, and compression screws.
should be strongly considered when radiographs The implant of choice must be capable of with-
fail to show signs of healing after 6–8 weeks of standing shearing forces with disrupt fracture
immobilization and certainly by 12 weeks. healing. Kirschner wires lack the compressive
Determination of healing can be difficult due to strength necessary to maintain fracture reduction,
the lack of callus deposition, and a CT scan may while the use of staples and plates, although
be necessary for definitive evaluation. This is displaying satisfactory healing rates with acute
especially prudent in athletes and young laborers, fractures, raise concern hardware impingement
as research has demonstrated faster return to play and damage to the surrounding articular cartilage
and work and decreased overall medical cost and often require later hardware removal [103,
with early surgical intervention [96, 97]. Also, 104]. New mini-plate technology may alleviate
prolonged long arm cast immobilization can lead some of these concerns, while theoretically
Fig. 7.8 a Chronic nonunion

of scaphoid waist fracture
without avascular necrosis of
the proximal pole.
b Images 10 weeks
postoperatively demonstrate
healed fracture after volar
nonvascularized bone grafting
and compression screw
fixation
provided increased torsional stability compared accurate screw positioning is critical, with the
to a screw [105, 106]. Nevertheless, compression greatest stability imparted by a screw positioned
screws remain the mainstay for current treatment in the center–center position of the bone on PA
of acute scaphoid fractures and nonunion. Her- and lateral views or perpendicular to the fracture
bert developed the initial headless screw. Even line [108, 109]. Additionally, studies by Trumble
without a compression design, they reported a et al. reported that screw placement within the
union rate of 100% for acute fractures and 83% central third of the proximal pole reduced time to
overall [107]. Subsequent advances have union by 50% [81, 109, 110]. Several companies
demonstrating increased compression through a have now developed compression screws, and
partially threaded or fully threaded, variable pitch little data exists comparing their effectiveness.
design. The addition of a cannulated system Therefore, screw choice remains largely depen-
utilizing guide wire placement under fluoro- dent on surgeon preference.
scopic guidance has improved accuracy of screw After implant choice, the next steps in the
placement. Studies have demonstrated that treatment algorithm are surgical approach and
choice of bone graft material. A combination of is performed to confirm the presence of the
fracture stability and alignment, bone reabsorp- fibrous nonunion, assess vascularity, and rule out
tion, presence of osteonecrosis, and previous ligamentous injury. Under arthroscopic visual-
surgical intervention is used to determine the ization, a dorsal capsular release is performed
surgical approach and bone graft source. using a curved hemostat through the 3,4 portal.
The arthroscope is then removed, and the sca-
phoid is drilled for the compression screw. The
7.5.2 Scaphoid Nonunion Without guidewire is left in place, and bone biopsy needle
Osteonecrosis is placed over it to deliver bone graft. Finally, a
compression screw 4 mm shorter than the mea-
Well-aligned, stable nonunions with evidence of sured length of the scaphoid is placed. Using this
substantial bone loss or osteonecrosis represent technique on 108 scaphoid fracture nonunions
the one category of fracture in which need for with either no displacement or a reducible
intra-operative bone graft remains controversial. humpback deformity, Slade and Gillon reported a
A study by Shah and Jones found that fractures 96% healing rate at 9 months. Ten cases of
with a stable fibrous nonunion or intact cartilage delayed healing required repeat percutaneous
cap healed without need for bone graft [111]. bone grafting [114]. Percutaneous bone grafting
Additional studies by Slade et al. and Ikeda et al. can also be performed through a limited volar
demonstrated similar results, reporting healing of approach, in the setting of a nondisplaced frac-
all nonunions treated with screw fixation alone if ture nonunion. Using this technique, only the
sclerosis was less than 1–2 mm on CT scan. Of scaphotrapezial joint in exposed in the volar
note, fractures treated within 6 months of initial approach. The volar beak of the trapezium is
injury healed faster than older injuries [112, removed to allow grafting and retrograde screw
113]. Fracture location determines surgical placement [36].
approach, with most authors recommending a Other authors would argue that a corticocan-
volar approach for waist fractures in order to cellous graft is required for structural stability
preserve the remaining dorsal blood supply and a and that only an open approach allows complete
dorsal approach for proximal pole fractures, evacuation of sclerotic bone, which can prevent
which allows improved access and reduces the healing. With open treatment of
risk of displacing the proximal pole with guide- well-vascularized nonunions, the location of the
wire placement [36, 65, 83]. fracture remains the primary determinant of the
Fractures that are well perfused, but with surgical approach. The volar approach described
substantial bone loss, require the use of bone by Russe facilitates access to the waist and distal
autograft to provide structural stability and pole of the scaphoid, while preserving dorsal
enhance healing potential. The choice of can- blood supply, especially in a fibrous nonunion or
cellous versus corticocancellous structural graft with an intact dorsal cortex [36]. The standard
remains an area of controversy. Geissler, Slade, Russe incision starts with a longitudinal incision
and Gillon have argued that arthroscopically proximally over the flexor carpi radialis
guided percutaneous fixation, and bone grafting (FCR) tendons and then angles obliquely in line
is adequate for healing in nondisplaced or mini- with the long access of the scaphoid to the level
mally displaced fractures with fibrous stabiliza- of the scaphotrapezial joint. The FCR tendon
tion of the fracture site [65, 114]. In their sheath is incised to allow ulnar retraction of the
technique, cancellous bone autograft is harvested tendon and protection of the palmar cutaneous
from either the distal radius or iliac crest using a branch of the median nerve. Care is taken to
bone biopsy needle. A guidewire is then place protect the deep branch of the radial artery,
within the scaphoid using a percutaneous of which is retracted radially. The volar radiocarpal
mini-incision dorsal approach. A second wire is ligaments are incised obliquely to allow access to
placed as an anti-rotation wire, and arthroscopy the scaphoid and later repair of the ligaments.
Preservation of as much RSC ligament as pos- procedure are varied and appear to rely most
sible prevents volar subluxation of the scaphoid heavily on the location of the fracture. Treating a
proximal pole [81, 110]. total of 84 patients, Dacho reported union rates of
The Matti-Russe procedure describes cortic- 82% overall and 81% in proximal pole fractures
ocancellous grafting through the volar approach [117]. However, Barton reported a healing rate of
and is particularly useful in the setting of a only 54%, with the majority of failures occurring
fracture with minimal displacement, a large bone in proximal pole fractures [118]. This is sup-
defect, and absence of a humpback deformity ported by Green’s study, which reported a heal-
[115]. Any fibrous tissue at the site of the non- ing rate of 71% that was negatively affected by a
union is removed, and an osteotome is used to higher incidence of failure to heal in proximal
expose the fracture line. Osteotomes or pole fractures [116].
high-speed burr is then used to cut a cortical Fractures that involve a viable proximal pole
window, remove a strip of cortical bone, and are best managed through the dorsal approach.
create a trough for graft placement. All necrotic The dorsal approach allows improved access to
bone is excised from the proximal fragment, and the proximal pole in order to remove nonviable,
punctate bleeding is a sign of preserved vascu- sclerotic bone and allows anterograde screw
larity. If necessary, the scaphoid is reduced. placement. The orientation of proximal pole
Kirschner wires can be placed in each fragment fractures is commonly from distal volar to
as joysticks to help facilitate reduction. A cortic- proximal dorsal, so a volarly placed screw runs
ocancellous graft equal in size to the cavitary the risk of both of displacing the proximal frag-
defect is then harvested from either the iliac crest ment and not crossing the center of the nonunion
or distal radius. The Green modification of the site [36]. The scaphoid may be approached from
procedure calls for two parallel grafts placed with either a longitudinal incision in the middle of the
the cancellous surfaces facing one another [116]. wrist centered over the radiocarpal joint or an
In the original description, the scaphoid is sta- oblique incision from the Lister’s tubercle to the
bilized with parallel Kirschner wires, although base of the second metacarpal [36, 65]. The
compression screws can also be used and provide capsule may then be divided longitudinally or
a more stable construct (Fig. 7.9). Upon com- with a capsular splitting [119] incision. In the
pletion of the procedure, the capsulotomy is dorsal capsular splitting incision, the dorsal
closed with nonabsorbable suture. Results of the radiocarpal ligament (DRC) is split from just
Fig. 7.9 Scaphoid nonunion

without avascular necrosis
treated with corticocancellous
graft from distal radius with
humpback deformity
corrected and screw in place
distal to Lister’s tubercle to the triquetrum, while treatment of these scaphoid nonunions with car-
the dorsal intercarpal ligament (DIC) is incised pal instability. The key components of his tech-
from the triquetrum to the distal pole of the nique are as follows: (1) the use of preoperative
scaphoid. The resulting ligamentous flap is ele- radiographs of the contralateral wrist to assess
vated from ulnar to dorsal for later repair. The correct scaphoid length and alignment; (2) the
distal aspect of the extensor retinaculum can be use of a volar approach; (3) the insertion of a
incised, and Lister’s tubercle can be utilized for wedge-shaped corticocancellous graft from the
either cancellous or corticocancellous bone graft. iliac crest to correct alignment after resection of
As with the volar approach, the nonunion site is the pseudoarthrosis; and (4) stabilization with a
identified and necrotic bone is removed with compression screw [65, 123]. Fisk’s technique
curettes or a high-speed burr. Kirschner wires used radial styloid bone graft, which Fernandez
placed in each pole of the scaphoid and used as modified due to the greater compressive strength
joysticks may be beneficial in reducing the of the iliac crest. Fernandez reported radio-
fracture. A guidewire for a cannulated screw graphic healing in 19 of 20 patients treated with
system is placed under fluoroscopic guidance this technique with an average time of from work
once the fracture has been reduced and bone graft of 8.9 weeks [123]. Eggli et al. found similar
has been placed. Placement of the wire is facili- results, with 35 of 37 patients achieving union.
tated by flexion of the wrist. A second eccentric Preexisting osteonecrosis was present in both of
wire can be placed to prevent fracture rotation the 2 cases that failed to heal [124].
with drilling and screw placement. Upon com-
pletion of the case, the capsulotomy is closed
with nonabsorbable suture. The literature exam- 7.5.3 Scaphoid Nonunions
ining dorsal approach and bone grafting in the with Osteonecrosis
absence of osteonecrosis is limited. Inoue et al.
[120] reported on 16 patients, 13 of which went It is generally accepted that vascularized bone
on to radiographic union. graft is necessary in the setting of proximal pole
In either a volar or dorsal approach to a osteonecrosis, or when traditional bone grafting
fracture with extensive bone loss, resulting in has failed to achieve union (Fig. 7.10). This is
concerns about stability after screw fixation, a supported by studies that demonstrated unac-
second screw or Kirschner wire can be inserted ceptable union rates in the setting of
as an augment [121, 122]. The use of a second osteonecrosis [115, 116, 120, 124–126]. In a
screw requires prior planning, as the first screw canine study, Sunagawa et al. reported a union
must be positioned to either side of the scaphoid rate of 73% if vascularized graft was used and
to allow space for the second. The Kirschner wire 0% with conventional graft [127]. Multiple graft
is typically left in place until there is radiographic sources have been described, the most common
appearance of fracture consolidation. Garcia et al. being grafts harvested from the distal radius and
reported on a total of 19 patients treated with rotated on an intact pedicle. Other techniques
placement of 2 compression screws and bone include pedicle graft from the second metacarpal
graft through either a volar or dorsal approach. and ulna, free grafts from the medial femoral
All fractures went on to union at an average of condyle, iliac crest and rib costochondral junc-
3.6 months [122]. tion, and arterialization.
For patients with a fixed humpback deformity Techniques for graft harvest from the distal
of the scaphoid or with a large cavitary lesion of radius are based on the comprehensive blood
the volar scaphoid that results in a DISI defor- supply of the distal radius described by Sheetz
mity, a volar approach is required for complete et al. in 1995 [101]. Their close proximity to the
correction and stabilization of the deformity. scaphoid confers some advantages over other
Fernandez has described a modification of a techniques. Minimal extra exposure and dissec-
procedure originally developed by Fisk for tion are required for their harvest, and they are
Fig. 7.10 a Chronic

nonunion of scaphoid waist
fracture. b 4 months after
compression screw fixation
with volar nonvascularized
bone graft and continued
nonunion of fracture. c,
d Images after revision
fixation of nonunion with
dorsal 1,2 intercompartmental
supraretinacular artery
vascularized bone graft and
screw and k wire fixation with
healing of fracture
rotated on an intact pedicle, eliminating the need proximal to the radiocarpal joint. The 2,3 ICSRA
for microvascular anastomosis [85]. The dorsal has a longer pedicle and may allow for greater
graft is typically based on either the 1,2 or 2,3 graft rotation [85]. An additional graft based
intracompartmental supraretinacular arteries, commonly used in the treatment of lunate
while the volar graft is from the volar carpal osteonecrosis (Kienböck disease).
artery [65, 83, 85]. The nutrient vessels of the Harvest of the 1,2 ICSRA was first described
distal radius from which the dorsal grafts are by Zaidemberg et al. in 1991 and is likely the
based are named by their relationship to the most commonly used graft source in part due to
extensor compartments of the wrist and the its consistent location on the extensor retinacu-
extensor retinaculum [65]. The 1,2 intercom- lum [36, 65, 128]. A dorsal radial skin incision is
partmental supraretinacular artery (ICSRA) lies centered over the radiocarpal joint between the
superficial to the extensor retinaculum between first and second dorsal compartments to allow
the first and second dorsal compartments, while exposure to the scaphoid and distal radius.
the 2,3 ICSRA is also superficial and between the A tourniquet is typically used for surgery, but
second and third compartments [65]. The 1,2 exsanguination is not performed in order to allow
ICSRA takes off from the radial artery 5 cm identification of the artery. Branches of the
superficial sensory radial nerve must be identified [129]. Straw et al. [130] reported less successful
and protected. Meticulous blunt dissection is results, with only 6 of 22 patients achieving
performed until the 1,2 ICSRA is identified as it union. Of note, they used Kirschner wires for
courses dorsally from the radial artery to lie fixation, which may have negatively impacted
superficial on the extensor retinaculum between union rates.
the first and second compartments. The artery As mentioned previously, vascularized bone
typically appears as a thin red stripe. The first and graft from the volar distal radius has also been
second extensor compartment tendons are described, based on the volar carpal artery. Gras
released. The tendons of the second compartment and Mathoulin reported a union rate of 89.5% in
are retracted ulnarly, and the first compartment 38 patients treated with the procedure, all of who
tendons are retracted radially. The arterial pedicle had undergone previous surgical intervention.
is mobilized by making parallel incisions in the The average time to radiographic union was
periosteum on either side of the artery at the 10.8 weeks [131]. This technique may be par-
desired length for the graft. The pedicle should ticularly useful with the combination of
not be mobilized more than 10–15 mm proximal osteonecrosis and a humpback deformity or large
to the joint line, as this is the area where the volar bone defect of the scaphoid that requires a
nutrient vessels enter the bone. The pedicle is volar wedge graft for deformity correction. Other
then freed to nearly the level of the radial artery options for this problem include a dorsal
to allow rotation of the graft. The desired size approach with a radial styloidectomy to facilitate
graft is then harvested. volar placement of the ICSRA, a dorsal approach
We recommend the use of osteotomes for with a vascularized second metacarpal bone graft
graft harvest, although some surgeons will utilize pedicle that is rotated volarly, or two separate
an oscillating saw. The artery is ligated proxi- approaches with a volar wedge graft to correct
mally, and the tourniquet should be deflated to the deformity and ICSRA graft to provide vas-
confirm blood flow through the pedicle. The cularity [83].
nonunion site is prepared with complete excision Similar to the grafts of the distal radius, the
of any necrotic bone as previously described. second metacarpal vascularized graft is a pedi-
Preparation of the nonunion can also be per- cled graft that can be rotated into the scaphoid
formed prior to graft harvest. Kirschner wire defect. The graft is based on second dorsal
joysticks are again useful for control of the sca- metacarpal artery or the dorsal intercarpal arch
phoid with graft placement and fracture reduc- and can be harvested from the dorsal head or
tion. Compression screws are used for fixation as base of the metacarpal. The pedicle does not
they have significantly better union rates than cross the wrist joint, theoretically decreasing the
Kirschner wires [129]. If fracture fragments are risk of occlusion due to vessel kinking. It can be
completely unstable, fixation can be placed first, harvested from a single dorsal approach or a dual
followed by impaction of the graft through a approach, in which it is rotated and placed in the
cortical window [65]. volar scaphoid defect [85]. Mathoulin and Bru-
The use of the 1,2 ICSRA graft has demon- nelli [132] reported a union rate of 93% and time
strated mixed results within the literature. to union of 4 months in 15 patients treated with
Zaidemberg et al. [128] reported union in all 11 grafts harvested from the metacarpal head.
patients treated with the procedure in their orig- Despite the high union rate, only 10 patients had
inal technique article. 34 of 50 nonunions treated acceptable functional outcomes and 2 sustained
by Chang et al. went on to heal at an average of radial nerve irritation. Sawaizumi et al. [133]
15.6 weeks post-surgery. Risk factors for per- used a graft from the proximal metacarpal,
sistent nonunion included advanced age, believing it allowed harvest of a larger piece of
osteonecrosis of the proximal pole of the sca- bone. They reported that all 14 patients treated
phoid, preoperative humpback deformity, lack of with the procedure went on to union at an aver-
screw fixation, tobacco use, and female gender age of 10.2 weeks. Of note, the presence of
preoperative osteonecrosis was not included in utilizing the deep circumflex iliac vessels as a
their study. pedicle [136]. Arora et al. [136] treated 21
Ulnar bone graft utilizing the ulnar artery as a patients with documented osteonecrosis and who
pedicle represents another option for treatment of has failed previous surgical intervention. They
avascular scaphoid nonunions. The procedure reported a union rate of 76%, with an average
has some advantages, including a predictable time to union of 17 weeks and no donor site
vascular pedicle, large periosteal layer, and complications. Al-Jabri et al. [137] performed a
decreased donor site morbidity compared to free systematic review of studies in which patients
vascularized grafts [85, 134]. The greatest dis- with nonunion were treated with either a medial
advantage is that the graft utilizes the ulnar artery femoral condyle or iliac crest free graft. They
as a pedicle, which requires division and recon- reported a combined union rate of 100% in 56
struction of the artery with an interpositional vein patients treated with a medial femoral condyle
graft. In a study by Guimberteau and Panconi, all graft and 87.7% in 188 patients treated with an
8 patients treated with the procedure both went iliac crest graft.
on to radiographic union at an average of Proponents of rib costochondral free vascu-
4.6 months and returned to their previous level larized bone graft tout its use in the setting of
of work or athletic activities without postopera- significant collapse of the proximal pole of the
tive complications [134]. Given the possible scaphoid [138–140]. A horizontal incision is
complications associated with division and placed over the 9th rib, and the osteocartilage-
reconstruction of the ulnar artery and that nous surface of the costochondral joint is
patients had failed an average of 2 prior surg- exposed. An osteochondral plug is harvested
eries, the procedure might be most useful in the with an oscillating saw and contoured to match
setting of revision surgery, in which distal radius the anatomy of the proximal scaphoid. The
pedicle graft has already been used. proximal pole of the scaphoid is resected, and
With regard to free vascularized grafts, tech- the graft is implanted and fixed to the distal
niques for harvesting from the medial femoral pole with either a screw or removable Kirschner
condyle, iliac crest, and rib costochondral surface wires. Sandow reviewed 47 patients treated with
have been described. Free vascular grafts provide the procedure [138]. All patients reported
enhanced structural support, such as that needed improvement in functional scores, with none
to correct a humpback deformity. Potential requiring additional salvage procedures at an
complications include morbidity from a second average follow-up of 3 months. In another study
surgical incision for graft harvest and failure of with longer follow-up, Veitch et al. reported
microvascular technique for anastomosis of the functional improvement in 13 of 14 patients
graft [85]. The medial femoral condyle graft is [139]. Graft union to the residual scaphoid was
based on a pedicle from the descending genicu- 100% in both studies. Tropet et al. [140]
late vessels or the superior medial geniculate reported on 18 patients treated with technique at
vessels [135]. Jones et al. [135] compared 1,2 an average follow-up of 4.1 years, finding
ICSRA and medial femoral condyle grafts in the excellent or good results in 15 cases, fair in 2
treatment of scaphoid nonunions with cases, and poor in 1 case, in which there was
osteonecrosis of the proximal pole. All 12 subluxation of the graft.
patients treated with medial femoral condyle Direct implantation of the second dorsal
graft healed, compared to only 4 of 10 patients metacarpal artery or the dorsal index artery into
treated with 1,2 ICSRA graft. Average time to the scaphoid has also been described and pre-
healing was 13 weeks in the medial femoral dominately used in the setting of previous sur-
condyle group, compared to 19 weeks in the 1,2 gery and limited vascularized bone graft options.
ICSRA. The iliac crest free graft allows harvest Known as arterialization, the procedure was first
of structurally sound tricortical iliac crest graft described by Hori et al. in a canine model [141].
Fernandez and Eggli [142] reported the use of the patients who have failed previous surgery, sig-
procedure in 11 patients, 10 of who went on to nificant delay in treatment or with preoperative
heal their fracture. Despite the high union rate, osteonecrosis. Ricardo et al. have reported
50% required subsequent operations, including accelerated time to union (by 38 days) in patients
radial styliodectomy and limited carpal with osteonecrosis who were treated with a bone
arthrodesis. Two other case studies in patients stimulator [145]. No current evidence supports
who had undergone previous surgery reported use of one stimulator type over another, and no
successful fracture healing [143, 144]. comparative trials have been performed to date.
Author’s Preferred Treatment Algorithm: In
patients who present with an established sca-
phoid nonunion with evidence of AVN, our 7.5.4 Salvage Procedures
preferred method is compression screw fixation for Scaphoid Nonunion
with vascularized bone grafting using a 1,2 Advanced Collapse
ICSRA pedicled graft. However, in patients who (SNAC)
have had previous dorsal wrist surgery or trauma
or who have a significant humpback deformity With significant radiocarpal or midcarpal
that we do not feel would be correctable from the degenerative changes, surgical treatment of sca-
dorsal side, consideration is given to a volar phoid nonunion is unlikely to produce a mean-
approach and volar vascularized, pedicled graft ingful outcome. As mentioned earlier in the
[131]. Patients who present with very proximal chapter, scaphoid collapse leads to a DISI pattern
pole fractures are offered attempted fixation with of carpal instability, with degenerative changes
one of these techniques prior to any considera- progressing in a pattern that resembles scaphol-
tion of free vascularized reconstruction. unate ligamentous instability advanced collapse
(SLAC) [78]. Stage I SNAC degenerative chan-
7.5.3.1 Postoperative Management ges are confined to the radial styloid. Stage II
Following conventional or vascularized bone involves the entire radioscaphoid joint. In
graft for scaphoid nonunion, patients are initially Stage III, degenerative changes spread to the
placed in a short arm splint. This is converted to capitolunate joint, and in Stage IV, pancarpal
a short arm cast at 2 weeks that is maintained arthritis is seen.
until the 3 month postoperative mark or radio- Stage I wrists can be treated with scaphoid
graphic healing. At 3 months, patients are con- fixation and bone grafting combined with radial
verted to a removable short arm splint that is styloidectomy. With resection of the styloid, it is
removed only for gentle range of motion exer- important to protect volar radial ligament
cises with therapy. Splinting is continued until attachments. Resection of greater than 1 cm of
radiographs show fracture healing. Additionally, the radial styloid can result in destabilization of
any lifting with the affected extremity is delayed the origin of the radioscaphocapitate ligament
until union is apparent on radiographs. Patients [65].
should be counseled preoperatively that they will In Stage II, progression of degenerative
develop significant wrist stiffness and substantial changes to involve the entirety of the radio-
therapy will be required to regain strength and scaphoid joint necessitates the use of a motion
range of motion. They should also be warned that preserving salvage procedure. Excision of the
healing can be delayed in the setting of preop- distal scaphoid pole is an option in lower demand
erative radiographs, with average time to union patients with fractures of the scaphoid waist and
between 3 and 5 months post-surgery [128–131]. distal. Malerich et al. [145] reported improve-
Although it is not part of our practice, supple- ments in grip strength of 134% and range of
mental use of a bone stimulator (either motion of 85% in 19 patients treated with the
low-intensity pulsed ultrasound or electrical procedure. There was no radiographic progres-
stimulation) can be considered, especially in sion of DISI deformity at average of 4 years
postoperatively. Ruch et al. demonstrated satis- radiolunate joint after the procedure is per-
factory results in 13 patients at an average of formed, so it is important to confirm intraopera-
4 years after distal pole excision [146]. However, tively that the joint is without significant
6 of 13 had radiographic worsening of DISI degenerative changes. The procedure is also
deformity. For more active patients, proximal more technically demanding than PRC and
row carpectomy (PRC) and scaphoid excision obtaining consolidation of the fusion can be
with 4 corner fusion are surgical options. Com- challenging. Inadequate correction of the lunate
pared to fusion, proximal row carpectomy has the can lead to fusion of the capitate in a dorsally
advantages of being a less technically demanding subluxed position and painful radiocapitate
procedure, with no risk of nonunion, shorter abutment that prevents wrist extension [65].
postoperative period of immobilization, and Implant design is an area of ongoing research.
minimal postoperative rehabilitation. Compar- Techniques with Kirschner wires, circular plates,
ison of PRC to scaphoid excision and 4 corner compression screws, and staples have all been
fusion has demonstrated improved patient satis- utilized with the goal of developing rigid,
faction, grip strength, and range of motion in low-profile implants. An ideal implant will also
patients treated with PRC [147]. In relation to provide adequate bone compression and mini-
their contralateral extremity, patients can expect mize soft tissue irritation. Outcomes from the use
to regain 70–80% of their grip strength and 50– of circular plates have been mixed. Skie et al.
60% of their range of motion [147]. Radiographs [149] reported union of only 29 of 37 patients
will often demonstrate progression of arthritis to (78%). Despite achieving union in 10 of 11
the capitolunate joint, but this has little functional patients treated with a circular plate, Chung et al.
impact, as most patients remain asymptomatic found that the majority of patients continued to
[148]. experience persistent wrist pain and 1 patient
With progression of arthritis to the midcarpal required removal of painful hardware [150].
joint in Stage III, proximal row carpectomy is no Finally, Shindle et al. reported complications in 9
longer an option, and scaphoid excision and 4 of 16 patients treated with a circular plate,
corner fusion are indicated (Fig. 7.11). The including nonunion (25%), delayed union (6%),
articulation between the capitate, lunate, hamate, dorsal impingement (25%), radial styloid
and triquetrum is the location of the fusion, with impingement (6%), and broken screws (13%)
the goal of maintaining a congruent radiolunate [151].
joint. Biomechanical studies have demonstrated The degenerative involvement of the radiolu-
that load is preferentially transferred to the nate joint or presence of pancarpal arthritis in
Fig. 7.11 a Scaphoid nonunion advanced collapse stage III arthritic changes. b Treated with scaphoid excision and
four corner fusion
Stage IV is an indication for wrist arthrodesis. It like the scaphoid, the proximal pole of the cap-
should be noted that this is pain alleviating sal- itate is almost entirely intra-articular and the head
vage procedure with significant impact on wrist is nearly completely covered by articular carti-
function and the decision to proceed should be lage. The tenuous blood supply from the head
based on a collaborative discussion between the renders the capitate vulnerable to osteonecrosis
surgeon and patient. [155]. This is further complicated by the liga-
mentous instability of a perilunate dislocation
and resulting fracture instability.
7.5.5 Scaphocapitate Fracture If recognized acutely, these injuries should be
Syndrome treated with open reduction and internal fixation
of the capitate at the same time as definitive
The capitate is centered in the carpus where it is management of concomitant injuries including
well protected and is a rare site of carpal fracture. scaphoid fracture and perilunate dislocation. The
In fact, capitate fractures account for only 1–2% capitate can be approached through a standard
of all carpal fractures. A capitate fracture may dorsal approach between the third and fourth
occur in isolation, but more commonly is part of extensor compartments. The dorsal capsule
a trans-scaphoid perilunate fracture dislocation. should be incised though the dorsal capsule
Termed by Fenton as scaphocapitate syndrome, it splitting approach described earlier in the chap-
will result in a capitate nonunion if not correctly ter, with preservation and later repair of the DRC
diagnosed and adequately addressed [65, 152]. and DIC ligaments to the triquetrum. There will
With wrist hyperextension and an axial load often be a dorsal capsular tear present in the
resulting in fracture of the scaphoid, the capitate setting of a perilunate dislocation, and this should
is able to hyperextend. Impaction of the neck of be incorporated into the capsular approach
the capitate on the dorsal ridge of the distal whenever possible. Flexion of the wrist facili-
radius produces a transversely oriented fracture tates access to the head of the capitate. The
of the capitate neck. As the wrist returns to fracture is reduced and commonly stabilized with
neutral position, shortening of the carpus pre- 2 headless screws. Literature on scaphocapitate
vents reduction of the proximal fracture frag- syndrome consists entirely of small case series
ment. With further wrist flexion, the distal pole and case reports, the largest of which consists of
of the capitate fragment exerts a flexion force on three patients [153, 154, 156–160]. The consen-
the proximal fragment, causing the unstable sus of these reports is that in the absence of
proximal fragment to rotate 180°. The final pro- osteonecrosis, both acute and delayed open
duct is a malrotated fracture, in which the artic- reduction and internal fixation of these injuries
ular surface of the proximal pole faces distally in result in good short-term functional outcomes.
the fracture site [65, 152]. Clinical diagnosis can Treatment of capitate nonunion, especially
be challenging, as initial PA, lateral, and oblique with osteonecrosis, is less straightforward. In the
radiographs may have the appearance of a setting of delayed presentation, diagnosis, or
reduced capitate. As a result, these injuries may established nonunion, an MRI is useful to eval-
be missed initially and only diagnosed after uate for osteonecrosis of the proximal pole. The
fracture reabsorption occurs. Consequently, some goal of the treatment in capitate nonunion is not
surgeons recommend a low threshold to obtain a only fracture healing, but also re-establishment
CT scan for more definitive evaluation of capitate of capitate height [161]. Failure to do so will
irregularities in the setting of a greater arc injury result in overload of the scaphotrapezialtrapezoid
[153, 154]. and triquetral hamate joints and resulting early
The intraosseous blood supply of the capitate degenerative changes. A corticocancellous bone
is similar to the scaphoid, as dorsal and palmar graft may be necessary to correct capitate short-
blood vessels in the head provide vascular supply ening. When significant carpal degenerative
to the proximal pole by retrograde flow. Also, changes have developed, limited carpal fusion
may be required. Prior to collapse of the proxi- and hook. The incidence of limited blood flow to
mal pole, attempted restoration of blood flow the tip likely contributes to the risk osteonecrosis.
with vascularized bone grafts has been described, Initial clinical presentation usually consists of
but no long-term outcomes have been published chronic aching pain at the base of hypothenar
[155]. After collapse of the proximal pole, the eminence. The hook of the hamate forms the
problem is best managed by carpal fusion [65, lateral border of Guyon’s canal, which serves to
155]. transport the ulnar nerve and artery into the palm.
Patients will also frequently complain of ulnar
nerve paresthesias in the ring and small finger, as
7.5.6 Hook of the Hamate well as a loss of grip strength and exacerbation of
pain with grip activities. Some will describe a
Hamate fractures are uncommon injuries, specific injury, with initial acute pain changing to
accounting for only 2% of all carpal fractures a dull ache. Many will describe an inability to
[65]. Fractures of the hook of the hamate occur return to sports activities or decrease in perfor-
more frequently than hamate body fractures. mance, as a result of symptoms. Tenderness to
Hook of the hamate fractures are more common palpation will typically be present over the hook
in athletes than the general population, especially of the hamate, which is distal and radial to the
those involved in racquet sports, such as golf, pisiform. Given the close proximity of the ring
tennis, and baseball. Presentation following and small flexor tendons to the hook, pain is
injury may be delayed, and without acute often worsened with the hook of the hamate pull
immobilization, the risk of nonunion is test [164]. Resisted flexion of the ring and small
increased. finger tendons is applied. Pain is aggravated with
The anatomy of the hook of the hamate both resisted finger flexion in wrist ulnar deviation
places it at risk of injury and contributes to and improved with radial deviation. Infrequently,
fracture displacement and the potential for non- a chronic fracture will be identified only after
union. The hook protrudes volarly from the rupture of the profundus or superficialis tendons
hamate base into the hypothenar eminence where to ring or small fingers.
direct compression from the handle of a club or The injury is often difficult to identify on
bat can lead to fracture. The hook is the origin of standard PA, lateral, and oblique hand or wrist
the flexor digiti minimi, opponens digiti minimi, radiographs. The PA radiographs may provide
hypothenar muscles, and pisohamate ligaments some clues to an injury. These include sclerosis
and the distal insertion point of the transverse of the hamate in the area of the hook and absence
carpal ligament. Shear forces from each of these of the cortical ring sign in the distal, radial corner
muscles, as well as the adjacent flexor tendons of of the hamate, which represents the hook radio-
the ulnar digits, can all contribute to fracture graphically. With suspected injury based on
displacement. Vascular supply to the hamate has clinical history, the carpal tunnel view, supinated
been described by Failla and suggests why oblique view, and lateral view projected on the
fractures of the tip have an increased risk of first webspace with the thumb abducted have all
osteonecrosis compared to the body [162, 163]. been identified as tools to better visualize the
The interosseus blood supply to the body of the hamate [65]. The supinated oblique has the
hamate enters the radial base through a combi- highest sensitivity in clinical studies [165]. When
nation of three vascular pedicles, whereas the the diagnosis remains unclear, a CT scan is
hook is supplied by a nutrient vessel that origi- indicated [165, 166]. CT is the most effective of
nates from the ulnar artery and enters at the all modalities in diagnosing fracture, with a
tip. In Failla’s study, the nutrient vessel to the sensitivity of 100% and specificity and accuracy
hook of the hamate was only present in 71% of both greater than 90%. CT is also useful in ruling
specimens. Further, minimal collateral flow out a bipartite hamate, an anatomic variant pre-
exists between the nutrient arteries of the base sent in a minority of patients.
When diagnosed acutely, nondisplaced frac- and nerve should then be carefully mobilized
tures of the hook are commonly treated with from around the base of the hook. The deep
casting and immobilization. Studies have sug- motor branch to the nerve branches from the
gested that delayed presentation of injury leads to dorsal ulnar aspect of the ulnar nerve, passing
poor results with conservative management deep to the flexor digiti minimi. The motor
[167]. In the setting of an established nonunion, branch frequently lies directly adjacent to a
healing rates are unacceptably low with cast fracture located at the base of the hook. Once the
immobilization alone. Case reports have reported artery and nerve are mobilized and safely
successful healing of a nonunion with cast retracted without tension, the fracture is subpe-
immobilization and use of low-intensity pulsed riosteally excised from the body of the hamate.
ultrasound [168]. As discussed earlier, patients The remaining periosteum can be closed over the
with untreated nonunion are at risk of flexor fracture site, or the fracture surface can be
tendon rupture. This has lead to a trend toward smooth with a ronguer to prevent further irrita-
surgical treatment of symptomatic nonunions. tion of the ulnar nerve and flexor tendons.
Historically, excision of the nonunited fracture As mentioned previously, return to play and
fragment has been recommended. This is par- activities is faster after fragment excision than
ticularly attractive in athletes, as it may offer a operative fixation or closed management. Heal-
more rapid return to play in comparison with ing of the surgical incision and resolution of pain
other treatment methods. There is concern that with impact activities are often determine the
excision may result in a negative effect on flexor time of return to play. In the largest, study of
tendon function, as the hook functions as a ful- baseball players to date, average return to play
crum for the flexor tendons to the small finger. after excision was 5.7 weeks [170]. Scar sensi-
Biomechanical studies have demonstrated that tivity is a common complication, and postoper-
hook excision results in decreased flexor tendon ative physical therapy focuses on scar
force, with worsening of the effect as the wrist manipulation and desensitization. Athletes will
moves into a position of extension and ulnar often wear specially padded gloves with initial
deviation [169]. As a result, some authors have return to play.
proposed attempted open reduction and internal
fixation with hook plates or compression screws.
Attempted fixation with or without bone grafting 7.6 Metacarpal and Phalangeal
has only been studied in small series or single Nonunions
case reports in the literature. Despite adequate
healing rates, no clinical benefit has been Nonunion is an uncommon complication after
demonstrated from fixation over fragment exci- fractures of the metacarpals and phalanges [171].
sion [169]. The risk of nonunion is increased when other
The surgical approach to the hook of the complicating factors are involved such as sig-
hamate is similar to the one used for a decom- nificant soft tissue injury, neurovascular insult,
pression of Guyon’s canal. A curvilinear incision infection, or bone loss. Inadequate fixation, fail-
is directed over the fragment. Care should be ure of proper reduction and maintenance of
taken to not cross over the proximal palmar reduction, and excessive motion through fracture
flexion crease, in order to avoid scar contracture. fragments can prevent fractures from uniting
The antebrachial fascia of the palm is divided, [172].
and the ulnar nerve and artery are identified Recognizing the presence of a nonunion is the
proximally and followed into Guyon’s canal. The first step in treating the complication. By defi-
transverse carpal ligament, which attaches to the nition, nonunion or delayed union is a fracture
tip of the hook and comprises the floor of without clinical or radiographic signs of healing
Guyon’s canal, is released from the insertion by 4 months after injury or failure of bony union
point on the hook of the hamate. The ulnar artery by 6 months post-injury [171, 172]. The key to
this principle is understanding that radiographic the fracture are rarely indicated, and the digit
evidence of union often lags behind clinical often becomes a hindrance to the overall function
findings and should not be used as the sole of the hand if retained. Consideration should be
indicator of fracture healing. Once a nonunion or given to arthrodesis or amputation in these sce-
delayed union is identified, the causal factors narios, especially when the zone of injury is more
should be explored. Signs and symptoms of distal along the ray or involving the lesser digits.
infection must be closely evaluated. Bone loss Hypertrophic nonunions of the metacarpals
with segmental defects along the metacarpal or and phalanges are treated the same way as those
phalanx, a dysvascular wound bed with devital- involving other bones in the body. The etiology
ized bone, or significant crush injury with of this type of nonunion is inadequate stability of
insufficient soft tissue coverage, is all features of a fracture which inhibits the formation of bridg-
the mechanism of injury that hinder union. ing callus. Radiographic evidence of callus in the
However, it is often a failure of fixation that leads area of the fracture exists, but the fracture line
nonunion. The use of K wires alone in an remains clearly visible. The biology of the frac-
unstable fracture pattern or failure of hardware ture environment however is sufficient for bone
can lead to instability that prohibits healing formation and simply requires a more conducive
across fracture lines. construct for healing. In this setting, treatment
As with other nonunions and delayed unions, consists of increasing stability and compression
classifying the conditions as hypertrophic or through the fracture fragments to allow for bone
atrophic helps direct treatment. Hypertrophic healing. This is often achieved by compression
nonunions of the metacarpals and phalanges are screw fixation if fracture morphology is amen-
relatively rare, and most are instead atrophic able, followed by rigid neutralization plating, or
[173]. Hypertrophic nonunions are a result of rigid plating alone.
inadequate immobilization of fracture fragments Atrophic nonunions requires careful evalua-
resulting in excessive relative motion of the tion of contributing factors such as metabolic
fracture fragments, all in the setting of adequate disarray, poor nutrition, infection, gross insta-
blood supply and a healthy environment for bility, or failure to follow prescribed activity
formation of callus. This results in a robust restrictions during the initial
osteoblastic response without bridging fracture post-injury/operative period. Furthermore, seg-
lines. More frequently, however, nonunions of mental bone defects or devitalized bone at the
the metacarpals and phalanges are atrophic in fracture site are significant inhibitors of fracture
nature [173]. When an atrophic nonunion is seen union. After these factors are addressed, stable
in the hand, careful workup is indicated to define fixation with a plate-and-screw construct aug-
the etiology. Fracture biology is compromised in mented by bone graft to fill any bony deficiency
this scenario from factors such as infection, will maximize the odds of functional union [174].
neurovascular compromise, soft tissue stripping Fracture edges should be debrided, and any signs
or loss, general medical conditions of metabolic of fibrous bridging tissue or interposed avascular
disarray. bone should be removed prior to application of
the new construct. Importantly, however, all
patients must be counseled in regard to the high
7.6.1 Treatment likelihood of residual stiffness after this proce-
dure. Surgeons should greatly consider soft tissue
Nonoperative treatment of nonunion is indicated releases including tenocapsulolysis at the time of
in certain cases of significant soft tissue or nerve rigid fixation in order to maximize the odds of a
damage. When grossly contaminated crush inju- successful outcome and to facilitate early motion
ries or other mechanisms of severe fracture and for guided therapy regimens [175, 176]. Splint-
soft tissue compromise arise, a dysvascular digit ing to avoid contracture and appropriate edema
often results. Attempted fixation and coverage of control and analgesia to allow for the early
Table 7.1 Bone graft algorithm

Location nonunion Avascular necrosis (yes/no) Graft type
Scaphoid Yes 1,2 ICSRA, volar vascularized graft
No Volar distal radius, olecranon, iliac crest
Other carpal Yes 5-4 ECA, 2nd metacarpal artery
No Olecranon, iliac crest
Distal radius (small defect) – Olecranon, iliac crest cancellous
Distal radius (large defect) – Iliac crest tricortical graft
ICSRA intercompartmental supraretinacular artery
ECA extracompartmental artery
phases of therapy is vital to reaching a good 7. Ring D, Prommersberger KJ, Gonzalez del Pino J,
functional outcome. Capomassi M, Slullitel M, Jupiter JB. Corrective
osteotomy for intra-articular malunion of the distal
Author’s Preferred bone graft algorithm: The part of the radius. J Bone Joint Surg Am. 2005;87
type of bone graft used or harvested as well as (7):1503–9.
allograft versus autograft is determined by the 8. Rikli DA, Regazzoni P. Fractures of the distal end
type of nonunion as well as location of the of the radius treated by internal fixation and early
function. A preliminary report of 20 cases. J Bone
nonunion (Table 7.1). For example, hypertrophic Joint Surg Br. 1996;78(4):588–92.
nonunions may be able to be treated without graft 9. Leslie BM, Medoff RJ. Fracture specific fixation of
or with limited allograft, such as demineralized distal radius fractures. Tech Orthop. 2000;15:336–52.
bone matrix. However, it is our opinion that 10. Medoff RJ. Essential radiographic evaluation for
distal radius fractures. Hand Clin. 2005;21(3):279–
nonunions associated with segmental defects or 88.
voids or atrophic nonunions are best treated with 11. Melone CP Jr. Articular fractures of the distal
autograft. radius. Orthop Clin North Am. 1984;15(2):217–36.
12. Hak DJ, Fitzpatrick D, Bishop JA, Marsh JL, Tilp S,
Schnettler R, Simpson H, Alt V. Delayed union and
nonunions: epidemiology, clinical issues, and finan-
cial aspects. Injury. 2014;45(Suppl 2):S3–7.
References 13. Bushnell BD, Bynum DK. Malunion of the distal
radius. J Am Acad Orthop Surg. 2007;15(1):27–40.
1. Fernandez DL, Ring D, Jupiter JB. Surgical man- 14. Adams BD, Grosland NM, Murphy DM, McCul-
agement of delayed union and nonunion of distal lough M. Impact of impaired wrist motion on hand
radius fractures. J Hand Surg Am. 2001;26(2):201–9. and upper-extremity performance. J Hand Surg Am.
2. Segalman KA, Clark GL. Un-united fractures of the 2003;28(6):898–903.
distal radius: a report of 12 cases. J Hand Surg Am. 15. Prommersberger KJ, Fernandez DL, Ring D,
1998;23(5):914–9. Jupiter JB, Lanz UB. Open reduction and internal
3. Wolfe SW. Distal radius fractures. In: Wolfe SW, fixation of un-united fractures of the distal radius:
Pederson WC, Hotchkiss RN, Kozin SH, editors. does the size of the distal fragment affect the result?
Green’s operative hand surgery. 6th ed. Philadel- Chir Main. 2002;21(2):113–23.
phia: Elsevier; 2010. p. 561–638. 16. Ring D. Nonunion of the distal radius. Hand Clin.
4. McKee MD, Waddell JP, Yoo D, Richards RR. 2005;21(3):443–7.
Nonunion of distal radial fractures associated with 17. Gwathmey FW Jr, Brunton LM, Pensy RA,
distal ulnar shaft fractures: a report of four cases. Chhabra AB. Volar plate osteosynthesis of distal
J Orthop Trauma. 1997;11(1):49–53. radius fractures with concurrent prophylactic carpal
5. Prommersberger KJ, Fernandez DL, Ring D, tunnel release using a hybrid flexor carpi radialis
Jupiter JB, Lanz UB. Open reduction and internal approach. J Hand Surg Am. 2010;35(7):1082–8.
fixation of un-united fractures of the distal radius: 18. Ring D, Jupiter JB. Nonunion of the distal radius.
does the size of the distal fragment affect the result? Tech Hand Up Extrem Surg. 2002;6(1):6–9.
Chir Main. 2002;21(2):113–23. 19. Harness NG, Jupiter JB, Orbay JL, Raskin KB,
6. Graham TJ. Surgical correction of malunited frac- Fernandez DL. Loss of fixation of the volar lunate
tures of the distal radius. J Am Acad Orthop Surg. facet fragment in fractures of the distal part of the
1997;5(5):270–81. radius. J Bone Joint Surg Am. 2004;86-A(9):1900–8.
20. Jupiter JB, Marent-Huber M; LCP Study 34. Wallwork NA, Bain GI. Sigmoid notch osteoplasty
Group. Operative management of distal radial for the chronic volar instability of the distal
fractures with 2.4-millimeter locking plates. A mul- radioulnar joint: a case report. J Hand Surg Am.
ticenter prospective case series. J Bone Joint Surg 2001;26(3):454–9.
Am. 2009;91(1):55–65. 35. Sagerman SD, Zogby RG, Palmer AK, Werner FW,
21. Jupiter JB, Fernandez DL. Complications following Fortino MD. Relative articular inclination of the
distal radial fractures. J Bone Joint Surg Am. distal radioulnar joint: a radiographic study. J Hand
2001;83:1244–65. Surg Am. 1995;20(4):597–601.
22. Bakker AJ, Shin AY. Fragment-specific volar hook 36. Waitayawinyu T, McCallister WV, Nemechek NM,
plate for volar marginal rim fractures. Tech Hand Trumble TE. Scaphoid nonunion. J Am Acad
Up Extrem Surg. 2014;18(1):56–60. Orthop Surg. 2007;15(5):308–20.
23. Mithani SK, Srinivasan RC, Kamal R, Richard MJ, 37. Palmer AK, Werner FW. The triangular fibrocarti-
Leversedge FJ, Ruch DS. Salvage of distal radius lage complex of the wrist—anatomy and function.
nonunion with a dorsal spanning distraction plate. J Hand Surg Am. 1981;6(2):153–62.
J Hand Surg Am. 2014;39(5):981–4. 38. Gofton WT, Gordon KD, Dunning CE, Johnson JA,
24. Hanel DP, Ruhlman SD, Katolik LI, Allan CH. King GJ. Soft-tissue stabilizers of the distal radioul-
Complications associated with distraction plate nar joint: an in vitro kinematic study. J Hand Surg
fixation of wrist fractures. Hand Clin. 2010;26 Am. 2004;29(3):423–31.
(2):237–43. 39. Adams BD, Lawler E. Chronic instability of the
25. Stucken C, Olszewski DC, Creevy WR, distal radioulnar joint. J Am Acad Orthop Surg.
Murakami AM, Tornetta P. Preoperative diagnosis 2007;15(9):571–5.
of infection in patients with nonunions. J Bone Joint 40. Ward LD, Ambrose CG, Masson MV, Levaro F.
Surg Am. 2013;95(15):1409–12. The role of the distal radioulnar ligaments, inter-
26. Jones AL, Bucholz RW, Bosse MJ, Mirza SK, osseus membrane, and joint capsule in distal
Lyon TR, Webb LX, et al; BMP-2 Evaluation in radioulnar joint stability. J Hand Surg Am.
Surgery for Tibial Trauma-Allgraft (BESTT-ALL) 2000;25(2):341–51.
Study Group. Recombinant human BMP-2 and 41. Adams BD, Holley KA. Strains in the articular disk
allograft compared with autogenous bone graft for of the triangular fibrocartilage complex: a biome-
reconstruction of diaphyseal tibial fractures with chanical study. J Hand Surg Am. 1993;18(5):919–
cortical defects. A randomized, controlled trial. 25.
J Bone Joint Surg Am. 2006;88(7):1431–41. 42. Schuind F, An KN, Berglund L, Rey R,
27. Aro HT, Govender S, Patel AD, Hernigou P, Perera Cooney WP III, Linscheid RL, Chao EY. The
de Gregorio A, Popescu GI, Golden JD, Chris- distal radioulnar ligaments: a biomechanical study.
tensen J, Valentin A. Recombinant human bone J Hand Surg Am. 1991;16(6):1106–14.
morphogenetic protein-2: a randomized trial in open 43. Ekenstam FAF, Hagert CG. Anatomical studies of
tibial fractures treated with reamed nail fixation. the geometry and stability of the distal radioulnar
J Bone Joint Surg Am. 2011;93(9):801–8. joint. Scand J Plast Reconstr Surg. 1985;19(1):17–
28. Adams BD. Distal radioulnar joint. In: Wolfe SW, 25.
Pederson WC, Hotchkiss RN, Kozin SH, editors. 44. Chidgey LK, Dell PC, Bittar ES, Spanier SS.
Green’s operative hand surgery. 6th ed. Philadel- Histologic anatomy of the triangular fibrocartilage.
phia: Elsevier; 2010. p. 523–60. J Hand Surg Am. 1991;16(6):1084–100.
29. May MM, Lawton JN, Blazar PE. Ulnar styloid 45. Zimmerman RM, Kim JM, Jupiter JB. Arthritis of
fractures associated with distal radius fractures: the distal radioulnar joint: from Darrach to total
incidence and implications for distal radioulnar joint joint arthroplasty. J Am Acad Orthop Surg. 2012;20
instability. J Hand Surg Am. 2002;27(6):965–71. (10):623–32.
30. Ekenstam F. Anatomy of the distal radioulnar joint. 46. Epner RA, Bowers WH, Guilford WB. Ulnar
Clin Orthop Relat Res. 1992;(275):14–8. variance—the effect of wrist positioning and roent-
31. Adams BD, Berger RA. An anatomic reconstruction gen filming technique. J Hand Surg Am. 1982;7
of the distal radioulnar ligaments for posttraumatic (3):298–305.
distal radioulnar joint instability. J Hand Surg Am. 47. Friedman SL, Palmer AK, Short WH, Levin-
2002;27(2):243–51. sohn EM, Halperin LS. The change in ulnar
32. Stuart PR, Berger RA, Linscheid RL, An KN. The variance with grip. J Hand Surg Am. 1993;18
dorsopalmar stability of the distal radioulnar joint. (4):713–6.
J Hand Surg Am. 2000;25(4):689–99. 48. Tomaino MM. The importance of the pronated grip
33. Tolat AR, Stanley JK, Trail IA. A cadaveric study X-ray view in evaluating ulnar variance. J Hand
of the anatomy and stability of the distal radioulnar Surg Am. 2000;25(2):352–7.
joint in the coronal and transverse planes. J Hand 49. Mino DE, Palmer AK, Levinsohn EM. Radiography
Surg Br. 1996;21(5):587–94. and computerized tomography in the diagnosis of
incongruity of the distal radio-ulnar joint: a 65. Geissler WB, Slade JF. Fractures of the carpal
prospective study. J Bone Joint Surg Am. 1985;67 bones. In: Wolfe SW, Pederson WC, Hotchkiss RN,
(2):247–52. Kozin SH, editors. Green’s operative hand surgery.
50. Scheker LR, Belliappa PP, Acosta R, German DS. 6th ed. Philadelphia: Elsevier; 2010. p. 639–707.
Reconstruction of the dorsal ligament of the trian- 66. Wolf JM, Dawson L, Mountcastle SB, Owens BD.
gular fibrocartilage complex. J Hand Surg Br. The incidence of scaphoid fracture in a military
1994;19(3):310–8. population. Injury. 2009;40(12):1316–9.
51. Lo IK, MacDermid JC, Bennett JD, Bogoch E, 67. Doornberg JN, Buijze GA, Ham SJ, Ring D,
King GJ. The radioulnar ratio: a new method of Bhandari M, Poolman RW. Nonoperative treatment
quantifying distal radioulnar joint subluxation. for acute scaphoid fractures: a systematic review
J Hand Surg Am. 2001;26(2):236–43. and meta-analysis of randomized controlled trials.
52. Wechsler RJ, Wehbe MA, Rifkin MD, Edeiken J, J Trauma. 2011;71(4):1073–81.
Branch HM. Computed tomography diagnosis of 68. Singh HP, Taub N, Dias JJ. Management of
radioulnar subluxation. Skeletal Radiol. 1987;16 displaced fractures of the waist of the scaphoid:
(1):1–5. meta-analyses of comparative studies. Injury.
53. Amrami KK, Moran SL, Berger RA, Ehman EC, 2012;43(6):933–9.
Felmlee JP. Imaging the distal radioulnar joint. 69. Yin ZG, Zhang JB, Kan SL, Wang P. Treatment of
Hand Clin. 2010;26(4):467–75. acute scaphoid fractures: systematic review and
54. Millard GM, Budoff JE, Paravic V, Noble PC. meta-analysis. Clin Orthop Relat Res.
Functional bracing for distal radioulnar joint insta- 2007;460:142–51.
bility. J Hand Surg. 2002;27(6):972–7. 70. Symes TH, Stothard J. A systematic review of the
55. Frykman G. Fracture of the distal radius including treatment of acute fractures of the scaphoid. J Hand
sequelae—shoulder-hand-finger syndrome, distur- Surg Eur. 2011;36(9):802–10.
bance in the distal radio-ulnar joint and impairment 71. Suh N, Benson EC, Faber KJ, Macdermid J,
of nerve function. A clinical and experimental Grewal R. Treatment of acute scaphoid fractures:
study. Acta Orthop Scand. 1967:Suppl 108:3+. a systematic review and meta-analysis. Hand (N Y).
56. Geissler WB, Fernandez DL, Lamey DM. Distal 2010;5(4):345–53.
radioulnar joint injuries associated with fractures of 72. Ibrahim T, Qureshi A, Sutton AJ, Dias JJ. Surgical
the distal radius. Clin Orthop Relat Res. versus nonsurgical treatment of acute minimally
1996;327:135–46. displaced and undisplaced scaphoid waist fractures:
57. Hauck RM, Skahen J III, Palmer AK. Classification Pairwise and network meta-analyses of randomized
and treatment of ulnar styloid nonunion. J Hand controlled trials. J Hand Surg Am. 2011;36
Surg Am. 1996;21(3):418–22. (11):1759–68, e1.
58. Peterson MS, Adams BD. Biomechanical evalua- 73. Heinzelmann AD, Archer G, Bindra RR. Anthro-
tion of distal radioulnar reconstructions. J Hand pometry of the human scaphoid. J Hand Surg Am.
Surg. 1993;18(2):328–34. 2007;32(7):1005–8.
59. Hui FC, Linscheid RL. Ulnotriquetral augmentation 74. Pichler W, Windisch G, Schaffler G, Heidari N,
tenodesis: a reconstructive procedure for dorsal Dorr K, Grechenig W. Computer-assisted
subluxation of the distal radioulnar joint. J Hand 3-dimensional anthropometry of the scaphoid.
Surg Am. 1982;7(3):230–6. Orthopedics. 2010;33(2):85–8.
60. Boyes JH. Surgical repair of joints. In: Bunnell S, 75. Gelberman RH, Menon J. The vascularity of the
editor. Bunnell’s surgery of the hand. Philadelphia: scaphoid bone. J Hand Surg Am. 1980;5(5):508–
Lippincott; 1970. p. 294–313. 13.
61. Johnston Jones K, Sanders WE. Posttraumatic 76. Weber ER, Chao EY. An experimental approach to
radioulnar instability: treatment by anatomic recon- the mechanism of scaphoid waist fractures. J Hand
struction of volar and radioulnar ligaments. Orthop Surg Am. 1978;3(2):142–8.
Trans. 1995–1996;19:832. 77. Horii E, Nakamura R, Watanabe K, Tsunoda K.
62. Kihara H, Short WH, Werner FW, Fortino MD, Scaphoid fracture as a “puncher’s fracture”.
Palmer AK. The stabilizing mechanism of the distal J Orthop Trauma. 1994;8(2):107–10.
radioulnar joint during pronation and supination. 78. Mack GR, Bosse MJ, Gelberman RH, Yu E. The
J Hand Surg Am. 1995;20(6):930–6. natural history of scaphoid non-union. J Bone Joint
63. Wallwork NA, Bain GI. Sigmoid notch osteoplasty Surg Am. 1984;66(4):504–9.
for chronic volar instability of the distal radioulnar 79. Little CP, Burston BJ, Hopkinson-Woolley J,
joint: a case report. J Hand Surg Am. 2001;26 Burge P. Failure of surgery for scaphoid
(3):454–9. non-union is associated with smoking. J Hand Surg
64. Tham SK, Bain GI. Sigmoid notch osseous recon- Br. 2006;31(3):252–5.
struction. Tech Hand Up Extrem Surg. 2007;11 80. Nolte PA, van der Krans A, Patka P, Janssen IM,
(1):93–7. Ryaby JP, Albers GH. Low-intensity pulsed
ultrasound in the treatment of nonunions. J Trauma. electromagnetic fields: a study continuation.

2001;51(4):693–702; discussion 702-3. J Hand Surg Am. 1992;17(5):910–4.
81. Trumble TE, Clarke T, Kreder HJ. Non-union of the 96. Rettig AC, Weidenbener EJ, Gloyeske R. Alterna-
scaphoid: treatment with cannulated screws com- tive management of midthird scaphoid fractures in
pared with treatment with Herbert screws. J Bone the athlete. Am J Sports Med. 1994;22(5):711–4.
Joint Surg Am. 1996;78(12):1829–37. 97. Davis EN, Chung KC, Kotsis SV, Lau FH, Vijan S.
82. Jupiter JB, Shin AY, Trumble TE, Fernandez DL. A cost/utility analysis of open reduction and internal
Traumatic and reconstructive problems of the fixation versus cast immobilization for acute
scaphoid. Instr Course Lect. 2001;50:105–22. nondisplaced mid-waist scaphoid fractures. Plast
83. Trumble TE, Salas P, Barthel T, Robert KQ III. Reconstr Surg. 2006;117(4):1223–35.
Management of scaphoid nonunions. J Am Acad 98. Bunker TD, McNamee PB, Scott TD. The Herbert
Orthop Surg. 2003;11(6):380–91. screw for scaphoid fractures: a multicentre study.
84. Buijze GA, Wijffels MM, Guitton TG, Grewal R, J Bone Joint Surg Br. 1987;69(4):631–4.
van Dijk CN, Ring D; Science of Variation 99. Hunter JC, Escobedo EM, Wilson AJ, Hanel DP,
Group. Interobserver reliability of computed tomog- Zink-Brody GC, Mann FA. MR imaging of clini-
raphy to diagnose scaphoid waist fracture union. cally suspected scaphoid fractures. AJR Am J
J Hand Surg Am. 2012;37(2):250–4. Roentgenol. 1997;168(5):1287–93.
85. Moon ES, Dy CJ, Derman P, Vance MC, Carl- 100. Kuhlmann JN, Mimoun M, Boabighi A, Baux S.
son MG. Management of nonunion following Vascularized bone graft pedicled on the volar carpal
surgical management of scaphoid fractures: current artery for non-union of the scaphoid. J Hand Surg
concepts. J Am Acad Orthop Surg. 2013;21(9):548– Br. 1987;12(2):203–10.
57. 101. Sheetz KK, Bishop AT, Berger RA. The arterial
86. Amadio PC, Berquist TH, Smith DK, Ilstrup DM, blood supply of the distal radius and ulna and its
Cooney WP III, Linscheid RL. Scaphoid malunion. potential use in vascularized pedicled bone grafts.
J Hand Surg Am. 1989;14(4):679–87. J Hand Surg Am. 1995;20(6):902–14.
87. Bain GI, Bennett JD, Richards RS, Slethaug GP, 102. Stark A, Brostrom LA, Svartengren G. Surgical
Roth JH. Longitudinal computed tomography of the treatment of scaphoid non-union: review of the
scaphoid: a new technique. Skeletal Radiol. literature and recommendations for treatment. Arch
1995;24(4):271–3. Orthop Trauma Surg. 1989;108(4):203–9.
88. Bain GI, Bennett JD, MacDermid JC, Slethaug GP, 103. Korkala OL, Kuokkanen HO, Eerola MS.
Richards RS, Roth JH. Measurement of the Compression-staple fixation for fractures,
scaphoid humpback deformity using longitudinal non-unions, and delayed unions of the carpal
computed tomography: intra- and interobserver scaphoid. J Bone Joint Surg Am. 1992;74(3):423–6.
variability using various measurement techniques. 104. Huene DR, Huene DS. Treatment of nonunions of
J Hand Surg Am. 1998;23(1):76–81. the scaphoid with the Ender compression blade
89. Smith ML, Bain GI, Chabrel N, Turner P, Carter C, plate system. J Hand Surg Am. 1991;16(5):913–22.
Field J. Using computed tomography to assist with 105. Leixnering M, Pezzei C, Weninger P, Mayer M,
diagnosis of avascular necrosis complicating Bogner R, Lederer S, et al. First experiences with a
chronic scaphoid nonunion. J Hand Surg Am. new adjustable plate for osteosynthesis of scaphoid
2009;34(6):1037–43. nonunions. J Trauma. 2011;71(4):933–8.
90. Morgan WJ, Breen TF, Coumas JM, Schulz LA. 106. Reigstad O, Thorkildsen R, Grimsgaard C,
Role of magnetic resonance imaging in assessing Reigstad A, Røkkum M. Is revision bone grafting
factors affecting healing in scaphoid nonunions. worthwhile after failed surgery for scaphoid non-
Clin Orthop Relat Res. 1997;336:240–6. union? Minimum 8 year follow-up of 18 patients.
91. Perlik PC, Guilford WB. Magnetic resonance J Hand Surg Eur. 2009;34(6):772–7.
imaging to assess vascularity of scaphoid non- 107. Herbert TJ, Fisher WE. Management of the frac-
unions. J Hand Surg Am. 1991;16(3):479–84. tured scaphoid using a new bone screw. J Bone
92. Sakuma M, Nakamura R, Imaeda T. Analysis of Joint Surg Br. 1984;66(1):114–23.
proximal fragment sclerosis and surgical outcome 108. Faucher GK, Golden ML III, Sweeney KR, Hut-
of scaphoid non-union by magnetic resonance ton WC, Jarrett CD. Comparison of screw trajectory
imaging. J Hand Surg Br. 1995;20(2):201–5. on stability of oblique scaphoid fractures: a
93. Trumble TE. Avascular necrosis after scaphoid mechanical study. J Hand Surg Am. 2014;39
fracture: a correlation of magnetic resonance imag- (3):430–5.
ing and histology. J Hand Surg Am. 1990;15 109. McCallister WV, Knight J, Kaliappan R, Trumble T.
(4):557–64. Central placement of the screw in simulated frac-
94. Roolker W, Ritt MJ, Bos KE. Spontaneous healing tures of the scaphoid waist. J Bone Joint Surg Am.
of a non-union of the scaphoid. J Hand Surg Br. 2003;85(1):72–8.
1998;23(1):86–7. 110. Trumble TE, Gilbert M, Murray LW, Smith J,
95. Adams BD, Frykman GK, Taleisnik J. Treatment of Rafijah G, McCallister WV. Displaced scaphoid
scaphoid nonunion with casting and pulsed fractures treated with open reduction and internal
fixation with a cannulated screw. J Bone Joint Surg scaphoid nonunion with avascular necrosis: a
Am. 2000;82(5):633–41. canine experimental study. J Hand Surg Am.
111. Shah J, Jones WA. Factors affecting the outcome in 2000;25(5):849–59.
50 cases of scaphoid nonunion treated with Herbert 128. Zaidemberg C, Siebert JW, Angrigiani C. A new
screw fixation. J Hand Surg Br. 1998;23(5):680–5. vascularized bone graft for scaphoid nonunion.
112. Slade JF III, Geissler WB, Gutow AP, Merrell GA. J Hand Surg Am. 1991;16(3):474–8.
Percutaneous internal fixation of selected scaphoid 129. Chang MA, Bishop AT, Moran SL, Shin AY. The
nonunions with an arthroscopically assisted dorsal outcomes and complications of
approach. J Bone Joint Surg Am. 2003;85-A(Suppl 1,2-intercompartmental supraretinacular artery pedi-
4):20–32. cled vascularized bone grafting of scaphoid non-
113. Ikeda K, Osamura N, Tomita K. Percutaneous unions. J Hand Surg Am. 2006;31(3):387–96.
screw fixation without bone graft for cystic-type 130. Straw RG, Davis TR, Dias JJ. Scaphoid nonunion:
scaphoid fractures. J Trauma. 2008;65(6):1453–8. treatment with a pedicled vascularized bone graft
114. Slade JF III, Gillon T. Retrospective review of 234 based on the 1,2 intercompartmental supraretinac-
scaphoid fractures and nonunions treated with ular branch of the radial artery. J Hand Surg Br.
arthroscopy for union and complications. Scand J 2002;27(5):413.
Surg. 2008;97(4):280–9. 131. Gras M, Mathoulin C. Vascularized bone graft
115. Russe O. Fractures of the carpal navicular. Diag- pedicled on the volar carpal artery from the volar
nosis, non-operative treatment, and operative treat- distal radius as primary procedure for scaphoid
ment. J Bone Joint Surg Am. 1960;42-A:759–68. non-union. Orthop Traumatol Surg Res. 2011;97
116. Green DP. The effect of avascular necrosis on Russe (8):800–6.
bone grafting for scaphoid nonunion. J Hand Surg 132. Mathoulin C, Brunelli F. Further experience with
Am. 1985;10(5):597–605. the index metacarpal vascularized bone graft.
117. Dacho A, Germann G, Sauerbier M. The recon- J Hand Surg Br. 1998;23(3):311–7.
struction of scaphoid pseudoarthroses with the 133. Sawaizumi T, Nanno M, Nanbu A, Ito H. Vascu-
operation of Matti-Russe. A retrospective larised bone graft from the base of the second
follow-up analysis of 84 patients. Unfallchirurg. metacarpal for refractory nonunion of the scaphoid.
2004;107(5):388–96 (article in German). J Bone Joint Surg Br. 2004;86(7):1007–12.
118. Barton NJ. Experience with scaphoid grafting. 134. Guimberteau JC, Panconi B. Recalcitrant non-union
J Hand Surg Br. 1997;22(2):153–60. of the scaphoid treated with a vascularized bone
119. Berger RA, Bishop AT, Bettinger PC. New dorsal graft based on the ulnar artery. J Bone Joint Surg
capsulotomy for the surgical exposure of the wrist. Am. 1990;72(1):88–97.
Ann Plast Surg. 1995;35(1):54–9. 135. Jones DB Jr, Moran SL, Bishop AT, Shin AY.
120. Inoue G, Shionoya K, Kuwahata Y. Ununited Free-vascularized medial femoral condyle bone
proximal pole scaphoid fractures. Treatment with transfer in the treatment of scaphoid nonunions.
a Herbert screw in 16 cases followed for 0.5-8 Plast Reconstr Surg. 2010;125(4):1176–84.
years. Acta Orthop Scand. 1997;68(2):124–7. 136. Arora R, Lutz M, Zimmermann R, Krappinger D,
121. Garcia RM, Leversedge FJ, Aldridge JM, Niederwanger C, Gabl M. Free vascularised iliac
Richard MJ, Ruch DS. Scaphoid nonunions treated bone graft for recalcitrant avascular nonunion of the
with 2 headless compression screws and bone scaphoid. J Bone Joint Surg Br. 2010;92(2):224–9.
grafting. J Hand Surg Am. 2014;39(7):1301–7. 137. Al-Jabri T, Mannan A, Giannoudis P. The use of the
122. Manske PR, McCarthy JA, Strecker WB. Use of the free vascularised bone graft for nonunion of the
Herbert bone screw for scaphoid nonunions. Ortho- scaphoid: a systematic review. J Orthop Surg Res.
pedics. 1988;11(12):1653–61. 2014;9:21.
123. Fernandez DL. Anterior bone grafting and conven- 138. Sandow MJ. Costo-osteochondral grafts in the
tional lag screw fixation to treat scaphoid non- wrist. Tech Hand Up Extrem Surg. 2001;5
unions. J Hand Surg Am. 1990;15(1):140–7. (3):165–72.
124. Eggli S, Fernandez DL, Beck T. Unstable scaphoid 139. Veitch S, Blake SM, David H. Proximal scaphoid
fracture nonunion: a medium-term study of anterior rib graft arthroplasty. J Bone Joint Surg Br. 2007;89
wedge grafting procedures. J Hand Surg Br. (2):196–201.
2002;27(1):36–41. 140. Tropet Y, Lepage D, Gallinet D, Obert L, Garbuio P,
125. Inoue G, Shionoya K. Herbert screw fixation by Vichard P. Articular reconstructions by a costo-
limited access for acute fractures of the scaphoid. chondral grafting (or osteochondral costal grafting).
J Bone Joint Surg Br. 1997;79(3):418–21. Bull Acad Natl Med. 2006;190(7):1439–56 (article
126. Robbins RR, Ridge O, Carter PR. Iliac crest bone in French).
grafting and Herbert screw fixation of nonunions of 141. Sunagawa T, Bishop AT, Muramatsu K. Role of
the scaphoid with avascular proximal poles. J Hand conventional and vascularized bone grafts in
Surg Am. 1995;20(5):818–31. scaphoid nonunion with avascular necrosis: a
127. Sunagawa T, Bishop AT, Muramatsu K. Role of canine experimental study. J Hand Surg Am.
conventional and vascularized bone grafts in 2000;25(5):849–59.
142. Fernandez DL, Eggli S. Non-union of the scaphoid: fracture: a variant of scaphocapitate fractures. BMJ
revascularization of the proximal pole with implan- Case Rep. 2014 Mar 31;2014. pii: bcr2014204041.
tation of a vascular bundle and bone-grafting. 157. Natera Cisneros L, Lamas Gómez C, Proubasta
J Bone Joint Surg Am. 1995;77(6):883–93. Renart I, Moya Gómez E. Fenton syndrome. Rev
143. Smith BS, Cooney WP. Revision of failed bone Esp Cir Ortop Traumatol. 2012;56(5):369–73 (arti-
grafting for nonunion of the scaphoid: treatment cle in Spanish).
options and results. Clin Orthop Relat Res. 158. Schliemann B, Langer M, Kösters C, Raschke MJ,
1996;327:98–109. Ochman S. Successful delayed surgical treatment of
144. Hori Y, Tamai S, Okuda H, Sakamoto H, Takita T, a scaphocapitate fracture. Arch Orthop Trauma
Masuhara K. Blood vessel transplantation to bone. Surg. 2011;131(11):1555–9.
J Hand Surg Am. 1979;4(1):23–33. 159. Arbter D, Piatek S, Wichlas F, Winckler S. The
145. Malerich MM, Clifford J, Eaton B, Eaton R, scaphocapitate fracture syndrome (Fenton). Hand-
Littler JW. Distal scaphoid resection arthroplasty chir Mikrochir Plast Chir. 2009;41(3):171–4 (article
for the treatment of degenerative arthritis secondary in German).
to scaphoid nonunion. J Hand Surg Am. 1999;24 160. Sawant M, Miller J. Scaphocapitate syndrome in an
(6):1196–205. adolescent. J Hand Surg Am. 2000;25(6):1096–9.
146. Ruch DS, Papadonikolakis A. Resection of the 161. Geissler WB. Carpal fractures in athletes. Clin
scaphoid distal pole for symptomatic scaphoid Sports Med. 2001;20(1):167–88.
nonunion after failed previous surgical treatment. 162. Failla JM. Osteonecrosis associated with nonunion
J Hand Surg Am. 2006;31(4):588–93. of the hook of the hamate. Orthopedics. 1993;16
147. Wyrick JD, Stern PJ, Kiefhaber TR. (2):217–8.
Motion-preserving procedures in the treatment of 163. Failla JM. Hook of hamate vascularity: vulnerability
scapholunate advanced collapse wrist: proximal row to osteonecrosis and nonunion. J Hand Surg Am.
carpectomy versus four-corner arthrodesis. J Hand 1993;18(6):1075–9.
Surg Am. 1995;20(6):965–70. 164. Pajares-López M, Hernández-Cortés P,
148. Mulford JS, Ceulemans LJ, Nam D, Axelrod TS. Robles-Molina MJ. Rupture of small finger flexor
Proximal row carpectomy vs four corner fusion for tendons secondary to asymptomatic nonunion of the
scapholunate (Slac) or scaphoid nonunion advanced hamate hook. Orthopedics. 2011;34(2):142.
collapse (SNAC) wrists: a systematic review of 165. Andresen R, Radmer S, Scheufler O, Adam C,
outcomes. J Hand Surg Eur. 2009;34(2):256–63. Bogusch G. Optimization of conventional X-ray
149. Skie MC, Gove N, Ciocanel DE, Smith H. Man- images for the detection of hook of hamate
agement of non-united four-corner fusions. Hand (N fractures. Rontgenpraxis. 2006;56(2):59–65.
Y). 2007;2(1):34–8. 166. Andresen R, Radmer S, Sparmann M, Bogusch G,
150. Chung KC, Watt AJ, Kotsis SV. A prospective Banzer D. Imaging of hamate bone fractures in
outcomes study of four-corner wrist arthrodesis conventional X-rays and high-resolution computed
using a circular limited wrist fusion plate for stage II tomography. An in vitro study. Invest Radiol.
scapholunate advanced collapse wrist deformity. 1999;34(1):46–50.
Plast Reconstr Surg. 2006;118(2):433–42. 167. Carroll RE, Lakin JF. Fracture of the hook of the
151. Shindle MK, Burton KJ, Weiland AJ, Domb BG, hamate: acute treatment. J Trauma. 1993;34
Wolfe SW. Complications of circular plate fixation (6):803–5.
for four-corner arthrodesis. J Hand Surg Eur. 168. Sakuma Y, Iwamoto T, Momohara S. Ununited
2007;32(1):50–3. fracture of the hook of hamate treated with
152. Fenton RL. The naviculo-capitate fracture syn- low-intensity pulsed ultrasound in an older
drome. J Bone Joint Surg Am. 1956;38-A(3):681–4. middle-aged patient. Clin J Sport Med. 2014;24
153. Kim YS, Lee HM, Kim JP. The scaphocapitate (4):358–9.
fracture syndrome: a case report and literature 169. Demirkan F, Calandruccio JH, Diangelo D. Biome-
analysis. Eur J Orthop Surg Traumatol. 2013;23 chanical evaluation of flexor tendon function after
(Suppl 2):S207–12. hamate hook excision. J Hand Surg Am. 2003;28
154. Hamdi MF. The scaphocapitate fracture syndrome: (1):138–43.
report of a case and a review of the literature. 170. Bachoura A, Wroblewski A, Jacoby SM, Oster-
Musculoskelet Surg. 2012;96(3):223–6. man AL, Culp RW. Hook of hamate fractures in
155. Vander Grend R, Dell PC, Glowczewskie F, Les- competitive baseball players. Hand (N Y). 2013;8
lie B, Ruby LK. Intraosseous blood supply of the (3):302–7.
capitate and its correlation with aseptic necrosis. 171. Green DP. Complications of phalangeal and
J Hand Surg Am. 1984;9(5):677–83. metacarpal fractures. Hand Clin. 1986;2(2):307–28.
156. Burke NG, Cosgrave CH, O’Neill BJ, 172. Jupiter JB, Koniuch MP, Smith RJ. The manage-
Kelly EP. Transstyloid, transscaphoid, transcapitate ment of delayed union and nonunion of the
metacarpals and phalanges. J Hand Surg. 1985;10 175. Weckesser EC. Rotational osteotomy of the meta-
(4):457–66. carpal for overlapping fingers. J Bone Joint Surg.
173. Ring D. Malunion and nonunion of the metacarpals 1965;47(4):751–6.
and phalanges. J Bone Joint Surg Am. 176. Creighton JJ, Steichen JB. Complications in pha-
2005;87:1380–8. langeal and metacarpal fracture management.
174. Smith FL, Rider DL. A study of the healing of one Results of extensor tenolysis. Hand Clin. 1994;10
hundred consecutive phalangeal fractures. J Bone (1):111–6.
Joint Surg. 1935;17:91–109.
Acetabular and Pelvic Nonunions
8
Kyle F. Dickson, MD, MBA
motion of the pelvis can be detected in this

8.1 Nonunion of the Pelvis: Clinical
manner or pain is induced at the nonunion site. In
Assessment and Pain
these chronic cases, radiographic single-leg
stance anteroposterior (AP) views are some-
Although pain is not always present in non-
times more helpful to show instability, as will be
unions, it is often the primary reason for a patient
reviewed later.
to seek medical consultation (Figs. 8.1, 8.2, 8.3
Pain secondary to nonunion of the pelvis is
and 8.4). The pain is commonly secondary to
often present during weight bearing and
instability of the pelvis, or malreduction, and is
improves with rest. Because weight is transmit-
most frequently located posteriorly in the
ted posteriorly through the pelvis, pain is more
sacroiliac (SI) region [1]. A nonunion can ini-
commonly associated with the sacrum or SI
tially be nondisplaced, but, due to lack of heal-
nonunions (see Fig. 8.1). Nonunions of the
ing, may displace and develop an associated
anterior pelvic ring are rarely painful because
deformity (see Fig. 8.3). Furthermore, a non-
less than 10% of the body’s weight is transmitted
union of one part of the pelvis, especially in
through the anterior part of the pelvis [4]. When
osteoporotic patients, can lead to insufficiency
the rare case of a painful nonunion of the anterior
fractures in other areas of the pelvis (Figs. 8.4
pelvic ring does present, it is often following a
and 8.5). Posterior pelvic pain associated with
protracted course and multiple consultations with
malunion often improves after correction of the
medical specialists (gynecologists, general sur-
malunion, although the reason for this is less
geons, urologists, rheumatologists, etc.) (see
apparent than with correction of nonunions [2,
Figs. 8.2 and 8.3). The patient may also experi-
3]. Some residual chronic pain often occurs. In
ence low back pain secondary to the pelvic
an acute injury, instability is readily apparent on
deformity, or neurogenic pain that radiates to the
physical examination of the pelvis. This is more
ankle secondary to compression or distraction of
difficult to appreciate in chronic nonunions. In
the nerves at the level of the roots or the lum-
these situations, the physician’s hands are placed
bosacral plexus. Scarring within the nerve is a
on each of the anterior superior iliac spine (ASIS)
common cause of chronic pain.
and the pelvis is rocked from side to side. Subtle
Patients may also complain of pain while
sitting or lying down. The two major causes for
this are pelvic malalignment that causes sitting
or lying imbalance and ischial nonunions that
K.F. Dickson (&) result in painful motion of the fracture upon
Southwest Orthopedic Group, Baylor University,
sitting. The pain with nonunion is due to the
6560 Fannin St, Suite 1016, Houston, TX 77030,
USA micromotion that is occurring at the nonunion
e-mail: kyledickson99@gmail.com site.

184 K.F. Dickson, MD, MBA
8 Acetabular and Pelvic Nonunions 185
b Fig. 8.1 a Anteroposterior (AP) pelvis at presentation sacrum for bleeding bone through a posterior approach.
2 years after the accident with pain in the sacrum treated d AP pelvis 1 year postoperative showing no lucency
previously with cannulated screws. b Axial CT scan around the screws and solid fixation. Patient ambulating
showing nonunion of the sacrum. c Initial postoperative with foot pain with minimal pelvic pain. e Axial CT 6
AP pelvis after open reduction internal fixation of the months postop showing bridging bone
sacral nonunion with iliac crest bone graft and burring
8.1.1 Deformity
monitoring. Disruption of peripheral nerves
Pelvic deformity is responsible for complaints in should be evaluated by nerve
many clinical areas—pain, gait abnormalities, conduction/electromyography tests. Peripheral
genitourinary system, etc.—and is beyond the disruptions may be repaired with possible sal-
scope of this chapter. Occasionally, failure to vage of some function or more consistent return
heal or a nonunion will eventually displace, of protective sensation. Myelograms and mag-
resulting in an associated malalignment (see netic resonance imaging (MRI) are used to rule
Fig. 8.3). Additionally, initial displacement or out spinal nerve avulsions.
deformity can separate the bone enough to cause Our studies on malunions and nonunions
a nonunion (Fig. 8.6). show that 57% of the patients had a preoperative
nerve injury and only 16% were resolving post-
operatively [2, 3]. Only one patient in our studies
8.1.2 Neurologic Injuries would not have the nonunion/malunion surgery
again, and this was due to a postoperative nerve
Permanent nerve damage is a common cause of complication. The patient underwent two opera-
disability following pelvic injuries. A nerve tions on a 16-year-old nonunion that was extre-
injury occurs in 46% of the patients with an mely mobile. An L5 nerve root injury occurred
unstable vertical pelvis [5]. The most commonly from likely reducing a vertical malreduced
affected nerve roots are L5 and S1, but any root hemipelvis almost 5 cm. The patient required
from L2 to S4 may be damaged. In Huittinen’s reoperation for persistent nonunion. At the time
[5] study of 40 nerve injuries, 21 (52.5%) were of the second operation, the posterior fixation
traction injuries, 15 (37.5%) were complete dis- was changed. The complaints of deformity were
ruptions, and 4 (10%) were compression injuries. completely resolved, but the patient still suffered
Interestingly, the lumbosacral trunk and superior from pain in the L5 nerve distribution, despite
gluteal nerve sustained traction injuries, while having a stable pelvis.
most of the disruptions occurred in the roots of
the cauda equina. Compression injuries occurred
in the upper three sacral nerve foramina in 8.1.3 Patient Expectations
patients with fractures of the sacrum. Further-
more, the traction and nerve disruption injuries An important aspect of the preoperative assess-
occurred in the vertically unstable pelvic injuries, ment is to discover a patient’s understanding and
while the compressive nerve injuries occurred expectations regarding their clinical problem.
following lateral compression of the pelvis. Lat- Significant discussion is necessary prior to
eral compression injuries of the pelvis often making a decision for surgery. The patient must
impact portions of the sacral bone into the fora- make the final decision based upon realistic goals
men, resulting in compression of the nerve, and and an understanding of the risk of complica-
may require decompression if neurologic exam- tions. Specific symptoms of deformity, such as
ination worsens. limb shortening, sitting imbalance, vaginal
A thorough neurologic examination is neces- impingement, and cosmetic deformity, are
sary to determine any preoperative deficits, and expected to be reliably addressed by surgery. The
for intra-operative as well as postoperative nerve patient must be cautioned, however, that while
b Fig. 8.2 a Anteroposterior (AP) view of the pelvis CT scan image illustrating nonunion of the inferior rami.
demonstrating normal alignment with a nonunion of the e Intra-operative photograph demonstrating positioning of
superior and inferior rami after 1 year of symptoms. the patient. f Intra-operative photograph demonstrating
b Magnified view of the nonunion. c Coronal CT scan location of the inferior rami. g Postoperative AP view of
image illustrating nonunion of the superior rami. d Axial the pelvis
the majority of the deformity can be corrected, translational displacements around a three ordi-
the actual anatomic result is usually less than nate axis.
perfect. In our series of pelvic malunions, only Plain X-rays will often show the anterior rami
76% of our reductions had less than 1 cm of nonunions (see Figs. 8.2, 8.3, 8.4 and 8.5). A CT
residual deformity [2, 3]. scan with sagittal and coronal reconstructions are
Posterior pelvic pain in the absence of a required to define the posterior pelvic nonunions
demonstrable nonunion or instability is often (see Figs. 8.1 and 8.4). When viewing the CT, all
difficult to explain and may not completely or of the slices need to be studied. Musculoskeletal
reliably improve with the correction of the pelvic radiologists sometimes call a nonunion because a
deformity. Ninety-five percent of patients with particular slice may not have any apparent boney
malunion of the pelvis report improvement of bridging, but following the individual pieces will
their pain; however, only 21% have complete often show some slices with boney bridging,
relief of their posterior pain [2, 3]. Radiographic eliminating the diagnosis of nonunion.
evidence of SI joint arthrosis is not a reliable Evaluating the nonunion site radiographically
indication of the cause of posterior pelvic pain. can determine whether the nonunion is hyper-
However, in patients with a pelvic nonunion, a trophic (needs stability), oligotrophic (may need
significant reduction in pain is seen (see stability and biology), or atrophic (needs biol-
Figs. 8.1, 8.3, 8.4, and 8.5). ogy). Atrophic nonunions, especially in the pel-
vis, should alert the surgeon to the possibility of
a hormonal, nutritional, or medical problem.
8.1.4 Radiographic Assessment Working with a bone endocrinologist is helpful
in these difficult cases (see Figs. 8.4 and 8.5).
Radiographic assessment includes five standard Infection is always a concern in any nonunion, so
pelvis X-ray views (AP, both 45° oblique views, C-reactive protein (CRP), erythrocyte sedimen-
40° caudad, and 40° cephalad), a weight-bearing tation rate (ESR), and white blood count (WBC)
AP X-ray, computed tomography (CT) scan, and a are routinely ordered. In infected nonunions (see
three-dimensional CT. The CT scan can be used to Figs. 8.1 and 8.4), these parameters are impor-
make a three-dimensional pelvic model. This tant to follow; the use of these labs on and off
model helps the surgeon to understand the defor- antibiotics can help determine when the infection
mity and plan preoperatively. The displacement is controlled enough to fix the nonunion.
and rotation of all fragments need to be understood
so that appropriate release and reduction of frag-
ments can be obtained. An obturator oblique 8.1.5 Treatment
clearly shows the SI joint on the ipsilateral side,
while a single-leg weight-bearing AP determines
stability of the nonunions. Technetium bone scans The best treatment is prevention [1, 6–9]. But
may be helpful in identifying the activity of the even after adequate treatment, nonunions can
nonunion (atrophic or hypertrophic) but are not occur (see Figs. 8.1, 8.2, 8.3 and 8.4). The
routinely ordered. Together, these multiple plain problem of malunions and nonunions appears
films and CT scans are used to assess nonunions most commonly after inadequate initial treatment
and deformities of the pelvis. The displacements of displaced fractures and unstable pelvic ring
are often complex and include rotational and injuries (see Fig. 8.5) [10]. In the acetabulum,
Fig. 8.3 a Anteroposterior (AP) X-ray of the pelvis from of inferior rami. Groin pain resolved and impotence
the time of injury. b AP X-ray at 6 months postinjury improved but had an increase posterior SI pain and
demonstrating malunion of the sacrum and sacroiliac joint abnormal gait with intoeing. f Axial CT scan image
and a painful inferior rami nonunion. A rotational defor- demonstrating 15° internal rotation deformity of the left
mity of the hemipelvis is also apparent. c Axial CT scan hemipelvis. g AP X-ray status postsacral osteotomy and
image demonstrating ischial nonunion. d AP X-ray after correction of deformity. h, i Pre- and postoperative pelvic
screw fixation of the nonunion (surgeon mistakenly fixed inlet X-rays illustrating correction of rotational
a malunion). e Axial CT demonstrates a healed nonunion malalignment
Fig. 8.3 (continued)
nonunions can occur in transverse fractures when or with an external fixator. These nonunions
these fractures are treated nonoperatively, with progress to malunions. From the technical
inadequate fixation (e.g., one-third tubular plates standpoint, late correction is very difficult; the
without lag screws instead of reconstruction anatomy is altered and less recognizable, and the
plates and lag screws), or there are medical potential complications are increased. Osteo-
problems with the patient (e.g., malnutrition, tomies can easily damage the structures that lie
vitamin D deficiency, rickets). In pelvic non- on the opposite side of the bone. Scarring around
unions, the same possible causes exist, nonop- nerves prevents the fragments from moving
erative or inadequate fixation and medical freely without causing nerve palsy.
comorbidities. Because most of the weight is Indications for surgery include pain, pelvic
transferred through the posterior pelvis, anterior ring instability, and clinical problems relating to
rami nonunions rarely exist and, if present, are the pelvic deformity (gait abnormalities, sitting
rarely symptomatic. However, if not treated, rami problems, limb shortening, genitourinary symp-
nonunions in elderly patients can lead to addi- toms, vaginal wall impingement, etc.). A thor-
tional pathological fractures (see Fig. 8.4). More ough knowledge of pelvic anatomy is required to
commonly nonunions occur posteriorly and are understand the three-dimensional deformity.
due to unrecognized instability of the pelvis with Furthermore, extensive preoperative planning is
vertical instability that is treated nonoperatively needed to determine the proper order of
Fig. 8.4 a Anteroposterior (AP) X-ray after a bladder showing the percutaneous placement of bilateral iliosacral
sling (no trauma but an insufficiency fracture). b Axial CT screws. e Postoperative AP pelvis 8 weeks out showing
scan showing the bilateral sacral fractures. c Axial CT all fixation. f AP pelvis at 17-month follow-up
scan of the rami nonunion. d Postoperative AP X-ray
Fig. 8.5 a–c Anteroposterior (AP), iliac oblique, and deformities. d Three-dimensional CT reconstruction of
inlet X-rays of pelvis demonstrating the pelvic ring the pelvis. e, f Postoperative AP and iliac oblique X-rays
nonunion with translation, flexion, and internal rotation of the pelvis
exposures for release, reduction, and fixation. always taken regardless of how the nonunion
Because each patient is different, it behooves the looks. We have seen a number of culture-positive
surgeon to individualize the treatment. less virulent organisms that do not form pus but
Previous literature focused on simple non- infect and prevent bones from healing. The two
unions. These patients often do not require sides of the bone are exposed with as minimal
extensive anterior and posterior ring releases and stripping as possible. It is important to remember
reduction, and respond to in situ fusion only (see that the blood supply comes predominantly from
Fig. 8.4). Pennal and Massiah showed that the soft tissue. The surgeon burrs both sides of
patients treated with surgery are significantly the bone until cortical bleeding is seen. The
better than those treated conservatively [11]. In surgeon then either burrs or drills inside the
their study, 11 out of 18 surgery patients returned nonunion site to penetrate any cortical cap and
to preinjury occupation versus five out of 24 get bleeding from the cancellous bone. The sur-
conservatively treated patients. In nonunion geon then lays cancellous strips from bleeding to
cases with significant displacement, in situ bleeding bone. This is covered with good
fusions are unrewarding and leave the patient bleeding muscle. The fixation is either
with complaints related to deformity as well as intra-medullary (i.e., iliosacral screws) or plate
significant pain (see Fig. 8.3). fixation next to the cancellous graft not covering
Surgical technique in the presence of signifi- it.
cant deformity can be complex and is beyond the Nonunions of rami fractures are rare. If they
scope of this chapter, but one must be familiar occur, they are often located in the medial aspect
with correction of deformities along with the of the pubis bone or in the symphyseal region.
management of the nonunion. This chapter will Because more than 90% of weight bearing is
focus on healing the bones that generally require posterior, many nonunions of the anterior pelvic
only one stage versus the more complex ring are asymptomatic. Because these are so rare,
three-stage reconstruction as described by they are often not part of the differential diag-
Letournel [2, 3, 7]. nosis, and some patients can be evaluated by
A radiolucent table with image intensification several specialists (obstetrics and gynecology,
is commonly used for the procedure. The Judet general surgery, etc.) before an X-ray identifies a
table is also useful for positioning the leg any- painful nonunion (see Fig. 8.2). Often, treatment
where in space to help with the approaches and of symptomatic superior rami nonunion will heal
exposure. Somatosensory-evoked potentials and the inferior rami nonunion, especially if the
motor-evoked potentials have been used on some nonunion is hypertrophic or the underlying
patients intra-operatively that require significant medical comorbidity has been corrected (see
correction of vertical displacement and but are Fig. 8.5). However, there are cases where plating
not routinely used. both the superior and the inferior rami is
required, especially in cases of atrophic non-
unions (see Fig. 8.2).
8.1.6 Pelvic Nonunions A Foley catheter is always placed preopera-
tively. A Pfannenstiel incision is made 2 cm
Painful nonunions without deformity can be cephalad from the symphysis. The decussation of
treated with stabilization, bone graft, or both. the fascia fibers of the rectus abdominis marks
A technetium bone scan can indicate activity of the division between the two heads of the rectus.
the nonunion (atrophic [requires bone graft] or The two heads are split, with extreme care being
hypertrophic [requires stabilization]). In most taken to avoid entering the bladder. The surgeon
cases, it is not necessary, and surgery involves then inspects the bladder to detect any perfora-
both bone graft and stabilization. tions. The Foley should be palpated to ensure
The preparation of the bone is the same no that the urethra is intact. A malleable retractor or
matter where the nonunion occurs. Cultures are lap sponge is then used to hold the bladder away
Fig. 8.6 a Anteroposterior (AP) X-ray of the pelvis head. d AP X-ray of the pelvis 3 months postinjury. e–g
demonstrates significant lateral displacement of the Axial CT scan images demonstrate extensive callous
femoral head. b, c Axial CT scan images demonstrating around the fracture site. h AP X-ray of the pelvis
displacement of iliac wing and dislocation of femoral following repair of the nonunion
from the symphysis pubis. Two Hohmann surgeon to also perform a Pfannenstiel incision
retractors are used to retract the two heads of the as well. Rarely, in already osteoporotic bone,
rectus from the superior surface of the symphysis rami nonunion can cause enough instability to
pubis. The superior surface of the superior rami cause insufficiency fractures posteriorly (see
is cleaned for the plate, but the anterior insertion Fig. 8.4). Elderly simple rami fractures need to
of the rectus remains intact. A large Weber clamp be worked up for osteoporosis to prevent this
or pelvic reduction clamp can be used anteriorly disastrous complication. The initial treatment of
to hold the symphysis together or rami fracture these insufficiency fractures includes touchdown
together. Usually, a six-hole 3.5 mm curved weight bearing with no range of motion or
reconstruction plate is then implanted. Clinical exercises to that side for eight weeks. If the
research supports the implantation of this device patient continues to be symptomatic, the pelvis is
[12]. Rarely, when a patient has a very small stabilized with posterior iliosacral screws (for
symptomatic nonunion of the pubis or late posterior insufficiency fractures) and/or platting
painful instability of the symphysis, a symphy- of rami fractures with another eight weeks of
seal fusion is indicated. When a fusion of the touchdown weight bearing. These patients are
symphysis is needed, an additional four-hole aggressively treated (parathyroid hormone, vita-
plate is used anterior to the symphysis with min D, calcium, etc.) by a bone endocrinologist.
cancellous bone graft. Additionally, when fusion Furthermore, the failure of rami to heal properly
of the symphysis is indicated, an eight to ten-hole may cause further displacement and eventual
plate is used rather than a six-hole plate superi- pelvic malunion (see Fig. 8.3).
orly. Through the Pfannenstiel approach, the SI For iliac wing nonunions, the lateral window
joints can be visualized and the quadrilateral of the ilioinguinal approach is performed (see
surface exposed via the modified Stoppa Fig. 8.5). Most of these have occurred with
approach [13]. Therefore, a plate can be placed nonoperative treatment or in a patient with a
from the symphysis to the SI joint along the brim significant bone-healing problem. The L5 nerve
superiorly bilaterally. Furthermore, a plate can be runs 2 cm medial to the SI joint and must be
placed within the pelvis from the symphysis protected. If vertical translation has occurred,
along the quadrilateral plate to the SI joint. Plates mobilization of the nerve is required to reduce
or screws can be used on the inferior rami (see the hemipelvis without causing nerve palsy.
Fig. 8.2) via a direct approach, with the patient in Bicortical 3.5-mm screws are used in a recon-
the lithotomy position. This position allows the struction plate. Anterior plating of the SI joint
with two three-hole plates positioned 90° to each [14, 15], but in our series of sacral injures, these
other (first plate along the brim with one screw in were rarely required [10]. There is theoretical risk
the sacrum and 2 screws in the ilium) can give of compression of the nerve roots in comminuted
excellent fixation of the SI joint. Since the joint sacral fractures using a lag screw. However, using
from anterior has about a 15° medial slope, fully threaded screws risks maintaining a gap at
angling the drill in this direction will provide the fracture site, potentially increasing the risk of
longer screws. The use of a long oscillating drill, nonunion. In the author’s experience of greater
which allows tactile feel of the second cortex and than 1000 lag iliosacral screws, with only 2 cases
prevents over penetration with the drill, is rec- of nerve palsy that eventually resolved, use of a
ommended because of its flexibility and safety. fully threaded screw is rarely used. Furthermore,
Alternatively, percutaneous iliosacral screws can the pelvic nonunions referred to our institution, if
be placed. Iliac wing nonunions usually require operated on, usually have a malreduced pelvis,
plate fixation only without involvement of the SI and a fully threaded screw possibly increases the
joint. Depending on the reduction required, the risk of nonunion.
iliac portion of the ilioinguinal and the Pfan- Patients are touchdown weight bearing or
nenstiel may be adequate for the combination wheelchair transfer for only 12 weeks postoper-
iliac wing and rami nonunion. However, in atively. After adequate healing, range of motion
chronic nonunions of both the pelvis and the and strengthening exercises are instigated.
acetabulum, a full ilioinguinal is usually required
(see Figs. 8.4 and 8.5).
Sacral nonunions (see Fig. 8.1), due to limited 8.1.7 Malunions and Displaced
visualization from an anterior approach, almost Nonunions of the Pelvis
always are operated on through a posterior
approach. A longitudinal approach two centime- To treat symptoms related to deformity of the
ters lateral to the posterior superior iliac spine is pelvis, a reduction of the pelvis is required
made [10]. The gluteus maximus is raised off of because a simple in situ fusion will be unre-
the iliac crest, lumbodorsal fascia, and paraspinal warding and not completely relieve the pain (see
muscles, exposing the posterior SI joint and Fig. 8.3). If a nonunion leads to pelvic deformity,
ligaments. The bone is prepared as above, being osteotomy, and release of the hemipelvis is
very careful of the sacral nerve roots. Fixation is required to reduce the deformity and alleviate
usually obtained with two 6.5 mm, symptoms (see Fig. 8.3) [16]. Correction of the
16-mm-thread-length iliosacral screws. Again, deformity is beyond the scope of this chapter.
the use of an oscillating drill is recommended for
safety and so that three cortices are entered but
not the fourth. Additional stability can be 8.1.8 Results
achieved by placing one or two posterior recon-
struction plates from one iliac wing to the other In our series that included both nonunions and
iliac wing. These plates act as a tension band and malunions, the time frame from injury to opera-
are less prominent if placed caudad to the poste- tion in our series averaged 42 months (range from
rior superior iliac spine. These tension band plates 4 months to 14 years) [2, 3]. In our initial series,
(typically 12–14 holes) have three screws in both 9% (3/34) had pure iliac wing nonunions, but the
iliac wings, with one screw from shooting combination fracture dislocation of the SI joint
between the inner and outer cortical tables and the produced 35% (12/34) nonunions. By far the
second and third shorter screws transversing the most common nonunions were in the sacrum,
two tables. Iliosacral bars are also an option; which were 56% (19/34) of the nonunions. Most
however, they are usually prominent and were not rami fractures associated with the nonunions had
used in our series [2, 3]. More recently, healed, but there were some cases of continued
trans-sacral screw fixation has been described symphyseal instability. Sixty-six percent (23/35)
were initially treated nonoperatively, and seven of Case 2. Painful nonunion of the anterior pelvic
these had a previous attempt to fix their ring (see Fig. 8.2)
nonunion/malunion. The cases that were operated
were inadequately stabilized with an external A 67-year-old presented 1 year postinjury after
fixator or screw and plate fixation. The more multiple consultations with medical specialists.
recent trend is that of more cases being operated Patient found to have nonunion of the anterior
on, but there is inadequate fixation with an pelvic ring confirmed by radiographic imaging
external fixator, or adequate reduction is not (plain and CT). Patient underwent stabilization of
achieved with iliosacral screws. Operative time the superior and inferior rami with complete
averaged 7 h (range 1.5–10.4 h). Operative blood healing and resolution of pain. Patient stopped
loss averaged 1977 cc (range 200–7200 cc). using narcotics 1 week postop after being on
At follow-up (average, 3 years, 11 months; narcotics for over 1 year. Patient continues to be
range, 9 months to 11 years), all but one patient pain free >2 years postoperatively.
had a stable union of their pelvic ring.
Ninety-five percent of the patients were satisfied Case 3. Patient with a nonunion of the inferior
with the operation, and 100% of the patients ramus with associated deformity of left
were satisfied with the improvement of their hemipelvis (see Fig. 8.3)
preoperative deformity. As mentioned earlier, the
unsatisfied patient continues to have L5 nerve A 58-year-old presented 6 months postinjury
palsy. Now, with the experience of over a hun- with complaints of groin pain and impotence.
dred pelvic nonunion and malunion patients, Patient found to have nonunion of inferior rami
prevention is still the key. left side, which was fixed but associated malu-
Complications included loss of reduction, nion was not addressed at first surgery. Patient’s
neurologic injury, and vascular injury (external groin pain resolved and impotence improved.
iliac vein). There were no postoperative surgical Patient now had increase in posterior SI pain and
infections, although there were some preopera- abnormal gait in intoeing of left side. After
tive infections that had to be treated with a staged evaluation and determination of a left hemipelvis
approach (see below). Although residual low internal rotation deformity, surgical osteotomy of
back pain was present in most of the patients the sacrum with correction of the deformity was
preoperatively, 95% reported less pain following performed. Pain diminished and patient returned
surgery. to work as a train conductor.
Case 4. Rami nonunion leading to insuffi-

8.1.9 Pelvic Nonunion Case ciency bilateral sacral fractures after place-
Discussions ment of a vaginal bladder sling (see Fig. 8.4)
Case 1. Nonunion of the sacrum (see Fig. 8.1) A 53-year-old with a 4-month history of groin
pain and a 1-month history of being unable to
A 62-year-old was involved in a head-on ambulate. Patient had a trans-vaginal bladder
motorcycle collision with complicated urologic mesh placed 5 months previously and woke up
injuries and history of infections in both the (no history of any trauma) 1 month later with
symphysis and the sacrum. Patient initially trea- groin pain. The patient was eventually worked up
ted with iliosacral screws and symphyseal plat- after finding a right rami fracture with an MRI
ing. Patient with persistent sacral (posterior) pain and CT scan, which were negative for any other
at 2 years. Radiographic imaging revealed sacral lesion. She was found to be vitamin D deficient
nonunion. Patient underwent bone grafting and and was started on appropriate therapy. 3 months
revision fixation of the sacrum with a successful later or 7 months after the onset of symptoms
union. without any trauma, the patient presented to our
clinic with increasing bilateral posterior pain in the hip. Critical to the preoperative assessment
eventually diagnosed as bilateral insufficiency of acetabular nonunion is the condition of the
sacral fractures. Patient underwent stabilization femoral head. Evaluation of the hip joint is also
with iliosacral screws and anterior plating, can- important to determine how much cartilage
cellous bone graft, and placement of BMP-2 remains. Attempts to compensate for loss of
(Infuse® Medtronics) of the rami nonunion. The substance of the femoral head or the cartilage
right superior ramus had debridement of granu- have not been successful. The osteoarthritis
lation tissue suspicious for a chronic burnt-out rarely improves, and at best the deterioration is
infection that had negative cultures. Patient halted. Before attempting reconstruction of an
treated with 6 weeks of broad-spectrum intra- acetabular nonunion, the following must be
venous antibiotics. Patient was now able to understood:
ambulate (after 10 months of nonambulating)
with some pain. 1. The location and condition of the different
articular fragments and the bony columns
Case 5. Nonunion of the iliac wing and supe- supporting them,
rior and inferior rami with no history of 2. The extent and location of wear on the
trauma and orthopedic diagnosed malab- femoral head,
sorption syndrome (see Fig. 8.5) 3. The presence, location, and extent of
osteoarthritis, and
A 49-year-old presented 1 year postpelvic frac- 4. The presence, location, and extent of avas-
ture after being referred to multiple medical cular necrosis [7].
specialists before X-rays were taken. Patient was
unable to ambulate for the previous 9 months. In all cases, a total hip arthroplasty (THA) is
Patient was found to have a pelvic nonunion with considered an option (see Fig. 8.6). If there is
significant associated deformity. Patient under- complete cartilage loss involving more than 50%
went single-stage repair of the nonunion through of the dome, a THA is probably required
an ilioinguinal approach. Patient is ambulating (Fig. 8.7). Depending on the associated deformity,
without pain 5 years postop. the THA may need to be performed in conjunction
with reduction of the columns (Fig. 8.8).
8.2 Nonunion of the Acetabulum:

Clinical Assessment and Pain 8.2.1 Deformity
Pain associated with an acetabular nonunion Acetabular deformity and/or hip protrusio causes
usually occurs either at the nonunion site or in symptoms similar to what is seen in the pelvis:
the hip joint. An acetabular nonunion may have a gait abnormalities, sitting imbalance, and limb
slight malreduction at the joint causing pain length discrepancy (i.e., shortening of a trans-
because of increased intra-articular pressure verse fracture). Furthermore, protrusion of the
during weight bearing. This is due to articular femoral head centrally will cause a significant
incongruity reducing the contact area between decrease in motion (see Fig. 8.8). The difference
the head and the acetabulum, wear of the head between pelvic nonunions and acetabular non-
rolling over a malreduced fracture line, avascular unions is that nonunions of the acetabulum
necrosis, or osteoarthritis of the acetabulum. require early diagnosis to prevent the develop-
Symptoms include increasing severity of pain ment of severe arthritis after which the hip will
with hip motion, limp, and restriction of hip no longer be salvageable (see Figs. 8.7 and 8.8).
motion. Radiographic studies are used (as Radiographic analysis is critical (see below) to
described later) to determine whether a nonunion determine the type of the fracture present and the
exists and the extent and location of the damage amount and direction of displacement.
b Fig. 8.7 a, b Anteroposterior (AP) of the pelvis at the cuts showing bridging and the end-stage osteoarthritis.
time of the accident 5 years before presentation pre- and f AP pelvis postsingle-stage open reduction internal
postreduction of the hip dislocation. c AP pelvis at fixation and hip arthroplasty through a
presentation with a complete nonunion and end-stage Kocher-Langenbeck approach. g–j Six-month follow-up
osteoarthritis. d, e Axial CT showing a complete with both 45° oblique views, and AP and lateral of right
nonunion of the transverse fracture without any of the hip showing solid fixation and patient without pain
8.2.2 Genitourinary System 8.2.3 Neurologic Injuries

in Acetabular Nonunions
Genitourinary symptoms in acetabular nonunions
can present similar to what is seen in pelvic The neurologic injuries associated with acetabu-
nonunions. Besides the symptoms listed in the lar fractures are somewhat different than neuro-
pelvic section, a separate cyst of urine can be logic injuries in pelvic fractures and dislocations.
found where a rami fracture has perforated the The predominant nerve injury in acetabular
bladder. fractures is the common peroneal tract of the
Fig. 8.8 a–c Anteroposterior (AP) obturator oblique and central migration of the head. f, g AP and obturator
iliac oblique X-ray, respectively, following a fall from a oblique X-ray, respectively, showing the medial sublux-
standing position. d Axial CT showing the medial dome ation of the femoral head due to a nonunion. h Postop-
impaction and central migration of the head. e Coronal CT erative AP X-ray after repair of nonunion and arthoplasty
reconstruction showing the medial dome impaction and
sciatic nerve causing a foot drop. Additional excellent results if the delay exceeded 12 weeks
nerve injuries include the superior gluteal nerve from the time of injury. Once again, significant
(abductor weakness) and obturator nerve (ad- discussion is necessary prior to making a decision
ductor weakness and numbness of the inner for surgery. The patient must have realistic goals
thigh). Rarely, the femoral nerve may be injured. and an understanding of the risks and benefits of
A preoperative examination will often identify surgery. The patient needs to understand preop-
partial or complete muscle weakness. In acetab- eratively that success is limited and that THA is
ular nonunions, complete knowledge of the likely in the intermediate or long term.
anatomy is required to prevent a retractor iatro-
genic nerve injury. Mayo et al. described post-
operative nerve palsies following correction of 8.2.5 Radiographic Assessment
acetabular malunion in six percent of their cases
(three percent superior gluteal and three percent The radiographic analysis of acetabular non-
sciatic) [17]. unions uses the same radiographs as in the pelvic
section with the addition of MRI to look for
cartilage damage and avascular necrosis. Often,
8.2.4 Patient Expectations these injuries have areas of malunions and non-
unions in the same fracture line; that is, in a
For acetabular nonunion patients, the results are transverse fracture, the fracture heals
not as good as those seen in pelvic nonunions, supero-medially and has a nonunion
and the degree of difficulty and the need for postero-inferiorly. To make the diagnosis of a
precise anatomic reduction is an order of magni- nonunion (see Fig. 8.7), no bridging bone is
tude greater than that discussed for the pelvis. found anywhere along the fracture line. The
Nothing less then a perfect reduction of the patients with partial unions typically do not have
acetabulum is acceptable, and even in experi- pain from the nonunion site but due to hip
enced hands, in patients with nonunions of the arthritis. Like the pelvis, the patterns of dis-
acetabulum, 58% of the patients go on to develop placement of certain fracture types have been
arthritis [17]. Timing is also an important factor, determined. The way a both column or T-type
with 57% good to excellent results if operated on fracture displaces is somewhat consistent. The
within three weeks of the injury, and 29% good to anterior and posterior columns open up like
“saloon doors” as the head pushes medially (see preventing the bone from healing inferiorly.
Fig. 8.8). Drawing the fracture on a model is There are not enough cases to say definitively
mandatory to determine the rotation of the bro- whether operatively treated acetabulum fractures
ken pieces that either need to be released (non- have more or less percentage of nonunions.
unions) or osteotomized and released Treatment of trans-tectal transverse and trans-
(malunions) in order to obtain anatomic reduc- verse posterior wall without surgery reliably
tion. For instance, transverse fractures have two leads to healing. Sometimes a transverse non-
axes of deformity. The inferior piece rotates union is associated with the psoas tendon being
around an axis that travels down the symphysis caught in the transverse fracture line in the psoas
pubis with greater displacement posterior versus grove. In the handful of cases that were operated
anterior. The inferior transverse fracture rotates on by the author or were referred, a few simi-
around a second axis from that lies on a line from larities existed. The patients were not compliant
the symphysis pubis to the fracture site through with touchdown weight bearing for 8 weeks, had
the posterior column as the head pushes medially a transverse or “T”-type fracture, had a slight
(see Fig. 8.7). deformity, and eventually had THA (see
The edges of nonunions are typically seen Figs. 8.7 and 8.8).
radiographically as hypertrophied bone. Nar-
rowing of greater than 50% of the dome articular
surface is an indication for THA. Wear in other 8.2.6 Treatment
areas of the joint may be well tolerated. Inter-
estingly, some both-column fractures detach the The indications for surgery include displaced
whole articular surface. The femoral head acetabular nonunions that meet the indications for
remains congruent with the dome despite acute surgery (i.e., incongruence at the femoral
widening medially between the two columns and head or > 2-mm step off in the weight-bearing
medial translation of the entire joint. Medial dome). Pain without deformity is also an indica-
widening up to one centimeter may be well tol- tion for surgery [19]. Age has a limited role in the
erated; therefore, treating these malunions con- decision pathway. If the patient has physiologi-
servatively may be the best option. As cally good bone and is active, it is prudent to try to
mentioned, transverse fractures seem to have the save his hip. As mentioned, all of the patients seen
highest incidence of nonunions whether they are in our institution have a combination nonunion
operated on or not (see Fig. 8.7). As mentioned, with some deformity and require resection of
Letournel reported on a 0.7% nonunion rate in fibrous tissue and bone to achieve anatomic
nonoperatively treated acetabulum fractures with reduction. The key is the quality of the cartilage.
most of these being transverse fractures or the One can reconstruct the hip if there is at least 50%
transverse component of a transverse posterior of the dome cartilage intact. If there is already
wall acetabular fracture (7 total: 2 posterior col- complete loss of articular cartilage, the surgeon
umn, 3 transverse, 1 transverse posterior wall, must ask him/herself whether a successful total hip
and 1 posterior column of a both-column) [18]. can be performed with or without an osteotomy of
Even those that heal can still have the inferior the acetabulum to correct the deformity. In the case
part of the posterior column fracture line remain of nonunion, the nonunion always has to be sta-
unhealed. Letournel described this as a bilized first prior to the THA or there will be a
nonunion/malunion and had 14 of these, includ- significant increase in cup loosening (see Figs. 8.7
ing 3 posterior column fractures, 1 anterior col- and 8.8). If the hip is out of the socket either
umn, 2 T-shaped, 2 transverse and posterior wall, medially or laterally (see Fig. 8.6), or the patient
and 6 both-column fractures [18]. The key to has not started weight bearing, usually the carti-
these is the resection of the boney bridge to be lage is preserved and the joint can be salvaged.
able to correct the deformity. This may be due to The treatment of the nonunion site is similar to
the synovial fluid tracking into that area, the treatment described in the pelvic nonunion
section. Preoperative laboratory work includes sides of the acetabulum. Sequential approaches can
possible diagnosis of other contributing diseases be used for these more complex patterns in the acute
such as nutritional, endocrine, or infectious setting but do not work well in established non-
abnormalities. WBC, ESR, and CRP are routinely unions if some reduction will be required. Both
ordered. Elevation of these tests may prompt a sides need to be freed up prior to reduction, so
two-stage revision; the first stage involves an simultaneous approaches is an option, but EIF is
open debridement with multiple site biopsies for preferred to get to both sides of the bone, particu-
cultures and alpha defensin, removal of previ- larly in transverse and “T”-type fractures. The EIF
ously placed hardware, and antibiotic polymethyl is used to salvage the hip joint but is not used if a
methacrylate (PMMA) beads (typically, use one THA will be performed. Either a KL [20] or an
bag of tobramycin [1 g] cement with an addi- anterior Smith-Peterson approach can give enough
tional 3 vials of vancomycin [1 g/vial, 3 g total] exposure for plating the nonunion prior to per-
and 2 vials of tobramycin [1.2 g/vial, 3.4 g forming a THA (see Figs. 8.7 and 8.8). In all cases,
total]). Culturing the nonunion site; burring to the fibrous tissue is removed from the fracture site,
bleeding cortical bone; drilling through potential including intra-articularly through a capsulotomy.
end caps, stimulating cancellous bleeding; The edges of the nonunion can be sclerotic and need
bridging the nonunion site with cancellous bone; to be freshened up so that there is bleeding from
and covering with healthy bleeding muscle are both ends, and cancellous bone graft is packed into
the key concepts for obtaining union. Impor- the gap. Intra-operative traction with
tantly, all callous or healed bone that is prevent- subluxation/dislocation of the hip allows the
ing anatomic reduction is removed. intra-articular nonunion to be corrected, and stabi-
lization is performed with standard compression
plate techniques [21]. Intra-articular osteotomies
8.2.7 Acetabular Nonunions are reserved for when the acetabulum is to be sal-
vaged (see Fig. 8.6). Displaced nonunions require
Simple acetabular nonunions are treated in a mobilization of the fragments similar to acetabular
similar manner to pelvic nonunions. Unfortu- malunions, with direct intra-articular visualization.
nately, the more common scenario is that by the If greater than 50% of the dome has osteoarthritis, a
time a diagnosis of acetabular nonunion is made, THA is performed, usually without mobilizing the
the patient already has complete loss of the artic- fractured fragments but treating the nonunion site
ular surface (see Figs. 8.7 and 8.8). It then and stabilizing the nonunion with a plate without
becomes imperative to treat and stabilize the attempting an anatomic reduction prior to per-
nonunion prior to doing a THA; this can be done at forming THA. The exception to this, as mentioned,
the same setting. If THA is performed without is where a protrusion of the femoral head into the
stabilization of the nonunion, >80% of these cases pelvis requires lateralization of the head and
will have loosening of the acetabular component. reduction of the two columns medially with plate
Anatomic reduction of the joint is required if stabilization prior to doing a THA.
salvaging the acetabulum (see Fig. 8.6). If there is
significant protrusio of the femoral head within the
pelvis, even in the face of doing a THA, osteo- 8.2.8 Malunions and Displaced
tomies and reduction of the two columns to keep Nonunions
the THA out of the pelvis are performed. of the Acetabulum
The choice of approach in nonunions is similar
to the acute setting: The Kocher-Langenbeck Most of the acetabular nonunions have some
(KL) for nonunions of the posterior column and displacement that requires reduction if salvage is
wall, the ilioinguinal for the anterior wall and col- desirable, and are complicated by the fragments
umns, and the extended iliofemoral (EIF) for all being healed or scarred down in the wrong
other fractures that require the surgeon to be on both position (see Fig. 8.6). Complete release of the
bone and associated soft tissue is required for A 79-year-old fell from standing position suf-
anatomic reduction of the joint. Interestingly, fering a “T”-type acetabulum with medial dome
bone healing is much more rapid than cartilage impaction and central migration of the femoral
healing, so osteotomies through old malunions head. Patient underwent operative repair of
can be visualized more easily intra-articularly acetabular fracture, but there was a failure of
than extra-articularly. Also, reduction can be reduction of the quadrilateral surface. Patient
visualized intra-articularly to ensure congru- with pain and discomfort in left hip with estab-
ency. Discussion of acetabular malunions is lished nonunion. Patient underwent an anterior
beyond the preview of this chapter. Smith-Peterson approach, fixation of the anterior
and posterior columns of the “T”, and an anterior
THA. A buttress plate over the brim onto the
8.2.9 Acetabular Nonunion Case quadrilateral plate was placed to prevent protru-
Discussions sion (custom-made lone star plate) and buttress
plate along the pelvic brim buttressing the lone
Case 1. Nonunion of the acetabulum (see star plate using the femoral head as a medial
Fig. 8.6) bone graft and performing a THA. Patient
2 years out walking without a walking aid, no
A 65-year-old initially presented with an open groin or thigh pain, but with some trochanteric
anterior column acetabular fracture. Surgery was pain.
delayed until 3 months postinjury due to the
condition of the soft tissues. Patient had devel-
oped a nonunion of the acetabulum, which
underwent single-stage open reduction internal 8.3 Conclusion
fixation through an ilioinguinal approach. Patient
ambulating without pain and no joint narrowing Stabilization of nondisplaced pelvic nonunions,
one year postop. especially posteriorly, has been proven to be
successful in returning patients to their preinjury
Case 2. Nonunion of a juxtatectal transverse status [11]. Treatment of acetabular nonunions is
acetabulum leading to degenerative hip not as successful due to pain coming from the hip
arthritis (see Fig. 8.7) joint in the form of osteoarthritis or avascular
necrosis. Operative correction of nonunions of
A 49-year-old presented 5 years after a trans- the acetabulum can give excellent results if the
verse acetabulum fracture that was treated con- joint does not already have significant damage.
servatively. Patient was never compliant with the The results of surgery in the setting of nonunion
previous touchdown weight-bearing recommen- are not as good as those of acute treatment of
dations. Patient was found to have a symptomatic acetabular fractures or pelvic ring injuries. Once
nonunion and end-stage osteoarthritis based on the nonunion has established itself and chronic
radiographic examination. Patient underwent symptoms develop, the probability of surgical
single-stage repair of the nonunion and THA. reconstruction returning the patient to his or her
Patient did well and his pain resolved. Unfortu- preinjury status is decreased. Also, the rate of
nately, the patient was trying to outrun the police complications is higher for late surgical treatment
on a motorcycle, and at the time of this writing [3, 17]. Prevention by acute anatomic reduction
was in prison. and internal fixation of unstable pelvic injuries
and anatomic articular reductions in acetabular
Case 3. Nonunion of a transtectal “T”-type fractures is the best treatment for pelvic and
acetabular fracture (see Fig. 8.8) acetabular nonunions.
References 12. Matta JM, Tornetta P 3rd. Internal fixation of

unstable pelvic ring injuries. Clin Orthop Relat Res.
1996;329:129–40.
1. Semba RT, Yasukawa K, Gustilo RB. Critical 13. Cole JD, Bolhofner BR. Acetabular fracture fixation
analysis of results of 53 Malgaigne fractures of the via a modified Stoppa limited intrapelvic approach.
pelvis. J Trauma. 1983;23(6):535–7. Description of operative technique and preliminary
2. Dickson KF, Matta JM. Surgical reduction and treatment results. Clin Orthop Relat Res.
stabilization of pelvic nonunions and malunions. 1994;305:112–23.
Paper presented at: The 63rd Annual Meeting of the 14. Beaule PE, Antoniades J, Matta JM. Trans-sacral
American Academy of Orthopaedic Surgeons; 1996; fixation for failed posterior fixation of the pelvic ring.
Atlanta, Georgia. Arch Orthop Trauma Surg. 2006;126(1):49–52.
3. Matta JM, Dickson KF, Markovich GD. Surgical 15. Griffin DR, Starr AJ, Reinert CM, Jones AL, Whit-
treatment of pelvic nonunions and malunions. Clin lock S. Vertically unstable pelvic fractures fixed with
Orthop Relat Res. 1996;329:199–206. percutaneous iliosacral screws: does posterior injury
4. Tile M, editor. Fractures of the pelvis and acetabu- pattern predict fixation failure? J Orthop Trauma.
lum. Baltimore: Williams and Wilkins; 1984. 2006;17(6):399–405.
5. Huittinen VM, Slatis P. Nerve injury in double 16. Frigon VA, Dickson KF. Open reduction internal
vertical pelvic fractures. Acta Chir Scand. 1972;138 fixation of a pelvic malunion through an anterior
(6):571–5. approach. J Orthop Trauma. 2001;15(7):519–24.
6. Hundley J. Ununited unstable fractures of the pelvis. 17. Mayo KA, Letournel E, Matta JM, Mast JW, John-
Proceedings of the 33rd Annual Meeting of the son EE, Martimbeau CL. Surgical revision of
American Academy of Orthopaedic Surgeons. malreduced acetabular fractures. Clin Orthop Relat
J Bone Joint Surg Am. 1966;46A. Res. 1994;305:47–52.
7. Letournel E. Diagnosis and treatment of nonunions 18. Letournel E, Judet R, editors. Fractures of the
and malunions of acetabular fractures. Orthop Clin acetabulum. 2nd ed. Berlin: Springer; 1993.
North Am. 1990;21(4):769–88. 19. Zura RD, Kahler DM. A transverse acetabular
8. Matta JM, Saucedo T. Internal fixation of pelvic ring nonunion treated with computer-assisted percuta-
fractures. Clin Orthop Relat Res. 1989;242:83–97. neous internal fixation. A case report. J Bone Joint
9. Kellam JF. The role of external fixation in pelvic Surg Am. 2000;82(2):219–24.
disruptions. Clin Orthop Relat Res. 1989;241:66–82. 20. Weber M, Berry DJ, Harmsen WS. Total hip arthro-
10. Hsu JR, Bear RR, Dickson KF. Open reduction of plasty after operative treatment of and acetabular
displaced sacral fractures: techniques and results. fracture. J Bone Joint Surg. 1998;80-A(9):1295–1305.
Sacral fractures. Orthopedics. 2010;33(10):730. 21. Mohanty K, Taha W, Powell JN. Non-union of
11. Pennal GF, Massiah KA. Nonunion and delayed acetabular fractures. Injury. 2004;35(8):787–90.
union of fractures of the pelvis. Clin Orthop Relat
Res. 1980;151:124–9.
Proximal Femur Nonunions
9
Roman A. Hayda, MD
Given the frequency of fractures of the proximal reduction should prompt the clinician to initiate
femur, the active fracture surgeon is likely to evaluation and possible surgical intervention.
encounter patients with nonunion. Fractures of Physical examination and careful evaluation of
the proximal femur, to include the femoral neck, radiographs with comparison to initial postoper-
intertrochanteric, and subtrochanteric regions, are ative films provide valuable clues to early de-
common, particularly in the elderly, with over tection of nonunion. Defining the timing and
310,000 occurring annually in the USA alone adequacy of union may be difficult in relying on
[1]. In adults, almost all fractures in this region radiographic and clinical signs [2]. Scoring of
are treated with surgery. While nonunion is progression of union on anteroposterior and lat-
reported in all patterns of proximal femur frac- eral films in serial films using the RUSH
ture, it is most common in the femoral neck. (Radiographic Union Score for Hip) can be pre-
In the younger patient, proximal femur frac- dictive of union and provides an organized
tures occur with high-energy mechanisms, while method of assessment. Bhandari et al. demon-
in the elderly, low-energy falls are sufficient to strated in a series of femoral neck and inter-
cause displaced fractures with comminution. trochanteric fractures higher interobserver
With the exception of the elderly patient in which reliability in assessing fracture union [3, 4]. In
arthroplasty is performed for displaced femoral select cases, the use of a CT scan can define the
neck fractures, internal fixation of proximal fe- presence of bridging bone particularly when
mur fractures is performed with intramedullary hardware obstructs osseous anatomy. In most
devices, sliding screw and side plate, or cannu- cases, bridging bone in the presence of hardware
lated screws depending on the site of the fracture leads to continued consolidation. Magnetic res-
and surgeon preference. When proximal femur onance imaging and other modalities can help
fractures are appropriately fixed, union typically detect avascular necrosis, infection, and other
results in resolution of pain with restoration of pathology.
ambulatory capacity. When the diagnosis of nonunion is made, the
Nonunion or impending nonunion should be clinician should perform an organized evaluation
suspected when there is loss of reduction or there the patient. The radiographs and operative report
is continued pain. Any hardware failure or loss of should be evaluated for mechanical factors that
may have contributed to the development of
nonunion. These mechanical factors are specific
to the region being treated and are related to the
R.A. Hayda (&) type of fixation device. The clinician should
Rhode Island Hospital, Brown University Warren
assess whether the device was appropriate for the
Alpert School of Medicine, 2 Dudley St Suite 200,
Providence, RI 02905, USA type of fracture and whether it was appropriately
e-mail: roman_hayda@brown.edu implanted. Secondary questions then arise

208 R.A. Hayda, MD
whether the surgical revision should employ a A careful history that evaluates systemic
similar device or alternate implants are required. conditions and medications must be performed.
Additionally, the surgeon must consider if spe- The use of nonsteroidal anti-inflammatory drugs
cial extraction devices or techniques are required has been implicated in nonunion [7]. The inhi-
especially in the face of broken hardware. In bition of the cyclooxygenase pathway by nons-
select cases, the surgeon may elect to leave some teroidals impairs the recruitment of cellular
of the previous implant in place when extraction elements critical for the initiation of bone heal-
would pose risks without additional benefit. ing. Although intermittent use does not seem to
Biologic factors cannot be ignored when impair healing, consistent dosing has been shown
evaluating a patient with a nonunion [5, 6]. Failure to delay union in animal models and in clinical
to take these into consideration may lead to con- series [8]. Screening for infection with a com-
tinued failure after considerable effort on the part plete blood count, erythrocyte sedimentation rate,
of the patient and the surgeon. Biologic factors and C reactive protein may raise the index sus-
may be local as well as systemic. Local factors picion for infection although intraoperative cul-
that contribute to nonunion are devascularization ture is the optimal test. Nutritional status can be
of the bone whether secondary to the injury or evaluated with albumin level and total lympho-
surgical insult, infection, and even bone necrosis cyte count. Screening for vitamin D deficiency
due to irradiation (Fig. 9.1). Systemic factors and other endocrine abnormalities should be
include malnutrition, smoking, and diseases such considered [9]. Alternatively, referral to an
as diabetes and chronic illness particularly those internist or endocrinologist may assist in evalu-
treated with corticosteroids and other anabolic ation of metabolic contributors to nonunion.
suppressants. Endocrine and vitamin deficiencies Bisphophonates have been credited with
have also been implicated as contributors to reducing the risk of osteoporosis-related fractures
nonunion and should be assessed5. Whether the and are commonly prescribed in the elderly
causes are local or systemic, the surgeon and the population at risk these low-energy fractures.
patient should make an effort to mitigate the bio- However, they have been implicated in atypical
logic factors that contribute to nonunion. fractures of the femur, especially in the sub-
trochanteric area [10]. Their mechanism of action
for reducing osteoporosis is by limiting osteo-
clastic resorbtion of bone. However, in so doing
they may limit the critical function of bone
remodeling contributing to the formation of
atypical stress fractures of the femur often after 5
or more years of treatment, a time frame where
continued use offers no added benefit. At present,
there is no evidence that it impairs fracture union.
However, authors have noted that these atypical
fractures are prone to delayed union [10]. The
role of anabolic agents such as synthetic
parathyroid hormone to promote union especially
in a patient with osteoporosis is not yet defined
[11].
The surgical decision-making is specific for
the femoral neck, intertrochanteric, and sub-
trochanteric regions. However, in developing the
treatment plan for a nonunion, the surgeon must
Fig. 9.1 Subtrochanteric nonunion in irradiated bone take into account not only mechanical factors but
with failed hardware also biologic factors, local and systemic, as well
9 Proximal Femur Nonunions 209
as the patient’s functional needs. Balancing these avoid multiple procedures and prolonged dis-
issues with the potential risks of surgery will lead ability [19]. The propensity for nonunion can be
to a reasonable result. For instance, although attributed to biologic and mechanical factors. The
preservation of native anatomy especially the femoral neck is intracapsular and lacks perios-
femoral head is preferred, endoprosthetic teum, an important contributor to fracture healing.
replacement for failed treatment of femoral neck The blood supply enters through the circumflex
and intertrochanteric fractures may be the opti- vessels, which may be disrupted by the injury and
mal treatment even in a patient under 65 after surgery [20, 21]. Not only may vascular disrup-
careful consideration of the expected treatment tion impair fracture healing, but it may also
course and potential risks. In this discussion, the contribute to avascular necrosis.
surgeon should not only consider the quickest or In femoral neck fractures, the development of
most straightforward treatment option but also a nonunion is usually heralded by loss of
keep in mind the consequences of failure of the reduction especially varus collapse. Varus is
chosen course. With prosthetic replacement, the particularly unstable, leading to continued col-
potential for dislocation, periprosthetic fracture, lapse and ultimately failure. In this instance, the
and the devastating effect of infection must be surgeon should consider early revision. Short-
taken into account. The patient and surgeon must ening along the axis of the cannulated screws or
maintain awareness of the potential for such screw and side plate is often seen and may result
outcomes. In select cases, Girdlestone resection in union. However, progressive shortening
arthroplasty with its attendant limitations may without evidence of stabilization is a sign of
remain the best treatment option for problematic nonunion. Ultimately, the fixation device may
proximal femur nonunion especially in the face end up within the hip joint and failure requiring
of recalcitrant infection or a patient with signifi- revision (Fig. 9.2). Although any varus collapse
cant comorbidities or low functional demand should alert the surgeon for potential revision,
[12]. Alho et al. [22] suggested specific radiographic
criteria of failure. High rates of revision were
associated with a change in reduction of 10 mm,
9.1 Femoral Neck change in screw position by 5%, or screws
backing out by 20 mm .
Among fractures of the proximal femur, the Mechanically, the femoral neck is subject to
femoral neck is the most prone to nonunion. substantial forces while depending on its tra-
Although one series reported a nonunion rate of becular structure for mechanical strength [21].
59% and avascular necrosis rate of 86% [13], Certain patterns of injury, namely the vertically
most authors have reported nonunion rates as of oriented fracture, are prone to nonunion because
0–30% in young patients [14–16]. Union is of poorly controlled shear forces [23]. The Pau-
expected within three months of fixation although wels classification scheme based on the angle of
failure can be seen very early following fixation the fracture line to the horizontal defines 3 types:
particularly in inadequately fixed fractures or low, < 30 degrees; intermediate, between 30 and
noncompliant patients. These early failures 60 degrees, and high, greater than 60 degrees
should be considered as a nonunion even though (Fig. 9.3). Regardless of fracture pattern, fixation
3 or 6 months have not passed from the time of strategies that do not engage subchondral bone
fixation. Even surgically repaired nondisplaced will have limited ability to control this short bone
fractures have been noted to proceed to nonunion segment. The optimal treatment device has not
in up to 8% of cases indicating the multifactorial been definitively established with proponents of
nature of nonunion [17, 18]. The risk of fixation sliding screw and side plates and those favoring
failure in elderly patients with displaced fracture cannulated screws [24]. At present, data suggest
has led to the recommendation for arthroplasty to that sliding screw and side plate devices may be
210 R.A. Hayda, MD
Fig. 9.2 Progressive failure

of femoral neck “valgus
impacted” fracture with
hardware encroaching the
joint. Note failure into varus
and screws backing out (a, b)
especially varus, have impaired mechanics and

are prone to failure. In the young patient where
retention of the hip is critical, the surgeon should
strive for an anatomic reduction. Therefore, open
reduction is advocated since closed reduction
may not consistently achieve this goal. The
exception to this rule is a valgus impacted frac-
ture, usually seen in the elderly patient. This
fracture pattern is considered stable and is fixed
in situ.
9.1.1 Surgical Options
Fig. 9.3 Pauwels classification: Line A is drawn in line Options for revision in femoral neck fractures
with the femoral shaft; line B is drawn perpendicular to line
A. Lines C and D represent possible femoral neck fracture
nonunion are arthroplasty or revision fixation
lines. The angle created by line C or D to the perpendicular with or without osteotomy. In the older patient,
determines the Pauwels type: Pauwels 1, <30 (line C); arthroplasty is the primary form of treatment,
Pauwels 2, 30–60; Pauwels 3, >60 (line D) while a younger patient should be considered for
revision fixation. There is no particular age at
which revision fixation or arthroplasty should be
less prone to failure but may have a higher rate of done. Marti successfully performed osteotomies
avascular necrosis. In the case of cannulated up to age 70, while arthroplasty surgeons have
screws, placing the device adjacent to the cortex advocated replacement at age 60–65, especially
improves fracture stability by increasing screw as arthroplasty implants have improved and
spread and by taking advantage of cortical bone. patients are less tolerant of lengthy rehabilitation
Placing cannulated screws more centrally has [16, 19]. Careful consideration of expected clin-
been compared to placing them in an “empty ical course and potential complications should be
can” in the elderly patient contributing to fixation used in choosing treatment (Table 9.1) [26–34].
failure [25] . Revision fixation requires a healing period prior
The quality of reduction is a determinant of to the resumption of activities, which may last
stability [14]. Fractures with poor reduction, 3–6 months and does not guarantee success.
Table 9.1 Comparative table of treatment of femoral neck nonunion
9
Study Number Age Avg time Technique Follow-up % Complications Outcome

from index Union
surgery
(mos)
Marti et al. 50 19– 9 (2−60) Osteotomy; 120° 7.1 (3–13) 86 3 UTI, 3 DVT, 1 deep infection Harris score 91, Merle
[26] 76 double-angle blade plate, years (ankylosis) D’Aubigne 78% good to
local bone graft 14% THR (3 nonunion, 3 AVN, excellent; AVN in 22 pts, 3
1 failed hardware (Etoh abuse) requiring THR,
Anglen 13 34 21 (4–54) Osteotomy; 120 deg 2 years 100% AVN in 2 pts-THR Harris 93, no pain 7/8 return to
Proximal Femur Nonunions
et al. [27] (18– double-angle plate, local (0.7–3.5) work, avg LLD 1.5 cm
59) bone graft
Shoenfeld 4 35 6 Opening/closing wedge 1.25 100% None Merle D’Aubigne 16.5;
et al. [28] (24– osteotomy; cannulated 130– (0.25– cane 1 pt, no LLD but sl
42) 140° sideplate, local bone 2.3) years Trendelenburg 2/4
graft
Hartford 8 46 10 Closing wedge, cannulated 2 years 88% Unrelated death at 4 months Harris hip score 73, cane 1 pt.
et al. [29] (30– 135–155 sideplate, local Trendelenburg 4/7
65) bone graft
Said et al. 17a 37 130 single-angle blade, 3.5 (1–5) 97%a AVN 5 cases 61% pain free, LLD < 0.5 cma
[30] (18– closing wedge 1 THR for nonuniona
49)
Lecroy 22 (all with 28.7 18.3 Free fibula, cannulated 7 years 90% 2 nonunion (1 iliac crest graft, 1 60% progression of AVN but
et al. [31] AVN at screws Meyers graft); 4 hardware 90% retained native hip, Harris
presentation) removal for articular penetration score 78.9
Jun et al. 26 41 12 Free fibular in anterior 3.5 (1–5) 100% 1 infection, 1 AVN (THR) Harris score 88
[32] (22– trough years
60)
Elgafy 17 46 8 (4–22) 13 autograft 69% 4 nonunion-arthroplasty, 2 ankle
et al. [33] (24– non-vascularized fibula; pain
58) cannulated screws
6 allograft fibula; cannulated 33% 3 arthroplasty, 1 autograft fibula
screws
Wu et al. 26 38 1.4 (0.8–2) 17 osteotomy with sliding >2 years 100% 2 AVN, 1 nonunion osteotomy Osteotomy indicated if
[34] (17– screw (2–6) shortening was >1.5 cm
60)
211
9 sliding screw only 100% None

AVN avascular necrosis; UTI urinary tract infection; THR total hip replacement; LLD limb length discrepancy
a
Series included 19 acute fractures not reported separately
212 R.A. Hayda, MD
However, it allows for preservation of the native lesser abductor moment are generally well tol-
hip joint. Even the presence of low-grade avas- erated but should be discussed with the patient.
cular necrosis is not a contraindication to hip In this procedure, the nonunion is not exposed
preservation [15, 26, 31]. In the uncommon directly. A blade plate is most often used to
instance of biologic failure, the healing envi- compress the osteotomy, but a screw and side
ronment can be augmented by autogenous or plate may also be used [28, 29, 34]. A blade plate
vascularized bone graft. The Meyers graft [35, is technically more demanding but offers proven
36] which utilizes the bony insertion of the stability while requiring only a narrow corridor
quadratus femoris placed into the posterior of bone.
femoral neck or a vascularized fibula as advo- The critical step in performing the inter-
cated by LeCroy et al. [31] can provide the trochanteric osteotomy is preoperative planning
stimulus to achieve union. (Fig. 9.4). The concept is to convert the angle of the
fracture to less than 30° by performing an inter-
trochanteric osteotomy. Good anteroposterior pel-
9.1.2 Revision Fixation vis and lateral hip views are required. The angle of
the fracture to the horizontal is measured. Due to leg
Mechanical failure is most common with failure rotation, the precise angle may be hard to measure,
of fixation and development of a deformity on but may be facilitated by measuring it to a line
radiographic examination. In femoral neck non- perpendicular to the femoral shaft [27]. A closing
union attributed to mechanical failure, the prin- wedge osteotomy is planned to correct the angle of
cipal decision is whether osteotomy will improve the fracture to less than 30°. In the case of a fracture
the mechanical environment. On analysis of the angle of 70°, a wedge of 40° is planned.
nonunion, the new construct must assure ade- The position of the blade plate that allows for
quate stability in the face of physiologic forces to adequate fixation is templated. Inevitably, the
allow for healing. In the instance of a poorly biomechanics of the hip are altered with some
placed initial construct, revision fixation alone shortening of the neck and medialization of the
may be considered. Cannulated screws in the shaft. Some modifications of the osteotomy have
most optimal position or a screw and side plate been proposed that can minimize the anatomic
device with screw augmentation may allow for alterations. A combination of opening and clos-
healing. Wu, in a mixed series with and without ing wedge osteotomy or translating the
osteotomy, reported 100% healing in the group shaft laterally along the osteotomy are methods
treated with a screw and side plate alone [34]. used to limit biomechanical alterations but may
carry a theoretic risk of nonunion.
The intertrochanteric osteotomy is executed
9.1.3 Valgus Osteotomy on a radiolucent table. The surgeon must assure
adequate radiographic visualization of the hip
In high angle fractures often seen in the younger prior to create the sterile field. The proximal
patient, an osteotomy is indicated when a non- femur is exposed through a lateral approach, but
union develops. These nonunions are often the nonunion site is not opened. The previous
associated with varus and shortening. Properly hardware is removed. The projected osteotomy
executed osteotomies result in union rate of 70– site is marked as well as the position of the fix-
100% [26–29, 34]. As described by Pauwels, the ation device. The pathway for proximal fixation
osteotomy converts a vertically oriented fracture should be established prior to the osteotomy
with shear to a more horizontal orientation cre- since it is very difficult to place a blade plate after
ating compressive forces promoting union. In the osteotomy is completed. For a blade plate, a
considering an osteotomy, the duration of heal- guide wire and drill bits establish the pathway for
ing along with alterations of anatomy must be the chisel. The pathway starts on the greater
considered. These alterations of shortening and a trochanter skirts the piriformis fossa, across the
Fig. 9.4 Corrective osteotomy of femoral neck nonunion: an angle of 30°, and osteotomy is planned 1.5 cm distal to
a injury radiograph note vertical angle; b initial fixation; projected blade insertion site; e intraoperative templating;
c nonunion, note varus, and shortening; d surgical tem- f placement of chisel anteroposterior view; g placement of
plate—fracture angle is 70°, a 40° correction will result in chisel lateral view; h healed nonunion and osteotomy
nonunion and terminates in the inferior medial fixation. Displacement in the lateral view will
femoral head. The chisel is used to create a render subsequent arthroplasty if needed much
pathway for the blade plate. Frequent fluoro- more difficult. The harvested wedge is morselized
scopic views in both the anteroposterior and and placed around the osteotomy site. Postoper-
lateral views assure that correct placement. atively, touch weight bearing and then protected
The planned closing wedge osteotomy is weight bearing are allowed until healing is seen,
marked with Kirschner wires. The proximal cut is usually in 3 months. Patients are cautioned that
parallel with the blade plate track and at least 1.5– persistent limp and mild leg length difference are
2 cm distal to provide for a bone bridge. The common while preserving the native hip.
osteotomy is then performed with a saw while Results of valgus intertrochantgeric osteot-
cooling the blade with irrigation making sure not omy have been reported by a number of authors
to cause thermal necrosis. It is also important to consistently reporting union from 70 to 100% of
keep the osteotomy perpendicular to the shaft to patients with avascular necrosis occurring around
avoid flexion or extension. The selected blade 20% [26–29, 34]. Marti et al. in a series of 50
plate is seated, and then, the osteotomy is com- patients with a mean age of 53 achieved a union
pressed with a compression device placed distally rate of 86% [26]. Of 22 patients that had evi-
on the shaft. On the anteroposterior view, care is dence of osteonecrosis without collapse only 3
taken to avoid excessive medialization along the progressed requiring arthroplasty. An additional
osteotomy as it contributes to not only to limb 3 patients required arthroplasty for continued
shortening but also to loss of abduction and val- nonunion and one for periprosthetic fracture for
gus alignment of the knee. A longer blade and an arthroplasty rate of 14% with a mean
lateralization of the shaft limits this problem. follow-up of 7 years. In this series, the Harris hip
Close attention is paid on the lateral view and score was 91, while 78% were good or excellent
rotation to assure that these are maintained in final on the Merle d’Aubigné score. Anglen had 100%
214 R.A. Hayda, MD
union rate in 13 patients although 2 required greater trochanter. The fibula is placed into the
arthroplasty by 2 years for osteonecrosis [27]. head and neck and stabilized with a K-wire after
Ballmer et al. reported that in 17 patients, 71% the fibula is inserted. The critical vascular anas-
healed following the index procedure while an tomosis is then performed. Further stability is
additional 3 healed following additional bone obtained by placing cannulated screws parallel to
grafting for an overall union rate of 88% [37]. the free fibula. Postoperatively the patient is
Three, however, required subsequent arthroplasty nonweight bearing for 6 weeks and then gradu-
for avascular necrosis. In Hartford et al.’s series ally progressed thereafter. This technique is
of 8 patients treated with a sliding hip screw clearly technically demanding but is useful in the
union was achieved in all cases and Harris hip face of avascular necrosis.
score improved from 24 to 73 [29]. Another The series reported by LeCroy et al. describes
more recent series reported the use of a 22 patients (mean age 29) who underwent a
single-angled 130 degree plate [30]. In 36 vascularized fibula repair of a nonunion at a
patients (mean age 37), union was achieved in mean of 18 months following initial fixation
97%, while 5 developed avascular necrosis. 61% [31]. Union was achieved in all patients at an
reported no pain, and the average leg length average of 9 months but 2 required additional
deficit of 2.5 cm was corrected to 0.5 cm. The surgeries to achieve union. Osteonecrosis was
authors of this series felt that the single-angle present in all 22 cases at reconstruction surgery
plate limited medialization while optimizing (predominantly Steinberg Stage II, 12 cases, but
length. Other authors, using the double-angle also Stage I, 4 cases; Stage III, 2 cases; Stage IV,
plate, have been able to control this problem by 3 cases; even Stage V, 1 case). Despite
laterally translating the femur. Unfortunately, osteonecrosis progressing in 13 patients, the
more detailed outcomes measures are not avail- native hip was retained in 20 patients with an
able to judge functional status or compare average Harris hip score of 78.9. Although only
methods of treatment. 5 reported being able to participate in vigorous
physical activity that included running, 16 were
able to perform moderate activity.
9.1.4 Vascularized Graft Recently, an alternate method of free vascu-
larized fibula was reported by Jun and colleagues
A vascularized fibular strut or vascular iliac crest [32]. In the reported series of 26 cases, the free
has been proposed to enhance vascularity while fibula was inserted via an anterior approach into
correcting deformity and enhancing stability [31, a trough facilitating placement and simplifying
38, 39]. The fibular technique has been particu- vascular re-anastomosis. Twenty-four healed at
larly advocated in nonunions with associated an average of 5 months. One case of postperative
avascular necrosis provided the acetabulum is osteonecrosis was observed. Unlike the series of
intact, but this is technically demanding and Lecroy where all had osteonecrosis, in this series
requires microvascular anastomosis. only 1 patient had radiographic findings of
The surgical technique described involves osteonecrosis preoperatively. Outcomes were
removal of hardware and revision fixation. reported in terms of the Harris hip score, which
Through a Watson-Jones approach, a channel is improved on average to 87.9.
created in the lateral femur into the neck and The use of nonvascular fibular grafts has been
head to accommodate the harvested fibula. The reported by Elgafy and associates in 19 cases
necrotic bone in the head if present is removed with minimal varus malalignment [33]. In that
with special burrs. Correction of the femoral small series, nonvascular fibular sutografts
neck deformity is obtained from the mobility achieved union in 9 of 13 cases (69%). The same
gained when the channel is created. The defect in authors reported union in only 2 of 6 cases in
the head is filled with cancellous bone from the which a fibular allograft was used. Despite the
relative technical ease of such a technique, there stock, a press fit stem is preferred. However, the
is no advantage in using a nonvascular fibular screw holes may predispose to fracture. Pro-
strut in the treatment of femoral neck nonunion. phylactic cabling should be considered to avoid
iatrogenic fracture. If cementing is selected, the
surgeon should take measures to avoid extrusion
9.1.5 Arthroplasty of cement by plugging the holes with bone,
cement, or hardware. Intraoperatively, the surgi-
Arthoplasty particularly with modern bearing cal site should be inspected and excess cement
surfaces should be considered in femoral neck removed. The approach is dictated by previous
nonunion not only in the face of acetabular incisions, hardware, and surgeon preference
destruction or avascular necrosis. In the patient although the anterior approach may have lower
over 60, arthroplasty is likely to provide a reli- dislocation rates [46]. Postoperatively, hip posi-
able method to return the patient to activity but tion restrictions are dictated by the approach, but
does not match the excellent results achieved full weight bearing is usually allowed
with primary arthroplasty for osteoarthritis. immediately.
Although in the series of Marti et al. osteotomy Outcomes following salvage arthroplasty have
was generally preferred in patients up to age 70 been described in a limited degree but are con-
[26] and by Hitt to age 60 [19], modern tech- sistently not as good as primary arthroplasty for
niques and implants have lowered the age in osteoarthritis. McKinley and Robinson con-
which arthroplasty is considered. Particularly in cluded that cemented total arthroplasty was the
the younger patient, considerations should preferred treatment to salvage of failed femoral
include the potential for infection, dislocation, neck fixation with arthroplasty while noting
and periprosthetic fracture. Although the inci- results were inferior to primary arthroplasty for
dence for each of these complications is low, for fracture [47]. In a matched group of 107 dis-
the patient with an infection, for instance, the placed femoral neck fractures, primary arthro-
outcomes are greatly diminished. It is well plasty cases compared to 107 salvage patients
established that arthroplasty in the face of pre- with failed fixation, the primary arthoplasty
vious surgery raises the risk of infection [40]. group had fewer dislocations (8 vs. 21%) and
The surgeon planning the arthroplasty should fewer infections. The salvage group also had
also consider the influence of disuse and the worse function and lower implant survivorship at
previous hardware on fixation of the stem 5 and 10 years. A series by Mabry et al.
(Fig. 9.5). In the elderly patient in which
arthroplasty is the preferred treatment for non-
union, osteopenia of both the proximal femur and
the acetabulum may be present [41–45]. Press fit
components should be used with caution partic-
ularly when compared to the sclerotic bone pre-
sent in patients with osteoarthrtitis [46]. Relative
osteopenia of the acetabulum may present prob-
lems with preparation and fixation. Sharp ream-
ers in nonsclerotic bone may lead to excess
medializaton and even iatrogenic protrusion. Cup
preparation should be deliberate to avoid this
problem. Screw augmentation of cup fixation
should be also considered to assure stable fixa-
tion for ingrowth.
Fig. 9.5 Total hip arthroplasty for femoral neck non-
Secure fixation of the stem presents another union performed for case in Fig. 9.2. Note screw
challenge. In the younger patient with good bone augmentation of cup and cable of proximal femur
216 R.A. Hayda, MD
describes longer-term outcome of Charnley capacity [51–54]. These may present to the sur-
arthroplasty for femoral neck nonunion [48]. geon with or without signs of avascular necrosis.
With a mean age of 68 years at time of surgery Contrary to popular belief, these patients are not
and a mean of 12 years follow-up, 10-year sur- condemned to developing avascular necrosis or
vivorship was 93 and 76% at 20 years. Dislo- poor results although these cases are challenging.
cation was seen in 9% of cases. Revision surgery Successful repair with salvage of the native
was performed in 12 of 99 cases: 10 for aseptic femoral head can be achieved with careful plan-
loosening, 1 for dislocation, and 1 Girdlestone ning and execution of surgery. In addition to plain
arthroplasty for infection. In terms of function, radiographs, CT scan provides valuable infor-
96% had no or mild pain, while 33% required no mation regarding fracture orientation and bone
ambulatory aids and 36% used a single cane or loss. Evaluation of options and careful operative
crutch for support; 11% could not walk. planning should proceed with evaluation of the
degree of bony erosion at the fracture site. In the
young patient, hip preservation with appropriate
9.1.6 Arthroplasty for Failed fixation should be performed (Fig. 9.6).
Osteotomy Although some shortening of the femoral
neck is common, anteversion should be restored
In the event of a failed intertrochanteric along with critical coronal plane alignment.
osteotomy, arthroplasty offers a means of Fixation constructs such as those used in fresh
restoring function. The altered anatomy may fractures may be used if fracture orientation is not
present challenges to the surgeon in placing the vertical and reduction can be obtained. Cancel-
stem. Ferguson and colleagues noted a high rate lous bone graft should be considered. However, a
of complications in his series of 305 cases valgus intertrochanteric osteotomy should
although 79% were noted to have good to strongly be considered at the outset in most cases
excellent results [49]. There were technical even to include intermediate Pauwels angles.
challenges in 23% of cases with perioperative Lifeso and Younge in their series of 31
complication in 12 and 3.2% infection rate. neglected femoral neck fractures (ages 16–90)
Revision was required in 10 years for 18% of treated 19 with arthroplasty, 3 with internal fixa-
cases with stem loosening predominating. In tion, and 6 with valgus intertrochanteric osteot-
contrast, Boos et al. reported a case–control omy at an average of 34 months post-injury [55].
series of 74 arthroplasty cases following osteot- Three patients with nondisplaced fractures
omy with 74 primary arthroplasties finding few underwent fixation alone; however, only 1 had a
significant differences [50]. Ten-year survival good result. The 6 patients with displaced
was slightly less at 82% as opposed to 90%, but neglected fractures undergoing valgus osteotomy
perioperative complications were equal at 11%. were all under 40 years of age. Of these, 3 were
Revisions for sepsis were also slightly higher at considered to have a good result achieving union,
8% compared to 3%. while 2 had fixation failure and one avascular
necrosis. More detailed outcome data is not pro-
vided but demonstrates that successful salvage is
9.1.7 Neglected Femoral Neck possible in this young patient population with
Fracture neglected fractures. A more recent series of 32
patients in India achieved union in 91% of cases
Although uncommon in the developed world, [56]. The surgeons performed a valgus inter-
missed or delayed diagnosis of femoral neck trochanteric osteotomy at an average of 6 months
fracture may present for treatment weeks or post-injury. The average patient age was 42.
months after injury. Such an instance may be Follow-up at an average of 5 years revealed a
considered a “neglected” femoral neck fracture, Harris hip score of 82. Three developed non-
which presents with pain and limited ambulatory unions, and six had stage 3 or 4 avascular
Fig. 9.6 Delayed diagnosis of femoral neck fracture: a at utilized. Iliac crest graft was utilized to fill the defect.
presentation, b CT axial note apex anterior deformity, A universal distractor placed supra-acetabular provided
c CT coronal note varus, d healed after fixation. distraction intraoperatively
Combined Smith Peterson and lateral approach was
necrosis. The authors noted that fracture resorp- Nonetheless, nonunion must be in the differential
tion and excessive valgus correlated with poor if the patient has persistent pain after fixation or
outcome, while the presence of avascular necrosis if the hardware fails. Visualization of the non-
did not. Although there is limited data on treat- union may be difficult due to hardware obscuring
ment of this uncommon presentation of the the fracture line, which also may be out of the
neglected femoral neck, successful salvage is plane of standard radiographs. Computerized
achievable in the young patient where arthro- tomography is invaluable in differentiating non-
plasty is not desired. union from malunion. The lesser trochanter may
remain ununited, but this is usually not prob-
lematic. Other areas without bony bridging may
9.2 Intertrochanteric Fracture be seen but likely do not constitute a nonunion or
Nonunion delayed union unless 50% or more is not bridged.
No precise definition exists here making clinician
Intertrochanteric fractures unlike femoral neck judgment critical [2, 4].
fractures almost always heal although malunion Once the diagnosis of intertrochanteric non-
is not uncommon. The metaphyseal area with union is established, diagnostic workup based
its good blood supply promotes healing. on history and physical is performed with
218 R.A. Hayda, MD
appropriate laboratory and radiographic evalua- pression axially while correcting the varus
tion as outlined earlier in the chapter. Good deformity. The use of proximal femur-locking
quality anteroposterior pelvis and full femur plates may be successful in select cases but has
films are important to devise the surgical plan. been associated with a number of reported fail-
Aside from patient age and functional status, ures [57, 58].
decision-making depends on the status of previ- Varus deformity and lucency around the
ous hardware, the quality of the femoral head and implant are signs of nonunion. Hardware failure
neck, and critically whether the acetabulum is may be seen in the form of pull out or breakage
intact. In the younger, more active patient in an intertrochanteric nonunion. Intramedullary
achieving union is preferred, while in the older, devices may fracture at the junction of the lag
lower demand patient arthroplasty should be screw and the rod. Broken screws are removed
considered. with conventional techniques. A broken rod may
In the presence of satisfactory alignment with be more difficult. The use of an extraction hook
intact hardware and hip articulation, the option can capture the distal end allowing for its
for simple bone grafting with or without hard- removal.
ware revision can be successful. Otherwise, the Given the biomechanical advantages of
hardware should be revised. An intramedullary intramedullary devices and new designs, revision
device offers a biomechanical advantage while fixation is most commonly used by the author
also providing bone graft from reaming. The although no series is available to support this
intramedullary device lies in the axis of force of trend (Fig. 9.7). However, blade plate is an
the proximal femur, while a side plate is more appropriate option if a deformity needs to be
lateral, subjecting it to cantilevering when the corrected, if alternate proximal fixation is
side plate is loaded with weight bearing. With a required, or failure of exchange reamed nailing.
variety of angles available for the lag screw, an In the elderly patient or when the hip articula-
intramedullary implant can be placed to replace a tion is damaged, arthroplasty offers a reliable
variety of previous devices. A blade plate offers method for restoring mobility (Fig. 9.8). A hemi-
alternate fixation of the proximal segment arthroplasty can be used if the acetabulum is intact.
entering very proximally laterally while taking A unipolar head allows for improved stability
advantage of the often-pristine area of the infe- compared to a total arthroplasty at the risk of
rior head and neck. It can also allow for com- acetabular wear and arthritis. A calcar-replacing
Fig. 9.7 Nonunion of intertrochanteric fracture: a broken nail, b healed with revision with larger nail
antibiotics, and subsequent revision. Revision

implants were most commonly low angle blade
plates (11 cases), and autologous bone graft was
used in 17 cases. Other devices included
dynamic hip and dynamic condylar screws. Only
1 older generation intramedullary device was
used. In 4 cases, an osteotomy was performed to
correct the deformity. The one failure was con-
verted to a cemented arthroplasty. No patient
under 60 required ambulatory supports, 8
required a single cane, and 4 used a walker. The
authors commented that effective salvage is
possible in the physiologically younger patient
with good bone with the ability to achieve stable
fixation and an intact hip articulation.
Other reported series consisted of 11 or fewer
patients in a similar age group [60–62]. The Sar-
athy et al. series [61] consisted of previously
unoperated cases treated with a medial displace-
ment osteotomy, a technique predating the wide-
Fig. 9.8 Total hip arthroplasty for intertrochanteric
nonunion with fixation of trochanter spread use of internal fixation. Similar fixation
strategy was used in all 3 series; however, they
capitalized on fixation in the inferomedial femoral
stem can be used, or conventional or revision neck and utilized autologous bone graft. The union
arthroplasty stems can be used, provided that distal success rate was uniformly over 90%. Another
fixation is achieved. When planning for the small series reported on 4 cases of varus nonunion
arthroplasty, restoration of normal biomechanics and malunion treated with an intertrochanteric
with restoration of length and offset will provide osteotomy with a 120° blade plate similar to a
optimal results. However, changes from the sur- femoral neck nonunion [63]. All cases healed
gery and subsequent deformation may lead to while regaining an average of 2 cm of leg length
contracted soft tissues and weak abductors. The and the Harris hip score improved from 73 to 92.
most critical factor is to restore stability of the hip The experience with arthroplasty is similarly
by retaining critical soft tissue and establishing limited [64–67] as a treatment for inter-
correct length. Modular components assist in trochanteric nonunion. The principles and
achieving this goal. potential complications mirror the experience
There are few reported series of treatment of with arthroplasty treatment for femoral neck
intertrochanteric nonunions and all predate the nonunion. A series of 60 patients with an average
widespread use of intramedullary implants. Hai- age of 78 years from the Mayo clinic were
dukewych and Berry reported on 20 cases that treated with arthroplasty [64], 28 with hemi-
underwent revision, achieving union in all but arthroplasty, and the remainder with predomi-
one with resolution of pain [59]. The patients nantly cemented total hip arthroplasty.
were older with an average age of 58 (range 21– Calcar-replacing stems were used in 51 cases.
86) presenting at a mean of 12 months from Of 44 survivors not lost to follow-up, 39 walked
initial injury. The previous implant was a plate with minimal pain. Two patients sustained femur
device in all cases (14 sliding screw and side fractures during surgery managed by cerclage
plate) while only one was a cephalomedullary wiring. One patient had 2 episodes of dislocation
nail. Two cases of infected nonunion underwent each treated with closed reduction. Survivorship
staged treatment with removal, interval was 100% at 7 years and 87.5% at 10 years
220 R.A. Hayda, MD
indicating excellent durability. The two revisions In the surgical plan, the surgeon must take
were at 8 and 10 years for aseptic loosening and into account the status of the existing bone and
severe osteolysis. The authors noted that careful the hardware present. In many cases of nonunion
surgical technique can lead to very functional with a preexisting nail, exchange reamed nailing
results with few complications. Hemiarthroplasty allows for use of previous approaches while the
in cases of lower demand and well-preserved reaming provides bone graft when no significant
articular cartilage did not appear to result in deformity is present. Both reconstruction nails
symptomatic wear in this series. and third-generation cephalomedullary nails exist
Hammad et al. reported on a younger cohort in larger diameters of 13 and even 15 mm upon
of 32 patients with an average age of 64 [65]. request from the manufacturers enhancing not
Again, most underwent cemented arthroplasty only the strength of the nail but also the fill
but 5 were uncemented and 7 hybrid. Perioper- within the canal. The use of dynamic interlocking
ative complications were low with one fracture, can further enhance compression at the fracture.
one dislocation, and 3 deep venous thromboses. Alternatively, bone graft in situ can be consid-
On follow-up at a mean of 5 years, Harris hip ered when the fracture implant, nail, or plate is
scores were markedly improved from a preop- intact. However, the surgeon must be satisfied
erative score of 24–84 postoperatively. Com- that the construct provides appropriate stability
munity ambulation was noted in 88% although and no significant deformity is present.
half required a single cane or crutch. A blade plate or a 95° condylar plate placed
centrally into the inferior femoral head is a useful
alternative device particularly when alternate
9.3 Subtrochanteric Nonunion control of the proximal segment is felt to be
necessary (Fig. 9.9). With both of these devices,
Subtrochanteric fractures have historically pre- additional screws can be placed through the plate
sented greater difficulty in achieving satisfactory into the proximal segment enhancing its control.
union [68]. The high biomechanical loads and Locking proximal femur plates offer another
deforming forces present challenges to the sur- alternative here but have been associated with
geon and the implant. However, using modern failure [57, 58].
implants with the emphasis on preserving the Several retrospective clinical series report
local blood supply to bone, high union rates are results of nonunion treatment. Haidukewych and
achieved. Kinast and colleagues achieved union Berry reported a 95% union rate in a series of 24
rates of 100% using a blade plate and indirect cases in patients with an age of 55 years (range
reduction techniques [69]. Other modern series 16–88) presenting 21 months after injury [74].
particularly those using modern standard and The majority had initial treatment with plate
cephalomedullary nails noted union rates of 92– devices. Nonunion surgery consisted of hardware
100% [70–73]. removal and revision fixation. Fixation con-
The radiographic evaluation of this region is structs included: 15 nails (7 reconstruction, 1
more straightforward that the femoral neck and cephalomedullary, 7 standard antegrade); and 7
the intertrochanteric region in assessing union. plates (5 blade plates, 1 screw and side plate, 1
Nonetheless, computerized tomography may be dynamic condylar screw, and one dual plating).
useful to evaluate the location and extent of the Autograft was used in 8 patients although 1
nonunion. Clinical and radiographic assessment patient with four previous failed nailing attempts
of length, angulation, and rotation is also underwent vascularized fibular grafting in addi-
important in creating the surgical plan and tion to nail stabilization. Eleven of 21 walked
informing the patient of expected outcomes. without ambulatory aids. When shortening is
Flexion and varus deformity are often encoun- noted, restoration of native length can be
tered and should be corrected to the maximal achieved in one stage with use of bone graft and
extent possible. an intramedullary nail. A series reported 100%
Fig. 9.9 Subtrochanteric

nonunion: a nonunion after
first revision, b revision with
intramedullary allograft
fibula, autogenous bone graft
and blade plate
union and restoration of length in 21 cases of anastomosis, a stable construct must be created
aseptic nonunion with shortening of 2–5 cm at for both the fibular strut and the femur as whole.
presentation [75]. The fibula may be inset into the medullary canal
Another series of 29 patients treated exclu- or into a trough anteriorly or even medially.
sively with a cephalomedullary nail achieved A plate may then stabilize the femur. Union must
successful union in 88% of cases following the be achieved of the fibula to the proximal and
initial surgery and 96% with a single additional distal femur segments. The segment may then
procedure [76]. This group was slightly older gradually hypertrophy in response to physiologic
(average age 63) but also included young patients stress of loading the femur. Weight bearing and
with high-energy injuries, with 11 patients activity must be limited and gradually advance in
between the age of 26 and 50. Similar to the the course of the healing process.
Mayo series [74], 66% had been previously Duffy reported on 4 patients who sustained
treated with plate devices. Although they repor- proximal femur fractures following an average of
ted that 81% returned to their preinjury status, 5500 gray irradiation for bony and soft tissue
outcome data does not otherwise allow compar- malignancy in a mixed series of 17 patients with
ison of techniques to determine whether intra- radiation associated fractures in various locations
medullary devices or plating is superior. [77]. The fracture occurred at an average of
Decision-making should be based on the ability 111 months after irradiation. The age of these
to correct deformity if present and maintain fix- patients at the time of free fibular transfer was
ation in the proximal segment to achieve union. between 13 and 82 years of age and had an
In cases where the biologic environment for established nonunion for 15 months on average.
healing is severely compromised due to the In addition to the free fibula, the femur was
presence of segmental bone loss or from devas- stabilized with an intramedullary rod. Three of
cularization from multiple surgical procedures, four had excellent results, while the fair result
infection, or irradiation, a vascularized fibula strut was associated with infection requiring debride-
may allow for bypassing of the affected segment ment and had continued pain. It should be noted
into an environment that allows for healing. In that one of the excellent results sustained a sec-
addition to the technical challenge of the vascular ond fracture distal to the vascularized fibula
222 R.A. Hayda, MD
Fig. 9.10 Atypical fracture

of the proximal femur:
a nonunion with broken nail
(initially fixed in slight varus),
b healed after revision
fixation with larger nail
requiring a second vascularized fibula. The varus alignment, the use of a blade plate or
authors advise that fractures in irradiated bone dynamic condylar plate can correct the deformity
should undergo vascularized fibular grafting and compress across the fracture.
when nonvascular fibula graft with an appropri-
ate fixation fails to achieve union in 6 months.
In a very unusual circumstance, proximal 9.4 Summary
femoral replacement can be considered as a
reconstructive option. Stability of the hip joint Nonunion of the proximal femur is most com-
and the function of the hip abductors are a mon in the femoral neck although it may also be
challenge to restore. seen in the intertrochanteric and subtrochanteric
Another scenario for delayed or nonunion of region. Often the nonunion is evidenced by
the proximal femur occurs in association with the persistent fracture gap, increasing deformity, or
so-called atypical fractures femur in association broken hardware. It is imperative that the sur-
with prolonged bisphosphonate use (Fig. 9.10) geon perform not only a careful radiographic
[10]. These are often transverse fractures in the analysis of the nonunion but also investigate all
proximal femur with an associated thickening of potential contributing factors to the nonunion.
the cortex from the chronic stress. Trivial injury Repair of the nonunion then depends on correc-
can complete the fracture, or the patient may tion of the deformity, and an improved biome-
simply present with chronic pain in the femur chanical construct and where appropriate
with radiographic changes of a stress fracture. augmentation of the biologic potential for heal-
These atypical fractures occur usually after ing with the use of bone graft. Successful union
5 years of bisphosphonate use. can be achieved with a careful evaluation,
At the time of initial treatment, bisphospho- thoughtful planning, and skillful execution of the
nate treatment should be terminated. The use of surgical plan in most cases. In some cases
dynamic interlock allows for compression with depending on the patient’s age functional status
weight bearing. In the event of nonunion, and the status of the hip articulation, arthroplasty
exchange reamed nailing will typically result in offers the best option to functional restoration.
success. However, if initial treatment resulted in Careful discussion with the patient is required for
them to understand the expected postoperative 14. Swiontkowski MF, Winquist RA, Hansen ST Jr.
course along with the potential for complications. Fractures of the femoral neck in patients between the
ages of twelve and forty-nine years. J Bone Joint
Surg Am. 1984;66(6):837–46.
15. Angelini M, McKee MD, Waddell JP,
Haidukewych G, Schemitsch EH. Salvage of failed
References hip fracture fixation. J Orthop Trauma. 2009;23
(6):471–8.
1. Cornwall R, Gilbert MS, Koval KJ, Strauss E, 16. Haidukewych GJ, Rothwell WS, Jacofsky DJ,
Siu AL. Functional outcomes and mortality vary Torchia ME, Berry DJ. Operative treatment of
among different types of hip fractures: a function of femoral neck fractures in patients between the ages
patient characteristics. Clin Orthop Relat Res. of fifteen and fifty years. J Bone Joint Surg Am.
2004;425:64–71. 2004;86-A(8):1711–1716.
2. Morshed S. Current options for determining fracture 17. Parker MJ, Raghavan R, Gurusamy K. Incidence of
union. Adv Med. 2014, Article ID 708574. fracture-healing complications after femoral neck
doi:9.1155/2014/708574. fractures. Clin Orthop Relat Res. 2007;458:175–9.
3. Bhandari M, Chiavaras M, Ayeni O, Chakraverrty R, 18. Conn KS, Parker MJ. Undisplaced intracapsular hip
Parasu N, Choudur H, et al. Assessment group for fractures: results of internal fixation in 375 patients.
radiographic evaluation and evidence (AGREE) Clin Orthop Relat Res. 2004;421:249–54.
Study Group (AGREE Investigators Writing Com- 19. Hitt K. Femoral neck nonunion: osteotomy or
mittee). Assessment of radiographic fracture healing arthroplasty. Tech Orthop. 2001;17(4):434–42.
in patients with operatively treated femoral neck 20. Trueta J, Harrison MH. The normal vascular anatomy
fractures. J Orthop Trauma. 2013;27(9):e213–9. of the femoral head in adult man. J Bone Joint Surg
4. Bhandari M, Chiavaras MM, Parasu N, Choudur H, Br. 1953;35-B(3):442–461.
Ayeni O, Chakravertty R, et al. Radiographic union 21. Hughes PE, Hsu JC, Matava MJ. Hip anatomy and
score for hip substantially improves agreement biomechanics in the athlete. Sports Med Arthrosc
between surgeons and radiologists. BMC Muscu- Rev. 2002;10(2):103–14.
loskelet Disord. 2013;14:70. 22. Alho A, Benterud JG, Solovieva S. Internally fixed
5. Brinker MR, O’Connor D, Monla Y, Earth- femoral neck fractures: early preduction of failure in
man TP. Metabolic and endocrine abnormalities in 203 elderly patients with displaced fractures. Acta
patients with nonunions. J Orthop Trauma. 2007;21 Orthop Scand. 1999;70(2):141–4.
(8):557–70. 23. Bartonícek J. Pauwels’ classification of femoral neck
6. Koval KJ, Maurer SG, Su ET, Aharonoff GB, fractures: correct interpretation of the original.
Zuckerman JD. The effects of nutritional status on J Orthop Trauma. 2001;15(5):358–360. Review.
outcome after hip fracture. J Orthop Trauma. 1999;13 24. Goldstein C, Petrisor BA, Ferguson T, Bhandari M.
(3):164–9. Implants for fixation of femoral neck fractures. Tech
7. Dodwell ER, Latorre JG, Parisini E, Zwettler E, Orthop. 2008;23(4):301–8.
Chandra D, Mulpuri K, Snyder B. NSAID exposure 25. Gurusamy, Parker MJ, Rowlands TK. The compli-
and risk of nonunion: a meta-analysis of case–control cations of displaced intracapsular fractures of the hip
and cohort studies. Calcif Tissue Int. 2010;87 the effect of screw positioning and angulation on
(3):193–202. fracture healing. J Bone Joint Surg Br. 2005;87
8. Gerstenfeld LC, Thiede M, Seibert K, Mielke C, (5):632–634.
Phippard D, Svagr B, et al. Differential inhibition 26. Marti RK, Schüller HM, Raaymakers EL. Inter-
of fracture healing by non-selective and trochanteric osteotomy for non-union of the femoral
cyclooxygenase-2 selective non-steroidal neck. J Bone Joint Surg Br. 1989;71(5):782–7.
anti-inflammatory drugs. J Orthop Res. 2003;21 27. Anglen JO. Intertrochanteric osteotomy for failed
(4):670–5. internal fixation of femoral neck fracture. Clin
9. Matthews V, Cabanela ME. Femoral neck nonunion Orthop. 1997;341:175–82.
treatment. Clin Orthop. 2004;419:57–64. 28. Schoenfeld AJ, Vrabec GA. Valgus osteotomy of the
10. Blood T, Feller RJ, Cohen E, Born CT, Hayda R. proximal femur with sliding hip screw for the
Atypical fractures of the femur: evaluation and treatment of femoral neck nonunions: the technique,
treatment. JBJS Reviews. 2015;3(3):e1–8. a case series, and literature review. J Orthop Trauma.
11. Nardi A, Ventura L, Cozzi L, Tonini G, Zennaro R, 2006;20(7):485–91.
Celi M, Ramazzina E. The bone anabolic therapy. 29. Hartford JM, Patel A, Powell J. Intertrochanteric
Aging Clin Exp Res. 2013;25(Suppl 1):S121–4. osteotomy using a dynamichip screw for femoral neck
12. Rubin LE, Murgo KT, Ritterman SA, McClure PK. nonunion. J Orthop Trauma. 2005;19(5):329–33.
Hip resection arthroplasty. JBJS Rev. 2014;2(5):e3. 30. Said GZ, Farouk O, Said HG. Valgus intertrochan-
13. Protzman RR, Burkhalter WE. Femoral-neck frac- teric osteotomy with single-angled 130° plate fixation
tures in young adults. J Bone Joint Surg Am. 1976;58 for fractures and non-unions of the femoral neck. Int
(5):689–95. Orthop. 2010;34(8):1291–5.
224 R.A. Hayda, MD
31. LeCroy CM, Rizzo M, Gunneson EE, Urbaniak JR. 47. McKinley JC, Robinson CM. Treatment of displaced
Free vascularized fibular bone grafting in the man- intracapsular hip fractures with total hip arthroplasty:
agement of femoral neck nonunion in patients comparison of primary arthroplasty with early sal-
younger than fifty years. J Orthop Trauma. 2002;16 vage arthroplasty after failed internal fixation. J Bone
(7):464–72. Joint Surg Am. 2002; 84-A(11):2010–5.
32. Jun X, Chang-Qing Z, Kai-Gang Z, Hong-Shuai L, 48. Mabry TM, Prpa B, Haidukewych GJ, Harmsen WS,
Jia-Gen S. Modified free vascularized fibular grafting Berry DJ. Long-term results of total hip arthroplasty
for the treatment of femoral neck nonunion. J Orthop for femoral neck fracture nonunion. J Bone Joint
Trauma. 2010;24(4):230–5. Surg Am. 2004;86-A(10):2263–2267.
33. Elgafy H, Ebraheim NA, Bach HG. Revision internal 49. Ferguson GS, Cabanela ME, Ilstrup DM. Total hip
fixation and nonvascular fibular graft for femoral arthoplasty after failed intertrochanteric osteotomy.
neck nonunion. J Trauma. 2011;70(1):169–73. J Bone Joint Surg Br. 199;76(2): 252–257.
34. Wu CC, Shih CH, Chen WJ, Tai CL. Treatment of 50. Boos N, Krushell R, Ganz R, Müller ME. Total hip
femoral neck nonunions with a sliding compression arthroplasty after previous proximal femoral osteot-
screw: comparison with and without subtrochanteric omy. J Bone Joint Surg Br. 1997;79(2):247–53.
valgus osteotomy. J Trauma. 1999;46(2):312–7. 51. Sandhu HS, Sandhu PS, Kapoor A. Neglected
35. Meyers MH, Harvey JP Jr, Moore TM. The muscle fractured neck of the femura predictive classification
pedicle bone graft in the treatment of displaced and treatment by osteosynthesis. Clin Orthop Relat
fractures of the femoral neck: indications, operative Res. 2005;431:14–20.
technique and results. Orthop Clin North Am. 1974;5 52. Nagi ON, Dhillon MS, Goni VG. Open reduction,
(4):779–92. internal fixation and fibular autografting for
36. Baksi DP. Internal fixation of ununited femoral neck neglected fracture of the femoral neck. J Bone Joint
fractures combined with muscle-pedicle bone graft- Surg Br. 1998;80(5):798–804.
ing. J Bone Joint Surg Br. 1986;68(2):239–45. 53. Huang HC. Treatment of neglected femoral neck
37. Ballmer FT, Ballmer PM, Baumgaertel F, Ganz R, fractures in young adults. Clin Orthop Relat Res.
Mast JW. Pauwels osteotomy for nonunions of the 1986;206(5):119–126.
femoral neck. Orthop Clin North Am. 1990;21 54. Roshan A, Ram S. Early return to function in young
(4):759–67. adults with neglected femoral neck fractures. Clin
38. Hou SM, Hang YS, Liu TK. Ununited femoral neck Orthop Relat Res. 2006;447(6):152–7.
fractures by open reduction and vascularized iliac 55. Lifeso R, Younge D. The neglected hip fracture.
bone graft. Clin Orthop Relat Res. 1993;294:176–80. J Orthop Trauma. 1990;4(3):287–92.
39. Leung PC, Shen WY. Fracture of the femoral neck in 56. Verghese VD, Boopalan PR, Titus V, Oomen VT,
younger adults: a new method of treatment of Jepegnanam TS. Indices affecting outcome following
delayed and nonunions. Clin Orthop Relat Res. neglected femoral neck fracture after valus inter-
1993;295:156–60. trochanteric osteotomy. J Orthop Trauma. 2014;28
40. Cordero-Ampuero J, de Dios M. What are the risk (7):410–6.
factors for infection in hemiarthroplasties and total 57. Glassner PJ, Tejwani NC. Failure of proximal
hip arthroplasties? Clin Orthop Relat Res. 2010;468 femoral locking compression plate: a case series.
(12):3268–77. J Orthop Trauma. 2011;25(2):7683.
41. Franzén H, Nilsson LT, Strömqvist B, Johnsson R, 58. Johnson B, Stevenson B, Chamma R, Patel A,
Herrlin K. Secondary total hip replacement after Rhee S, Lever C, et al. Short-term follow-up of
fractures of the femoral neck. J Bone Joint Surg Br. pertrochanteric fractures treated using the proximal
1990;72(5):784–7. femoral locking plate. J Orthop Trauma. 2014;28
42. Hägglund G, Nordström B, Lidgren L. Total hip (5):283–7.
replacement after nailing failure in femoral neck frac- 59. Haidukewych GJ, Berry DJ. Salvage of failed
tures. Arch OrthopTrauma Surg. 1984;103(2):125–7. internal fixation of intertrochanteric hip fracture.
43. Nilsson LT, Jalovaara P, Franzén H, Niinimäki T, Clin Orthop Relat Res. 2003;412:184–8.
Strömqvist B. Function after primary hemiarthro- 60. Mariani EM, Rand JA. Nonunion of intertrochanteric
plasty and secondary total hip arthroplasty in femoral fractures of the femur following open reduction and
neck fracture. J Arthroplasty. 1994;9(4):369–73. internal fixation: Results of second attempts to gain
44. Tabsh I, Waddell JP, Morton J. Total hip arthroplasty union. Clin Orthop Relat Res. 1987;218:81–9.
for complications of proximal femoral fractures. 61. Sarathy MP, Madhavan P, Ravichandran KM. Non-
J Orthop Trauma. 1997;11(3):166–9. union of intertrochanteric fractures of the femur.
45. Turner A, Wroblewski BM. Charnley low-friction J Bone Joint Surg Br. 1995;77(1):90–2.
arthroplasty for the treatment of hips with late 62. Wu CC, Shih CH, Chen WJ, Tai CL. Treatment of
complications of femoral neck fractures. Clin cutout of a lag screw of a dynamic hip screw in an
Orthop. 1984;185:126–30. intertrochanteric fracture. Arch Orthop Trauma Surg.
46. Haidukewych G, Berry DJ. Salvage of failed treat- 1998;117(4–5):193–6.
ment of hip fractures. J Am Acad Orthop Surg. 63. Bartonícek J, Skála-Rosenbaum J, Dousa P. Valgus
2005;13(2):101–9. intertrochanteric osteotomy for malunion and
nonunion of trochanteric fractures. J Orthop Trauma. 71. Kang S, McAndrew MP, Johnson KD. The recon-
2003;17(6):606–12. struction locked nail for complex fractures of the
64. Haidukewych GJ, Berry DJ. Hip arthroplasty for proximal femur. J Orthop Trauma. 1995;9(6):453–63.
salvage of failed treatment of intertrochanteric hip 72. Pai CH. Dynamic condylar screw for subtrochanteric
fractures. J Bone Joint Surg Am. 2003;85-A(5):899– femur fractures with greater trochanteric extension.
904. J Orthop Trauma. 1996;10(5):317–22.
65. Hammad A, Abdel-aal A, Said HG, Bakr H. Total 73. French BG, Tornetta P 3rd. Use of an interlocked
hip arthroplasty following failure of dynamic hip cephalomedullary nail for subtrochanteric fracture
screw fixation of fractures of the proximal femur. stabilization. Clin Orthop Relat Res. 1998;348:95–
Acta Orthop Belg. 2008;74(6):788–92. 100.
66. Laffosse JM, Molinier F, Tricoire JL, Bonnevialle N, 74. Haidukewych GJ, Berry DJ. Nonunion of fractures of
Chiron P, Puget J. Salvage hip arthroplasty for the subtrochanteric region of the femur. Clin Orthop
trochanteric osteosynthesis failure in the elderly. Relat Res. 2004;419:185–8.
Acta Orthop Belg. 2007;73(6):729–36. 75. Wu C. Locked nailing for shortened subtrochanteric
67. Tabsh I, Waddell JP, Morton J. Total hip arthroplasty nonunions: a one-stage treatment. Clin Orthop Relat
forcomplications of proximal femoral fractures. Res. 2009;467(1):254–9.
J Orthop Trauma. 1997;11(3):166–9. 76. Barquet A, Mayora G, Fregeiro J, López L, Rienzi D,
68. Lundy DW. Subtrochanteric femoral fractures. J Am Francescoli L. The treatment of subtrochanteric
Acad Orthop Surg. 2007;15(11):663–71. nonunions with the long gamma nail twenty-six
69. Kinast C, Bolhofner BR, Mast JW, Ganz R. Sub- patients with a minimum 2-year follow-up. J Orthop
trochanteric fractures of the femur: Results of Trauma. 2004;18(6):346–53.
treatment with the 95 degrees condylar blade-plate. 77. Duffy GP, Wood MB, Rock MG, Sim FH. Vascu-
Clin Orthop Relat Res. 1989;238:122–30. larized free fibular transfer combined with autograft-
70. Wiss DA, Brien WW. Subtrochanteric fractures of ing for the management of fracture nonunions
the femur: results of treatment by interlocking associated with radiation therapy. J Bone Joint Surg
nailing. Clin Orthop Relat Res. 1992;283:231–6. Am. 2000;82(4):544–54.
Femoral Shaft Nonunions
10
Gil R. Ortega, MD, MPH and Brian P. Cunningham MD
niques, improved implants design, and the

10.1 Introduction and Background
development of biologic adjuvants, nonunions
still occur and represent a challenging diagnostic
In 1939, Dr. Küntscher introduced intramedul-
and clinical problem. Femoral nonunion also
lary (IM) nails and subsequently revolutionized
represent a serious socioeconomic challenge
the treatment of long bone fractures [1]. The
resulting in prolonged patient morbidity, gait
transition to intramedullary nails for the treat-
abnormality, multiple operations as well as high
ment of femoral shaft fractures has resulted
resource utilization, emotional impairment and
union rates ranging between 90 and 100% [2–4].
ultimately a prolonged inability to return to work.
These fractures are typically a high-energy injury
A systematic approach to the evaluation and
resulting in significant mortality risks, and they
treatment of femoral nonunions is required for a
are often encountered with other injuries.
successful outcome because of the multifactorial
Improved trauma care has provided decreased
nature of the problem. The origin of the process
mortality for severely injured patients, and it has
requires the diligent investigation of the fracture,
been suggested that this may result in an
fixation strategy previously employed, signs of
increasing number of complex femoral shaft
infection, and critical evaluation of the charac-
fractures and subsequently increase the number
teristics of the nonunion. Once the fracture has
of femoral nonunion [5–8]. Femoral shaft non-
been diagnosed, the nonunion must be charac-
unions can result from a number of etiologies
terized in order to identify the factors that pre-
including inadequate fixation stability, insuffi-
vented fracture union. When the personality of
cient blood supply, bone loss, or infection.
the nonunion has become clear, a surgical plan
Despite the success of treating femoral shaft
can be developed. A large number of tools are
fractures as well as improved surgical tech-
currently available in the orthopedic armamen-
tarium including dynamization, plate osteosyn-
thesis, external fixation, exchange nailing,
G.R. Ortega (&) adjuvant such as electrical or ultrasound stimu-
Department of Orthopaedic Surgery, Mayo Clinic
lation, bone grafting with autogenous or allo-
Arizona Orthopaedic Residency Program, Mayo
Clinic Arizona, HonorHealth Scottsdale Osborn genic bone, and biologic agents.
Trauma Center, Sonoran Orthopaedic Trauma Despite advances in technology and tech-
Surgeons, 3126 N. Civic Center Plaza, Scottsdale, nique, current union rates range from 50 to 100%
AZ 85251, USA
of patients, which is significantly lower than
e-mail: gortega@sotsmd.com
previously suggested [9, 10]. Surgeons must
B.P. Cunningham
counsel patients accurately and make every effort
Department of Orthopaedic Surgery, University of
Minnesota, Regions Hospital, St. Paul, MN 55101, to select the optimal intervention for each unique
USA situation. This chapter will discuss the

228 G.R. Oretega, MD, MPH and B.P. Cunningham, MD
epidemiology of femoral shaft nonunions and 10.3 Etiology

examine briefly the socioeconomic impact, as
well as briefly discuss the etiology of nonunions Various predisposing and contributing factors
with special attention to those factors, which are have been identified that predispose a patient to
unique to femoral shaft injuries. The diagnosis of nonunion. Identification of these factors is a
femoral shaft fractures including history, physi- critical element of the evaluation of femoral shaft
cal examination, and laboratory and imaging nonunions in order to create a surgical plan that
work up will be addressed. Finally, the treatment corrects the offending variables. Female gender
option will be discussed as well as a brief note on and age have been linked to decreased healing
adjuvant treatments including biologic agents. and potential increase in nonunion rate secondary
The last section will contain a few case examples to deceased estrogen levels and biologic activity,
that illustrate some of the challenges faced when respectively [16]. Metabolic and endocrine
treating femoral shaft nonunions. abnormalities are well-established etiologies of
nonunions. A recent series of patients with
unexplained nonunions found 83% of patients
10.2 Epidemiology were found to have a previously undiagnosed
metabolic or endocrine abnormality after being
The annual incidence of midshaft femur fractures evaluated by an endocrinologist; however,
is approximately 10–37 per 100,000 person per interestingly, the rate was 100% in the subset of
years with the incidence peaking among the patients with femoral shaft nonunions under-
young and then again in the elderly [11, 12]. scoring the importance of investigating this
Current estimates place femoral shaft nonunion group of patients [17]. Additionally, cigarette
rates above humeral shaft nonunions, but lower use, medical comorbidities, and NSAIDs use
than tibial shaft nonunions [13]. Extrapolated have all been correlated to and increased risk of
from those statistics using a 1–9% nonunion rate nonunion [18–21].
and the current population in the USA of 313 Fracture-specific characteristics can also
million then each year between 310 and 2790, increase the risk of nonunion in femoral shaft
patients in the USA will go onto a femoral shaft fractures. Disruption of the soft tissue envelope
nonunion [14]. through either an open fracture [22] or open
Fracture nonunions represent a difficult chal- reduction during intramedullary nailing [21] has
lenge for the surgeon, the patient, and thus also been shown to increase the risk of nonunion. The
to the health system and the social services degree of fracture comminution has also been
supporting them. Their average treatment man- shown to increase the risk of nonunion in open
agement requires large assets and long-lasting femur fractures with the assumption that a greater
therapies. Recent estimates have placed treatment degree of periosteal and soft tissue damage has
of femoral nonunions at $28,970 utilizing occurred [23] resulting in impaired physiology at
exchange nailing without biologic adjuvant [13]. the fracture site [24]. The presence of a fracture
This provides a starting point; however, other gap has also been indicated to increase the non-
surgical techniques, infectious etiology, and union risk [25] with the caveat that this variable
reoperation would all drastically increase the has to be taken in context with the fracture type
overall costs. Moreover, it is estimated that the (simple > comminuted) and fixation strategy
indirect costs for musculoskeletal conditions (compression or nail > bridge or external fixator).
represent about 80% of the total costs of these Infection has been well documented to increase
conditions and can be highly variable [15]. the risk of nonunion in all fracture types and is
Utilizing this data, the cost of treating femoral most applicable in femoral shaft fractures after
shaft nonunions ranges from 9 to 80. One million open fractures and prolonged open reductions.
dollars annually with this likely being at the low The surgical techniques utilized also play a
end of the spectrum. role in femoral shaft fracture healing. Although
10 Femoral Shaft Nonunions 229
there are many techniques to stabilize femoral surgeon to create a successful operative plan.
shaft fractures, the literature is most clear in Deformity is often rotational, but can be
addressing femoral shaft fracture healing when varus/valgus or a length discrepancy. Finally, the
these fractures are treated with a femoral nail. For surgeon should have a high degree of suspicion
example, in regard to femoral shaft fracture for femoral shaft nonunions to be infected with
nailing technique, the femoral shaft fracture careful inspection to assess warmth, induration,
union rates are similar with union rates near sinus tracts, fluctuate, and tenderness to
equal at 100% for antegrade nailing and 98% for palpation.
retrograde nailing [26]. According to a system- Radiographic evaluation should include
atic literature review, reamed femoral nailing has orthogonal views of the femur with additional
a higher union rate when compared to unreamed views to establish the mechanical axis of the
femoral nailing [27]. entire extremity (hip to ankle standing, tibia AP,
and Lateral). On occasion, computed tomogra-
phy (CT) is useful to analyzed the presence and
10.4 Diagnosis amount of bridging callus as well as quantifying
rotational deformities. While data is limited, CT
As with all clinical problems, the starting point has been shown to improve diagnostic accuracy
for evaluation is a thorough history. Target areas in tibial nonunions [29].
include a detailed discussion of the original Laboratory evaluation for patients with
injury with notes on the mechanism, time to femoral shaft nonunions has two main purposes.
treatment, postoperative course, and potential The exclusion of infection is a multifactorial
signs of infection such as wound healing issues process; however, preoperative laboratory anal-
or fevers. A concerted effort to gather records ysis should include an erythrocyte sedimentation
from previous surgeries can be extremely valu- rate, C-reactive protein level, and complete blood
able in understanding the original injury and count with differential. While not routinely uti-
fixation strategy. The most important component lized, the literature has suggested some value
of the history is pain with ambulation, with exists utilizing radionuclide and indium 111–la-
careful attention paid to the temporal association, beled leukocyte scans in equivocal cases [30].
character, and duration. Demographic informa- The gold standard to diagnosis infection contin-
tion must also be considered as well as medical ues to be tissue culture and all antibiotics should
comorbidities because both major factors can be discontinued 7–14 days prior to collection
increase the prevalence of nonunions. [31].
The physical exam should focus on four main
components, the soft tissue envelope, the motion
at the fracture site, the deformity, and the signs of 10.5 Classification
infection. Evaluation of the soft tissue envelope
should provide information regarding both the Nonunions are classified according to the criteria
original injury (open vs. closed), previous soft described by Müller [32] and Weber [33]. This
tissue reconstruction, reduction and fixation classification divides nonunions into viable or
strategy (open vs. closed reduction), and a care- nonviable subtypes. Viable nonunions are
ful exam of the vascular status of the extremity. defined as having biologic evidence of healing
Motion at the fracture site will most often result such as bone bleeding, bony soft tissues attach-
in pain with ambulation, and this is uniquely ments, and nonsclerotic bony edges.
sensitive for femoral shaft nonunion. In the Hypertrophic nonunions are well vascularized
developing world, union is often defined by a and show evidence of an attempt to heal the
simple clinical test called “squat and smile” as a fracture site. Radionuclide scans will be hot
surrogate for radiographic union [28]. Evaluation secondary to callus formation. This type of
of the deformity of the extremity enables the nonunion typically results from a fixation
strategy that allows gross motion at the fracture optimize medical comorbidities and nutritional
site. Hypertrophic nonunions are subdivided into: status. In addition to optimizing the host status,
any substance that interferes with osteogenic
1. ‘Elephant foot’ nonunions: These nonunions activity should be stopped (NSAIDs, cigarette
lay down abundant callus. use). A number of treatment options exist and
2. ‘Horse hoof’ nonunions: Moderate callus and will briefly be discussed including dynamization,
possibly signs of bone sclerosis. conversion of a plate to nail, exchange nailing,
3. Oligotrophic nonunions: Osteogenic activity plate augmentation, plate osteosynthesis, external
is minimal at the fracture site, minimal to no fixation, and adjuvant alternatives.
callus formation, and differ from atrophic
subcategory based on intact blood supply.
10.6.1 Dynamization
Atrophic nonunions show little or no evidence
to heal the fracture site typically without any Dynamization is the removal of those interlock-
callus formation. The fracture site forms sclerotic ing screws creating a length unstable nail and
mottled bone ends and biologic activity is mini- theoretically allowing compression at the fracture
mal. These types of nonunions require both an site with weight bearing [34]. Typically, a screw
improved fixation strategy as well as biologic is left in an oblong whole on the dynamized end
augmentation. Radionuclide scans demonstrate of the nail to provide rotational control of the
cold secondary to osteologic ischemia. Atrophic segment. While tibial shaft fractures have seen
nonunions are subdivided into: improved outcomes with this technique [35],
outcomes in femoral shaft nonunions is mixed
1. Torsion wedge nonunions: Fracture fragment [36, 37]. The literature reports a union rate of
with reduced blood supply and no evidence of approximately 50% [38], although numerous
callus formation. complications have been seen with this tech-
2. Comminuted nonunions: typically seen with nique. Shortening greater than 2 cm has been
at least one necrotic, with no evidence of reported in up to 20% of patients underscoring
callus formation. the importance of a careful analysis of fracture
3. Defect nonunions: Segmental bone loss pattern and identification of length stable injuries
makes fracture healing nearly impossible. [34, 36, 37]. These patients require close
May be viable or nonviable at the bone ends. follow-up, and this technique provides the best
4. Atrophic nonunions: Fibrous tissue is formed results with early utilization.
between atrophic bone at the fracture site.
10.6.2 Plate Conversion to a Nail
10.6 Treatment Options Revising a compression or bridge plate to a

reamed intramedullary nail has been reported
The overall goal of treating femoral shaft non- with excellent outcomes. The advantages include
unions is to achieve bony healing, however there open access to the fractures site for correction
are other considerations as well including, but and removal of any fibrous tissue as well as
not limited, to curing infection, correction of biologic activity of the reamings and relative
malalignment, rehabilitating the patient, and stability of the intramedullary nail. Outcomes
return to previous functional level. The treatment have reported as union rates ranging from 91 to
strategy should consider all the goals for the 100% with an average time to union of 6 months
patient and maybe altered secondarily to one [39–42]. The addition of autologous bone graft
component. The patient factors must be accoun- did not improve outcomes further demonstrating
ted for and every effort should be made to the beneficial biologic activity of intramedullary
reamings [42]. This technique has limited utility nailing [44, 48]. Recently, the concept of plate
in the USA where the vast majority of adult augmentation has been introduced for nonunions
femoral shaft fractures are treated primarily with [49, 50]. The benefits of this technique capitalize
intramedullary nails. on the load-sharing properties of the nail while
adding the compression and rotational control
created by addition of the plate as well direct
10.6.3 Exchange Nailing access to the fracture site. The development of
locked plates has simplified the surgeon’s ability
Exchange nailing refers to the technique of to utilize the technique of augmentative plating
removing the previous intramedullary implant, through reliable unicortical screw fixation.
reaming the medullary canal to a larger diameter Outcomes in the literature from augmentative
and implanting a second intramedullary nail with plate fixation for femoral nonunions are limited.
a significantly larger diameter. The benefits of A recent systemic review of femoral shaft non-
this technique include improved biomechanics at unions demonstrates an average union rate of
the fracture site with a larger nail (r4) and 96% with an average time to union of 6 months
improved endosteal contact through isthmic [47, 50, 51]. Challenges with this technique are
elongation, augmentative biologic activity of the chiefly the disruption of the soft tissue envelope
intramedullary reamings, preservation of the soft and the inability to correct any pre-existing
tissue envelope, and early weight bearing. The deformity.
amount of reaming deposited at the fractures site
has been speculated to be limited, and therefore,
a role for open bone grafting in some situations 10.6.5 Plate Fixation
may exist [43].
Outcomes from exchange nailing have been Weber and Čech advocate debridement with the
mixed with union rates reported from 53 to 100% use of plates for “mechanical rest and massive
[44–46]. A recent systematic review demon- cancellous autograft” [52]. Muller and Rosen
strated 343 patients treated with exchange nailing first described the use of the plate compression
in 11 studies averaged a union rate of 73% (n = principle in the treatment of femoral nonunions
251) at an average of 7 months [47]. Efforts [32, 53]. Plate osteosynthesis provides advan-
should be made to increase the diameter of the tages over other techniques, specifically com-
new nail by 2 mm typically necessitating ream- pression and direct access to the nonunion site
ing the canal 1.5 mm larger than the anticipated where bone grafting and/or biologic agents can
implant. While exchange nailing is clearly not also be supplemented to the plate fixation. In
the previously described panacea for all femoral these situations, the surgeon should give con-
shaft nonunions, it does provide some clear sideration to plate osteosynthesis. As previously
advantages to more invasive treatments. This discussed, plating can be particularly helpful in
technique is recommended for viable nonunion metadiaphyseal nonunions of the femur where
in particular because adequate biologic activity direct endosteal contact cannot be attained. Plate
will lead to union with improved biomechanical fixation has also been shown to have improved
environment. biomechanical profile compared to a retrograde
nail with regard to axial and torsional stability
[54]. Without question, an infected nonunion can
10.6.4 Augmentative Plate Fixation be amenable to debridement, open reduction, and
compression with plate fixation and has been
Femoral shaft nonunions that result from fracture reported with successful outcomes in a
patterns with extensive comminution and large single-staged procedure [55].
segmental defects or metaphyseal–diaphyseal Outcomes from plate fixation have historically
location have had poor outcomes with exchange been excellent. In one large retrospective series
of femoral shaft nonunions utilizing a wave plate, 10.6.7 Masquelet Technique

union rates were 98% after a single procedure
[56]. A more recent prospective series utilized a The clinical management of segmental bone
subvastus approach and indirect reduction tech- defects in femoral shaft nonunions is a particu-
niques with a AO 95° blade plate and reported larly challenging situation. Current treatment
91% union rate at 3 months [57]. While excellent options include the use of autografts for defects
outcomes have been achieved with plate <5 cm; however, techniques for larger defects
osteosynthesis, there are some considerable require treatments with high morbidity such as
downsides. Compared with closed medullary vascularized bone grafting or intercalary bone
nailing, plate osteosynthesis has a higher risk for transport [66–69]. Masquelet and colleagues [70,
infection, greater blood loss, and further devas- 71] reported successful repair of wide diaphyseal
cularization to soft tissues [58, 59]. A further defects >25 cm with concurrent severe soft tissue
challenge with plate fixation is patient compli- loss in patients utilizing a technique of induced
ance with restricted postoperative weight bearing membrane and autologous cancellous bone
which can also lead to decreased functional grafting in a two-step surgical procedure. This
rehabilitation and stiffness postoperatively [60]. technique was initially described in 1986 for the
reconstruction of extensive diaphyseal bone loss
up to 25 cm in length with an associated severe
10.6.6 External Fixation soft tissue injury [70]. This technique allows the
reconstruction of large diaphyseal defects even in
The circular external fixators (Ilizarov) have also the face of irradiated or infected recipient site
been used for the treatment of femoral nonunions [72]. The first stage consists of complete surgical
While small case series have been reported with debridement of all necrotic tissue followed by the
good results in aseptic nonunions [61], the prin- insertion of a polymethylmethacrylate cement
cipal role of this type of fixation is primarily for spacer into the defect, which leads to a mild for-
infected nonunions [62, 63]. Cyclic compression eign body reaction and the formation of a mem-
and distraction has been retrospectively investi- brane. After the soft tissue has stabilized and there
gated using Ilizarov techniques. Union was are no signs of infection, the second stage
achieved in an average of 5.8 months; however, involves removing the spacer and bone grafting.
the authors noted that pain control and This stage can be preformed at a variable time
pin-related complications were a major limitation from 2 to 6 months after the first stage. Careful
of this technique [64]. Advantages of the Ilizarov incision of the membrane and removal of the
technique include percutaneous application, cement spacer is then followed by bone grafting.
minimal blood loss, correction of the deformity Masquelet initially described cancellous autograft
in three planes, and leg length discrepancies, for bone grafting; however, many strategies have
while allowing the patient early weight bearing. been advocated including intramedullary ream-
Disadvantages of this technique include emo- ings, allograft, and combinations of the previ-
tional challenges for psychological impaired ously listed sources [73]. The graft is carefully
patients, pin tract infections, and risks of neu- placed into the membrane and the membrane is
rovascular injury at the time of wire insertion. then sutured closed to create a confined com-
The technical challenge and economic costs partment. The application of this technique has
estimated at $50,607 per patient relegates the been reported for femoral shaft nonunions with a
utility of this technique primarily to tertiary union rate of 88% in patients with segmental
referral centers [65]. defects ranging from 1 to 25 cm [74].
10.7 Biologic Augmentation 10.8 Summary
10.7.1 Adjuvant Treatments Femoral shaft nonunions are a debilitating clini-

cal problem for patients as well as a diagnostic
In addition to nonunion fixation strategies, ad- and technical challenge for orthopedic surgeons.
juvant therapy can promote union and improve Patients are burdened with gait abnormality,
biologic activity of nonviable nonunions. Adju- inability to return to work, re-operations, and
vant modalities include electrical stimulation, psychological impairment; while care for these
bone grafting and bone graft substitutes, and patients take a tremendous amount of resources
bone morphogenic proteins. and burdens hospitals systems with tremendous
While these modalities can be employed in costs. A systematic approach to the evaluation
isolation, they are almost exclusively utilized to and treatment of femoral nonunions is required
complement a previously described technique. for a successful outcome because of the com-
Electrical stimulation in isolation has shown plexity of the problem.
improved union rates; however, the data is Diligent investigation of the original fracture,
limited in femoral shaft fractures [75, 76]. previous fixation strategy, and critical evaluation
Autogenous bone, allograft bone, bone marrow of the characteristics of the nonunion are para-
aspirate, and BMPs, independently or in com- mount to develop a successful treatment algo-
bination to nonunion sites, can improve the rithm. At this time, there is no golden standard
biology of the local environment. These adju- for femoral nonunions and each patient must be
vants are typically utilized in nonviable non- approached as a unique situation requiring a
unions or in cases with bone loss in unique solution.
combination with a revised fixation strategy.
While isolated bone grafting has been reported
in the literature, this technique is rarely utilized 10.9 Case Discussions
at our institution [77] and has not shown a
benefit when converting fixation from a plate to
Case 1
intramedullary nail for femoral shaft nonunions
59-year-old female that suffered an initial
[78]. BMP has been studied in both animal
Type IIIA open segmental femur fracture that was
models and clinical trials, and while the indi-
treated with a variety of techniques including
cations remain controversial, it has been evalu-
initial external fixation, Masquelet technique,
ated in the treatment of femoral shaft nonunions.
plating and bone grafting, intramedullary nail
A retrospective review of 30 patients who had
fixation, and bone grafting. Femur healed after 12
atrophic aseptic femoral shaft nonunions and
months of treatments that required intramedullary
reported union rate of 87% at median period of
nail fixation and bone grafting as final treatment
6 months [79].
(Figs. 10.1, 10.2, 10.3, 10.4, 10.5 and 10.6).
Fig. 10.1 Initial injury X-rays (a–c)
Fig. 10.2 Post-initial debridement and external fixation (a–c)
Fig. 10.3 Initial Masquelet technique after debridements (a, b)

Fig. 10.4 Massive bone

grafting using the RIA or
reamer-irrigator-aspirator
(DePuy Synthes, Warsaw IN,
USA) and bone
morphogenetic protein-2 (off
label) (Medtronic,
Memphis TN, USA) (a, b)
Fig. 10.5 Femur shaft

nonunion at 8 months
postbone grafting (a, b)
Fig. 10.6 Femur shaft union

after removal of plate,
intramedullary nailing, and
combination of autograft and
allograft (a–d)
Case 2 compression plating, and bone grafting. Femur

18-year-old male with femoral shaft nonunion healed after 10 months of treatments that
treated with a variety of techniques including required compression plating, intramedullary nail
initial intramedullary nail fixation, exchange exchange, and bone grafting as final treatment
intramedullary nailing and bone grafting, (Figs. 10.7, 10.8, 10.9, 10.10 and 10.11).
Fig. 10.7 Initial femur

postoperative X-rays at 3
months with nonunion (a, b)
Fig. 10.8 Femur X-rays after

exchange nailing and bone
grafting (a, b)
Fig. 10.9 Femur X-ray after nail dynamization
Fig. 10.10 Immediate

postoperative femur X-rays
after compression plating in
addition to previous exchange
nailing and bone grafting with
nail dynamization (a, b)
Fig. 10.11 Femur shaft

union after compression
plating in addition to previous
exchange nailing and bone
grafting with nail
dynamization (a–d)
4. Wolinsky PR, McCarty E, Shyr Y, Johnson K.

References Reamed intramedullary nailing of the femur: 551
cases. J Trauma. 1999;46(3):392–9.
1. Küntscher G, Maatz R. Technik der marknagelung. 5. Babhutkar S, Pande K. Nonunion of the diaphysis of
Leipzig: Georg Thieme Verlag; 1945. [Book in long bones. Clin Orthop Relat Res. 2005;431:50–6.
German]. 6. Finkemeier CG, Chapman MW. Treatment of
2. Canadian Orthopaedic Trauma Society. Nonunion femoral diaphyseal nonunions. Clin Orthop Relat
following intramedullary nailing of the femur with Res. 2002;398:223–34.
and without reaming: results of a multicenter 7. Furlong AJ, Giannoudis PV, DeBoer P, Matthews SJ,
randomized clinical trial. J Bone Joint Surg Am. MacDonald DA, Smith RM. Exchange nailing for
2003;85-A(11):2093–6. femoral shaft aseptic non-union. Injury. 1999;30
3. Winquist RA, Hansen ST Jr, Clawson DK. Closed (4):245–9.
intramedullary nailing of femoral fractures: a report 8. Giannoudis PV, Matthews SJ, Smith RN. Exchange
of five hundred and twenty cases. J Bone Joint Surg nailing for femoral nonunion. J Orthop Trauma.
Am. 1984;66(4):529–39. 2000;14(7):522–3.
9. Kempf I, Grosse A, Rigaut P. The treatment of 24. Triffin P, Greg PJ. Depression of bone blood flow
noninfected pseudarthrosis of the femur and tibia after brunt trauma. Acta Orthop Scand. 1994;65
with locked intramedullary nailing. Clin Orthop (2):195–8.
Relat Res. 1986;212:142–54. 25. Pihlajamäki HK, Salminen ST, Böstman OM. The
10. Webb LX, Winquist RA, Hansen ST. Intramedullary treatment of nonunions following intramedullary
nailing and reaming for delayed union or nonunion of nailing of femoral shaft fractures. J Orthop Trauma.
the femoral shaft: a report of 105 consecutive cases. 2002;16(6):394–402.
Clin Orthop Relat Res. 1986;212:133–41. 26. Ostrum RF, Agarwal A, Lakatos R, Poka A.
11. Weiss RJ, Montgomery SM, Al Dabbagh Z, Prospective comparison of retrograde and antegrade
Jansson KA. National data of 6409 Swedish inpa- femoral intramedullary nailing. J Orthop Trauma.
tients with femoral shaft fractures: stable incidence 2000;14(7):496–501.
between 1998 and 2004. Injury. 2009;40(3):304. 27. Duan X, Li T, Mohammed AQ, Xiang Z. Reamed
12. Arneson TJ, Melton LJ 3rd, Lewallen DG, O’Fallon intramedullary nailing versus unreamed intramedul-
WM. Epidemiology of diaphyseal and distal femoral lary nailing for shaft fracture of femur: a systematic
fractures in Rochester, Minnesota, 1965–1984. Clin literature review. Arch Orthop Trauma Surg.
Orthop Relat Res. 1988;234:188–94. 2011;131(10):1445–52.
13. Kanakaris N, Giannoudis PV. The health economics 28. Zirkle LG Jr. Diversity drives design. American
of the treatment of long-bone non-unions. Injury, Academy of Orthopaedic Surgeons. AAOS Now.
Int J Care Injured. 2007;38(Suppl 2):S77–84. 2012;6(8):47. http://www.aaos.org/news/aaosnow/
14. Norris BL, Nowotarski PJ. Femoral shaft fractures. aug12/youraaos7.asp. Accessed 12 Feb 2016.
In: Stannard JP, Schmidt AH, editors. Kregor PJ, 29. Bhattacharyya T, Bouchard KA, Phadke A,
editors. Surgical treatment of orthopedic trauma. Meigs JB, Kassarjian A, Salamipour H. The accuracy
New York: Thieme; 2007. p. 611–32. of computed tomography for the diagnosis of tibial
15. Bozic KJ, Rosenberg AG, Huckman RS, Herndon JH. nonunion. J Bone Joint Surg Am. 2006;88(4):692–7.
Economic evaluation in orthopaedics. J Bone Joint 30. Nepola JV, Seabold JE, Marsh JL, Kirchner PT,
Surg Am. 2003;85A(1):129–42. el-Khoury GY. Diagnosis of infection in ununited
16. Calori GM, Albisetti W, Agus A, Iori S, Tagliabue L. fractures: combined imaging with
Risk factors contributing to fracture non-unions. indium-111-labeled leukocytes and technetium-99
Injury. 2007;38(Suppl 2):S11–8. m methylene diphosphonate. J Bone Joint Surg
17. Brinker MR, O’Connor DP, Monla YT, Earth- Am. 1993;75(12):1816–22.
man TP. Metabolic and endocrine abnormalities in 31. Gristina AG, Naylor PT, Webb LX. Molecular
patients with nonunions. J Orthop Trauma. 2007;21 mechanisms in musculoskeletal sepsis: the race for
(8):557–70. the surface. Instr Course Lect. 1990;39:471–82.
18. McKee MD, DiPasquale DJ, Wild LM, Stephen DJ, 32. Müller ME. Treatment of nonunions by compression.
Kreder HJ, Schemitsch EH. The effect of smoking on Clin Orthop Relat Res. 1965;43:83–92.
clinical outcome and complication rates following 33. Weber BG, Čech O. Pseudoarthrosis: pathology,
Ilizarov reconstruction. J Orthop Trauma. 2003;17 biomechanics, therapy, results. Berne, Switzerland:
(10):663–7. Hans Huber Medical Publisher; 1976.
19. Schmitz MA, Finnegan M, Natarajan R, 34. Yokota H, Tanaka SM. Osteogenic potentials with
Champine J. Effect of smoking on tibial shaft fracture joint-loading modality. J Bone Miner Metab.
healing. Clin Orthop Relat Res. 1999;365:184–200. 2005;23:302–8.
20. Giannoudis PV, MacDonald DA, Matthews SJ, 35. Richardson JB, Gardner TN, Hardy JR. Dynamiza-
Smith RM, Furlong AJ, De Boer P. Nonunion of tion of tibial fractures. J Bone Joint Surg Br. 1995;77
the femoral diaphysis: the influence of reaming and (3):412–6.
non-steroidal anti-inflammatory drugs. J Bone Joint 36. Wu CC, Lee ZL. Low success rate of
Surg Br. 2000;82(5):655–8. non-intervention after breakage of interlocking nails.
21. Malik MH, Harwood P, Diggle P, Khan SA. Factors Int Orthop. 2005;29(2):105–8.
affecting rates of infection and nonunion in intrame- 37. Pihlajamaki HK, Salminen ST, Bostman OM. The
dullary nailing. J Bone Joint Surg Br. 2004;86 treatment of non-unions following intramedullary
(4):556–60. nailing of femoral shaft fractures. J Orthop Trauma.
22. Arslan H, Subaşý M, Kesemenli C, Ersuz H. Occur- 2002;16(6):394–402.
rence and treatment of nonunion in long bone 38. Wu CC, Chen WJ. Healing of 56 segmental femoral
fractures in children. Arch Orthop Trauma Surg. shaft fractures after locked nailing: poor results of
2002;122(9–10):494–8. dynamization. Acta Orthop Scand. 1997;68(6):537–
23. Noumi T, Yokoyama K, Ohtsuka H, Nakamura K, 40.
Itoman M. Intramedullary nailing for open fractures 39. Wu CC, Shih CH, Chen WJ. Effect of reaming bone
of the femoral shaft: evaluation of contributing grafting on treating femoral shaft aseptic non-union
factors on deep infection and nonunion using mul- after plating. Arch Orthop Trauma Surg. 1999;119
tivariate analysis. Injury. 2005;36(9):1085–93. (5–6):303–7.
40. Wu CC. Treatment of femoral shaft aseptic utilizing internal fixation. Injury. 2009;40(11):1220–
non-union associated with plating failure: emphasis 5.
on the situation of screw breakage. J Trauma. 56. Ring D, Jupiter JB, Sanders RA. Complex non-union
2001;51(4):710–3. of fractures of the femoral shaft treated by wave-plate
41. Megas P, Syggelos SA, Kontakis G. Intramedullary osteosynthesis. J Bone Joint Surg Br. 1997;79
nailing for the treatment of aseptic femoral shaft (2):289–94.
non-unions after plating failure: effectiveness and 57. Bellabarba C, Ricci WM, Bolhofner BR. Results of
timing. Injury. 2009;40(7):732–7. indirect reduction and plating of femoral shaft
42. Emara KM, Allam MF. Intramedullary fixation of nonunions after intramedullary nailing. J Orthop
failed plated femoral diaphyseal fractures: are bone Trauma. 2001;15(4):254–63.
grafts necessary? J Trauma. 2008;65(3):692–7. 58. Cove JA, Lhowe DW, Jupiter JB, Siliski JM. The
43. Furlong AJ, Giannoudis PV, DeBoer P. Exchange management of femoral diaphyseal nonunions.
nailing for femoral shaft aseptic non-union. Injury. J Orthop Trauma. 1997;11(7):513–20.
1999;30(4):245–9. 59. Rozbruch SR, Müller U, Gautier E, Ganz R. The
44. Weresh MJ, Hakanson R, Stover MD, Sims SH, evolution of femoral shaft plating technique. Clin
Kellam JF, Bosse MJ. Failure of exchange reamed Orthop Relat Res. 1998;354:195–208.
intramedullary nails for ununited femoral shaft 60. Abdel-Aa AM, Farouk OA, Elsayed A, Said HG.
fractures. J Orthop Trauma. 2000;14(5):335–8. The use of a locked plate in the treatment of ununited
45. Hak DJ, Lee SS, Goulet JA. Success of exchange femoral shaft fractures. J Trauma. 2004;57(4):832–6.
reamed intramedullary nailing for femoral shaft 61. Brinker MR, O’Connor DP. Ilizarov compression
nonunion or delayed union. J Orthop Trauma. over a nail for aseptic femoral nonunions that have
2000;14(5):178–82. failed exchange nailing: a report of five cases.
46. Yu CW, Wu CC, Chen WJ. Aseptic nonunion of a J Orthop Trauma. 2003;17(10):668–76.
femoral shaft treated using exchange nailing. Chang 62. Krishnan A, Pamecha C, Patwa JJ. Modified Ilizarov
Gung Med J. 2002;25(9):591–8. technique for infected nonunion of the femur: the
47. Somford MP, van den Bekerom MP, Kloen P. Oper- principle of distraction–compression osteogenesis.
ative treatment for femoral shaft nonunions, a J Orthop Surg. 2006;14(3):265–72.
systematic review of the literature. Stat Traum Limb 63. Saridis A, Panagiotopoulos E, Tyllianakis M,
Recon. 2013;8:77–88. Matzaroglou C, Vandoros N, Lambiris E. The use
48. Banaszkiewicz PA, Sabboubeh A, McLeod I. of the Ilizarov method as a salvage procedure in
Femoral exchange nailing for aseptic non-union: infected nonunion of the distal femur with bone loss.
not the end to all problems. Injury. 2003;34(5):349– J Bone Joint Surg Br. 2006;88(2):232–7.
56. 64. Inan M, Karaogtu S, Cirri F, Turk CY, Harma A.
49. Nadkarni B, Srivastav S, Mittal V. Use of locking Treatment of femoral nonunions by using cyclic
compression plates for long bone non-unions without compression and distraction. Clin Orthop Relat Res.
removing existing intramedullary nail: review of 2005;436:222–8.
literature and our experience. J Trauma. 2008;65 65. Patil S, Montgomery R. Management of complex
(2):482–6. tibial and femoral nonunion using the Ilizarov
50. Hakeos WM, Richards JE, Obremskey WT. Plate technique, and its cost implications. J Bone Joint
fixation of femoral nonunions over an intramedullary Surg Br. 2006;88(7):928–32.
nail with autogenous bone grafting. J Orthop 66. de Boer HH, Wood MB. Bone changes in the
Trauma. 2011;25(2):84–9. vascularized fibular graft. J Bone Joint Surg Br.
51. Chen CM, Su YP, Hung SH, Lin CL, Chiu FY. 1989;71(3):374–8.
Dynamic compression plate and cancellous bone 67. Han CS, Wood MB, Bishop AT, Cooney WP 3rd.
graft for aseptic non-union after intramedullary Vascularized bone transfer. J Bone Joint Surg Am.
nailing of femoral fracture. Orthopedics. 2010;33 1992;74(10):1441–9.
(6):393. 68. Pelissier P, Casoli V, Demiri E, Martin D,
52. Weber BG, Čech O. Pseudarthrosen. Bern: Hans Baudet J. Soleus-fibula free transfer in lower limb
Huber; 1973. [Book in German]. reconstruction. Plast Reconstr Surg. 2000;105
53. Rosen H. Compression treatment of long bone (2):567–73.
pseudarthroses. Clin Orthop Relat Res. 69. Green S, Jackson J, Wall D, Marinow H, Ishka-
1979;138:154–66. nian J. Management of segmental defects by the
54. Meyer RW, Plaxton NA, Postak PD, Gilmore A, Ilizarov intercalary bone transport method. Clin
Froimson MI, Greenwald AS. Mechanical compar- Orthop. 1992;280:136–42.
ison of a distal femoral side plate and a retrograde 70. Masquelet AC. Muscle reconstruction in reconstruc-
intramedullary nail. J Orthop Trauma. 2000;14 tive surgery: soft tissue repair and long bone
(6):398–404. reconstruction. Langenbecks Arch Surg. 2003;388
55. Prasarn ML1, Ahn J, Achor T, Matuszewski P, (5):344–6.
Lorich DG, Helfet DL. Management of infected 71. Masquelet AC, Fitoussi F, Begue T,
femoral nonunions with a single-staged protocol Muller GP. Reconstruction of the long bones by the
induced membrane and spongy autograft. Ann Chir interim results of a double-blind trial. Lancet. 1984;1
Plast Esthet. 2000; 45(3):346–53. [Article in French]. (8384):994–6.
72. Pelissier P, Martin D, Baudet J, Lepreux S, 76. Sharrard WJ. A double-blind trial of pulsed electro-
Masquelet AC. Behavior of cancellous bone graft magnetic fields for delayed union of tibial fractures.
placed in induced membranes. Br J Plast Surg. J Bone Joint Surg Br. 1990;72(3):347–55.
2002;55(7):598–600. 77. Wu CC, Chen WJ. Treatment of femoral shaft aseptic
73. Giannoudis PV, Faour O, Goff T, Kanakaris N, nonunions: comparison between closed and open
Dimitriou R. Masquelet technique for the treatment bone-grafting techniques. J Trauma. 1997;43(1):112–6.
of bone defects: tips-tricks and future directions. 78. Emara KM, Allam MF. Intramedullary fixation of
Injury. 2011;42(6):591–8. failed plated femoral diaphyseal fractures: are bone
74. Stafford PR, Norris BL. Reamer-irrigator-aspirator grafts necessary? J Trauma. 2008;65(3):692–7.
bone graft and Masquelet technique for segmental 79. Kanakaris NK, Lasanianos N, Calori GM, Ver-
bone defect nonunions: a review of 25 cases. Injury. donk R, Blokhuis TJ, Cherubino P. Application of
2010;41(Suppl 2):S72–7. bone morphogenetic proteins to femoral non-unions:
75. Barker AT, Dixon RA, Sharrard WJ, Sutcliffe ML. a 4-year multicenter experience. Injury. 2009;40
Pulsed magnetic field therapy for tibial non-union: (Suppl. 3):S54–61.
Distal Femoral Nonunions
11
Animesh Agarwal, MD
intra-articular component is a simple split, direct

11.1 Introduction
reduction of the joint with screw fixation and
subsequent retrograde nailing can be performed.
11.1.1 Distal Femur Fractures
In all cases, it is important to realign the
mechanical axis of the limb.
Distal femur fractures are defined as those frac-
Open reduction and internal fixation of the
tures involving the distal 9–15 cm of the femur.
distal femur historically required fixed angle
They may be entirely extra-articular (AO/OTA
devices [4]. The 95° dynamic condylar screw
Type A), partial articular (AO/OTA Type B) or
(DCS) and 95° blade plate provided excellent
intra-articular (AO/OTA Type C) [1].
fixation options for all extra-articular distal femur
The management of these fractures depends
fractures and select type C fractures with simple
on the type. The type A fractures are generally
intra-articular splits. More comminuted articular
best managed with either intramedullary
fractures required other options such as
(IM) nailing or open reduction and internal fix-
non-locking condylar plates which often times
ation (ORIF) [2]. The decision usually is based
required a second plate medially to provide suf-
on the amount of intact distal femur. If there is
ficient support. These gave way to locking plates,
less than 4 cm, generally ORIF is preferred to
with one of the earliest being the less invasive
obtain adequate fixation in the distal segment. If
stabilization system (LISS) [4]. Early reports of
above 4 cm, then there usually is enough of the
its use in the management of distal femur frac-
distal femur to allow for two interlocking screws
tures were encouraging [5, 6]. Weight and Col-
with a retrograde nail [3]. Antegrade nailing can
linge reported a 100% healing rate at 13 weeks
be used but meticulous detail must be adhered to,
[5]. In a larger series of 103 fractures, Kregor
to insure an anatomical restoration of the limb
et al. [6] reported a healing rate of 93% without
[2]. Type B fractures require ORIF with screws
adjunctive bone grafting. They had two
± small plates as needed since they are partial
nonunions.
articular fractures. The type C fractures generally
are best managed with ORIF with locking large
fragment plates to allow fixed angle fixation or
dedicated fixed angle devices; if the
11.1.2 Incidence of Nonunions
The incidence of nonunions in the management

of distal femoral fractures has been reported
A. Agarwal (&) historically anywhere from 0 to 19% in the lit-
Division of Orthopedic Traumatology, Department erature with variation depending upon the
of Orthopedics, University of Texas Health Science
Center, San Antonio, TX, USA implant used [4]. This was prior to the wide-
e-mail: Agarwal@uthscsa.edu spread use of current locked condylar plating

244 A. Agarwal, MD
systems. As locking plate constructs became 11.2.1 Mechanical Considerations

more widely used, the rate of nonunion seems to
have increased with an incidence as high as 32% Locked plating has been found to be a potential
[7–9]. Additionally, as the population increases, risk factor for the development of nonunion.
the number of geriatric distal femur fractures has Healing problems with the use of locked plates
increased with a nonunion incidence of 24% in have recently been reported as high as 32% [7–9].
this subgroup [10]. Hoffman et al. [7] reported an 18% nonunion rate
with locked plating and a 20% recalcitrant non-
union rate after secondary procedures in this
11.1.3 Ramifications of Nonunions subgroup. Henderson et al. [8] retrospectively
evaluated a group of 86 distal femur fractures (82
These distal femoral nonunions can be severely patients) treated with locked plating and found a
disabling and lead to poor function [11]. Many of nonunion incidence of 20% which was much
these patients have been unable to bear weight higher than the reported literature at the time.
through the affected limb for months if not years. They felt that callus inhibition was occurring from
These patients also have malalignment that too stiff of a construct. A similar decrease in callus
affects the mechanical axis of the lower extrem- formation when using locking plates for distal
ity. They often have a leg length discrepancy due femur fractures was seen in comparison to intra-
to bone loss from the injury, which is then medullary nails [13]. Lujan et al. [14] retrospec-
exacerbated by the multiple procedures that have tively evaluated 64 consecutive patients that
been performed. However, with proper attention underwent osteosynthesis of a distal femur frac-
to the principles of nonunion management, repair ture with either titanium plates or stainless steel
of the distal femoral nonunion can lead to healing plates. They found that locking constructs did
and improved function [12]. result in asymmetric callus formation, which was
inconsistent. Most notably they found that tita-
nium constructs exhibited significantly more cal-
11.2 Causes of Nonunions lus formation early on up to 12 weeks. Although
no increased risk of nonunion with the use of
Chapter 1 reviewed the various risk factors for stainless steel implants was seen in this study,
the development of nonunions. There are, others have indicated a potential relation-
however, factors that have been implicated in ship. Rodriguez et al. [15], in a multicenter ret-
nonunion development that are specific to the rospective study, showed that the use of stainless
distal femur. The majority of nonunions are steel plate was an independent risk factor for
related to the mechanical environment that is nonunion. They showed that the probability of
created by the fixation construct. Fortunately, intervention for a nonunion was 21% if a stainless
the femur has a circumferential soft tissue steel plate was used initially versus 4% if a tita-
envelope. However, damage to the soft tissues nium plate had been used at the index operation. In
can occur when the fracture is open. Addition- a follow-up study, Rodriguez et al. [16] showed a
ally, periosteal stripping and bone loss can 41% nonunion rate in stainless steel constructs,
occur in such high-energy injuries. This can but only 10% nonunion rate in titanium constructs
provide a significant biological insult and put which was statistically significant. However, the
the patient at risk for the development of a overall nonunion rate was 13.3% indicating that
nonunion as well. The surgeon can inadver- most of the cases in their series were treated with
tently cause additional stripping during fixation. titanium plates (239 T vs. 32 SS). They indicated
Thus, it is important to minimize soft tissue that the plate material was an independent pre-
dissection, especially on the medial side. dictor of nonunion development.
11 Distal Femoral Nonunions 245
Historically, fixed angle devices, such as the screws were used in a prospective observational
dynamic condylar screw, were the implants of study by Bottlang et al. [21]. Thirty-two patients
choice for distal femur fractures. The device had with 33 distal femur fractures were treated with
excellent results with 0–10% nonunion rates [4]. this technique, with thirty-one available for
In a multicenter study by the Canadian Ortho- follow-up. They had a statistically significant
pedic Trauma Society, the use of the DCS was increase in periosteal callus at weeks 12 and 24
revisited [17]. In a prospective randomized con- when compared to their previous published his-
trolled trial, the DCS group had a union rate of toric controls [14]. There was only one nonunion
91% compared to 52% of the LISS group at in this series. Despite the promising results, this
12 months. There was a higher complication and technique has not gained wide acceptance.
revision rate in the LISS group. A similar implant It is clear that too stiff of a construct can
to the DCS, the 95° angled blade plate was inhibit fracture healing, but conversely inade-
compared to the locking condylar plate (LCP) quate stability can lead to failure of the construct
[18]. They showed that the LCP group had more and a subsequent nonunion. It has been sug-
malunions and nonunions, and a statistically gested that by using a longer plate (>9 holes in
higher incidence of complications resulting in shaft length but with at least 8 holes proximal to
significantly more secondary procedures. In a the fracture) can minimize plate failure [20]. In a
study comparing the use of the LISS plate to that different study, plate length was not found to be
of locking compression plates (LCP), similar predictor of nonunion [16]. Most likely it is a
nonunion rates were seen—22.1% (LISS) and combination of plate length, the number of
20.7% (LCP) [19]. screws, plate material and the type of screws
In an effort to determine the cause of the placed that modulate the healing response.
healing issues, Bottlang et al. [20] evaluated the
biomechanics of locked plating in the distal
femur. They showed that biomechanically there 11.2.2 Biological Considerations
clearly was asymmetric interfragmentary motion,
with the least amount of motion at the cortex An open fracture with resultant bone loss or
adjacent to the plate. Clinically, this was seen as defect can be a predisposing factor to the
inhibition of callus formation and nonunion development of a nonunion [15, 19, 22]. An open
development of 19% despite intact hardware. In fracture was found to increase the probability of
an attempt to promote callus formation and an intervention for nonunion from 21 to 52% for
enhance the mechanical environment with locked a stainless steel plate and from 4 to 14% for a
plating, they described a technique called far titanium plate [15]. In a separate study, 37% of
cortical locking. In this technique, the screws open fractures required reoperation for a non-
locked into the plate but did not engage the near union, compared to only 10% of closed fractures
cortex and only obtained fixation in the far cor- [22].
tex. Biomechanically, this reduced the overall An open distal femur fracture is the result of a
stiffness of the construct and promoted inter- high-energy injury with resultant comminution,
fragmentary motion. In an animal model, they which itself has been suggestive of nonunion
compared far cortical locking to standard locking [11]. It has been suggested that bone grafting of
in a tibia gap model. The far cortical locking these highly comminuted injuries should be
group had significantly greater callus formation, considered early to help prevent failure of the
which was symmetric, had stronger calluses and hardware and subsequent nonunion [23]. Barei
complete healing. In a biomechanical cadaveric and Beingessner [24] bone grafted 55% of distal
distal femur model, the far cortical locking femurs with bone loss in their series at an aver-
technique showed an 81% decrease in construct age of 70 days, all of which achieved union.
stiffness and enables parallel interfragmentary Those with bone loss, which did not undergo
motion. Specially designed far cortical locking bone grafting, all healed. All of these were found
246 A. Agarwal, MD
to have posterior cortical continuity. The pres- et al. [22] found that proximal implant failure
ence of posterior cortical continuity despite bone was associated with a higher BMI. The implant
loss indicated that bone grafting was unneces- failure then leads to the development of a
sary. These open fractures also have significant nonunion.
soft tissue disruption, and thus, further insult
with extensive exposures can further disrupt the
already compromised soft tissue envelope. 11.3 Evaluation and Diagnosis
Infection has been reported in 0–10% after
ORIF. Many of the same things leading to a The general evaluation and diagnosis of non-
nonunion can predispose one to infection. unions has been covered in Chap. 1. The same
Infection itself has been shown to be a risk factor principles apply. However, specific points to
for the development of a nonunion [11, 15, 19]. address in relation to distal femoral nonunions
Infection was found to increase the probability of will be discussed below.
an intervention for nonunion from 21 to 66% for
a stainless steel plate and from 4 to 24% for a
titanium plate [15]. Thus, it is imperative, when 11.3.1 History
evaluating a nonunion, that infection is ruled out
with the appropriate laboratory studies and A clear understanding of the original mechanism
radiographic imaging (see Chap. 1). of injury can provide information to assist in
evaluating the nonunion. It is important to
understand the mechanism of injury of the orig-
11.2.3 Patient Considerations inal fracture. Was it a high- or low-energy
injury? Was it an open fracture? If it was open,
Patients with osteoporosis may have tenuous how many surgeries prior to definitive fixation?
fixation and are at risk for hardware failure. What was done at the time of the original sur-
Locked plating can certainly be helpful in such gery? This can be hard to ascertain if the patient
cases. Despite this, the geriatric population has has undergone several surgeries prior to their
high incidence of nonunion despite the use of presentation. Requesting the medical records
locked plates. Moloney et al. [10] performed a from the original surgeon can be enlightening.
multicenter retrospective cohort study of 176 Did they have any problems after fixation?
patients. The mortality at one year was 25% with Obtaining an accurate history regarding any
a 24% incidence of nonunion in the survivors. previous infection is paramount. It is important to
The long-term functional outcome of such inju- determine when weight bearing began especially
ries in the geriatric population has been very poor when there is hardware failure. Early failure may
as well [25]. indicate non-compliance with the postoperative
Non-compliance with weight bearing may put regimen. A social history should be obtained to
undue stress on plate fixation and can lead to include the use of nicotine, narcotics and illicit
early failure. Smoking can certainly delay frac- drugs. A careful medical history, to determine
ture healing and may lead to a nonunion [22]. whether any comorbidities contributed to the
Diabetes and other endocrinopathies can also development of the nonunion especially diabetes,
lead to a delay in healing or a nonunion [22]. is critical.
Morbid obesity has also been shown to be a
risk factor for the development of a nonunion
specifically in distal femur fractures [15, 22]. 11.3.2 Physical Examination
Obesity was found to more than double the
probability of an intervention for a nonunion The patient should be evaluated for gross motion
despite the material of the implant [15]. Ricci at the nonunion site in cases of hardware failure.
The limb should be inspected for signs of 11.3.5 Computed

infection such as erythema or draining sinus Tomography/Magnetic
tracts. Knee motion should be assessed as best Resonance Imaging
possible. In cases where the hardware has failed,
the patient may have too much discomfort or A computed tomography (CT) scan should be
pain to assess accurate range of motion. A thor- obtained to define the nonunion. If there is con-
ough neurovascular exam should be performed. cern for malrotation, a CT scan of both hips and
Many of these patients may have concomitant knees can be obtained to compare the injured
ligamentous injuries of the knee, which may side to the unaffected side for a more accurate
have gone unrecognized. Therefore, a careful determination. A magnetic resonance image
knee exam to assess for stability should be per- (MRI) is warranted in select cases where the
formed if possible. Gross motion at the nonunion hardware has already been removed (no metal
site may preclude an accurate assessment of knee artifact) and in infected cases to better assess the
stability. The patient should be evaluated for leg presence and extent of osteomyelitis. In general,
length discrepancy, as many of these patients will an MRI is not needed for the aseptic nonunion.
have developed shortening from the numerous
previous surgeries [26].
11.3.6 Nuclear Imaging
11.3.3 Laboratories These studies can be useful in evaluating non-

unions when there is a concern for infection. If
This has been covered previously, but a full laboratory studies (CBC, ESR and CRP) are
evaluation for infection (complete blood count elevated, then nuclear medicine studies may add
[CBC], erythrocyte sedimentation rate [ESR] and additional information. In the case of aseptic
C-reactive protein [CRP]) and metabolic issues nonunions, these studies are usually not indicated.
should be performed. Vitamin D deficiency
should be addressed. Endocrinopathies and other
metabolic abnormalities may require evaluation 11.4 Treatment
by an endocrinologist. Diabetics should have
better glucose control. Osteoporosis should be In a systematic review of the literature regarding
managed with appropriate medications. distal femoral nonunions, the most common
treatment involved fixed angle plating with can-
cellous autografting resulting in a 97.4% union
11.3.4 Radiographs rate [11]. It is important to determine whether
there are any causative factors which may have
Standard anteroposterior (AP) and lateral contributed to the nonunion. Correctable factors
radiographs of the entire femur should be should be addressed such as smoking cessation
obtained. Standing bilateral AP and laterals and vitamin D replacement (which should correct
from the hip to the ankle can help to assess for secondary hyperparathyroidism). Treatment is
any associated deformity with the nonunion. based on a number of factors. The type of non-
This also allows evaluation of the mechanical union, whether it is hypertrophic or atrophic, will
axis of the limb and to rule out any other determine whether bone grafting is needed. The
associated deformities in the tibia. Stress presence of intact or failed hardware can influ-
examination of the nonunion site can be ence the treatment of choice. For fractures that
obtained to determine whether any motion is were intra-articular, a determination of whether
present in the cases of stiff nonunions where or not the intra-articular portion has healed can
clinical evaluation may be equivocal. determine whether revision ORIF is needed or
248 A. Agarwal, MD
whether the nonunion is isolated to the 92.5% healing rate with the one surgery in the 27
meta-diaphyseal region. It is clear from the lit- patients with follow-up.
erature that no clear consensus exists as to the An alternative approach to complete revision
best treatment option for these nonunions [11, ORIF has recently been described by Holzman
12]. et al. [32], where a medial locking plate is
Chapman et al. [27] used either single or added to a preexisting intact lateral locking
double plate fixation with autologous bone plate construct. They treated 22 patients with 23
grafting in the management of distal femoral distal femoral nonunions with either the addi-
nonunions in 18 patients. In their retrospective tion of a medial plate and autogenous bone
review, they had 100% union rate. In another grafting when the lateral plate was stable (16
study by Bellabarba et al. [28], twenty patients cases) or a two-stage procedure where the bro-
with nonunions were managed with indirect ken lateral plate was removed and replaced,
reduction techniques and application of either a followed by a medial locking plate and bone
95° condylar blade plate, condylar buttress plate graft two months after the first stage (7 cases).
or a locking condylar plate. Only 45% (atrophic They had a 95.2% success rate in the 21 cases
and oligotrophic nonunions) underwent adjunc- with follow-up. They concluded that adding a
tive autologous bone grafting. They reported a medial plate in cases with stable lateral fixation
100% union rate. All of these patients had been was a successful alternative to complete revision
initially treated with similar plate screw con- surgery.
structs, but none had bone grafting as part of the The use of intramedullary nailing in the
original fracture treatment. The same authors had management of distal femoral nonunion has also
used similar techniques in a series of been studied. However, the early supracondylar
twenty-three patients with femoral nonunions nails initially developed were fraught with com-
that had been initially treated with intramedullary plications due to the multiple hole configurations
nailing for their femur fracture [29]. There were of these implants. Koval looked at a series of 16
only eight distal femoral nonunions. These were distal femoral nonunions treated with the supra-
all treated with a 95° condylar blade plate and all condylar nail and had only a 25% success rate
healed. The overall success rate for all fractures with a high rate of hardware failure and com-
was 91%. Bone grafting was performed on all plications [33]. Wu treated 21 distal femoral
biologically deficient nonunions. Gardner et al. nonunions with antegrade nails placed in a ret-
[12] reviewed a single surgeon case series of 31 rograde fashion and dynamically locked [26]. In
distal femoral nonunions treated with a fixed the 18 patients followed for an average of
angled implant. Lag screws across the nonunion 3.3 years, 88.9% healed at an average of
site were used in all patients as well as bone graft 4.2 months. All were bone grafted with autoge-
augmentation (71% autologous bone). They had nous bone graft obtained from the ipsilateral
a 97% union rate at 15.9 weeks with return to a medial tibial condyle at the time of the nailing. In
pre-injury functional status in 84% of patients. a similar series, Wu also treated 13 distal femoral
Deformity correction was an important part of nonunions where the initial fracture was treated
the treatment. Wang and Weng [30] treated with an antegrade nail [3]. They again utilized an
thirteen patients with distal femoral nonunions antegrade nail placed in retrograde fashion,
with open reduction and internal fixation com- locked dynamically with bone grafting from the
bined with both cortical allograft struts and medial tibial condyle. Plate fixation was added in
autogenous iliac crest bone grafts. They used some cases. They had a 100% union rate at an
predominately blade plates or condylar buttress average of 4.5 months.
plates and a few antegrade nails. They achieved Since many of these nonunions have associ-
100% union at an average of 5 months. Amorosa ated leg length discrepancy and deformity,
et al. [31] used 95° angled blade plates to treat 32 external fixation has been described as an option
cases of distal femoral nonunions. They had a for the management of distal femoral nonunions.
Ali and Saleh [34] treated 15 cases of distal 11.4.1 Treatment Based on Nonunion
femoral nonunion in which all had either a leg Type
length discrepancy or malalignment requiring
correction. Five of the cases were infected. They 11.4.1.1 Hypertrophic
had success in 14 of 15 (93.3%) cases with the Hypertrophic nonunions need stability and thus
one case uniting after intramedullary nailing. improvement of the mechanical environment is
They were able to correct angular deformities as paramount. These do not usually require bone
well as regain length in these patients. The big- grafting. In the majority of cases, the hardware
gest issue was poor motion with an average range has failed and revision of the fixation is required.
of motion of 80o. Either plate fixation or retrograde intramedullary
As a salvage procedure in patients with a nailing has been successful. If a retrograde
persistent nonunion of the distal femur, espe- intramedullary nail is utilized, the intra-articular
cially in the elderly, prosthetic replacement has component must be healed. The nail can address
been described [35–37]. Haidukewych et al. [35] the meta-diaphyseal component only. If the
performed a total knee arthroplasty (TKA) in 17 intra-articular nonunion is simple, lag screw
patients (ages 38–86; mean of 66) that had either compression with bone graft may be needed. The
failed treatment of a distal femur fracture or mechanical axis needs to be realigned regardless
nonunion. They had a five-year survivorship of of the implant used. The hypertrophic nonunion
91%. They did have a 29% rate of both is usually mobile enough to allow for deformity
intra-operative and postoperative complications. correction. If a nail is used, blocking screws can
They felt that it provided reliable pain relief as aid in deformity correction. If plates are used,
well as functional improvement, but the overall fixed angle devices can help correct the defor-
results were inferior to that of primary TKA. In mity. Small leg length discrepancies can be tol-
cases of the elderly patient with a persistent erated and managed with a shoe lift. Healing of
nonunion, the use of a megaprosthesis has also the hypertrophic nonunion is the goal.
been reported [36, 37]. These patients are cited as
having poor bone quality, arthritis, joint con- 11.4.1.2 Atrophic or Oligotrophic
tractures and previous implant failure. Revision The decision for bone grafting is clear and should
is felt to be a poor option in these elderly be performed in cases of atrophic or oligotrophic
patients. Vaishya et al. [37] treated ten patients nonunions. If the hardware is stable, autogenous
with a persistent nonunion and arthritis with a bone grafting can be performed without a need
megaprosthesis. All knees had satisfactory for hardware revision. If, however, the hardware
alignment and range of motion, but two patients has failed, then both revision fixation and bone
had minor wound problems. They felt that this grafting are required for a successful outcome.
was a viable one-stage salvage procedure for the As in all cases, the mechanical axis should be
patient with a difficult nonunion. The advantage re-established. If the joint component is healed,
for prosthetic replacement is that is allows for then retrograde nailing with use of the ‘reamer–
early ambulation [36]. irrigator–aspirator’ (Synthes, Paoli PA, USA) for
Many options exist for the treatment of distal harvesting of autogenous bone graft from the
femoral nonunions, and there is no clear algo- femoral canal can be performed. Fixation with a
rithm for the best treatment in terms of implant. retrograde nail that has fixed angle capabilities in
Revision plating, intramedullary nailing and the distal segment should be utilized. The
even circular external fixation are all viable reamings obtained can be packed into the non-
options but need to be based on the stability of union site.
the pre-existing fixation as well as the local
biology. Prosthetic replacement should be con- 11.4.1.3 Infected
sidered in the elderly with poor bone quality In cases of infected nonunions, a two-stage if not
and arthritis. three-stage procedure may be warranted. In the
250 A. Agarwal, MD
first stage, removal of the hardware, debridement of the nonunion type. Nail stabilization
of the infected nonunion site, obtaining cultures, allows for earlier weight bearing. Bone
application of antibiotic cement into the defect grafting provides a biological stimulant as
with or without temporary external fixation are many of these patients have already had
performed. Once the infection is cleared, stabi- several operations at the time of presentation.
lization along with placement of a cement spacer Correction of any deformity can usually be
is performed. Fixation can be with either a ret- accomplished with the nail as most are
rograde intramedullary nail (preferred) or a mobile. If the nonunion site is stiff, a fixed
locking plate or a fixed angle device. In the final angle device (95° angled blade plate or DCS)
stage, bone grafting into the defect is done after to correct the deformity may be a better
the cement spacer is removed (Masquelet tech- option as long as the joint injury is healed.
nique). If the amount of bone requiring 4. Failed Hardware with Nonunion of Intra-
debridement is extensive, the use of circular Articular Component: If the hardware has
external fixation and distraction osteogenesis to failed and the joint component is not healed,
fill the defect can be considered. This technique then complete removal of the previous plate
is highly specialized and should be undertaken and revision ORIF with restoration of the
by someone experienced. joint congruity, realignment of the mechani-
cal axis and bone grafting is needed. We
prefer to use a locking plate, either the lock-
11.4.2 Author’s Preferred Methods ing compression LISS plate or locking
of Treatment condylar plates after compression and fixation
of the joint component.
1. Stable Hardware (Rare) and Hypertrophic: 5. Use of External Fixation: Circular external
Adjunctive plate fixation can often provide fixation for nonunion management is reserved
sufficient stability to promote union. This for those cases where, despite bone grafting to
situation is rare. large defects, the nonunion persists. It can also
2. Stable Hardware and Atrophic/Oligotrophic: be useful in cases of septic nonunion where
If the hardware is stable, simply bone grafting internal fixation may be problematic despite
the nonunion site should be sufficient to debridement. It can also be used in cases of
promote union. The harvest site for the bone multiplanar deformities in combination with a
graft should be based upon the amount of nonunion, especially when it is a stiff hyper-
bone graft needed. trophic nonunion. The patient must understand
3. Failed Hardware (Common): In cases where the procedure and the length of time such a
the initial lateral locked plating has failed, the device will be on as it can be life altering
joint component has healed and the during the time the fixator is on the thigh.
meta-diaphyseal area has gone on to a mobile 6. Prosthetic Replacement: In cases of the
hypertrophic nonunion, removal of hardware elderly patient with a distal femoral nonunion,
and retrograde intramedullary nailing with a consideration to a total knee arthroplasty
nail allowing for a fixed angle distally works should be given. The ideal candidate should
well. The largest diameter nail should be used be one with poor bone quality where fixation
to obtain stability. With the advent of the may be problematic with ORIF. If they also
reamer–irrigator–aspirator (RIA) system have preexisting arthritis or as a result of the
(DePuy Synthes, Warsaw IN, USA), it is easy original injury, then a total knee arthroplasty
to obtain autogenous bone graft from the may be preferred. A megaprosthesis (distal
intramedullary canal at the time of reaming. femoral replacing) can be considered when
The bone graft can be packed into the non- the bone stock is deficient and unable to
union. This technique is our method of choice support a standard or stemmed total knee
for most distal femoral nonunions regardless arthroplasty.
11.4.3 Case Discussions Dual-energy X-ray absorptiometry (DEXA) scan

had been obtained by his primary care provider
and was normal. He had normal 25-OH vitamin
Case 1
D levels. His physical examination showed
Patient is a 50-year-old white male originally
well-healed surgical scars as well as traumatic
involved in motor vehicle accident (MVA) in
lacerations from the original injury, varus
2008. Patient sustained a right Grade III A open
malalignment of the limb at the nonunion site
distal femur fracture AO Type C3. He underwent
and flexion only to 30°. A CT scan with coronal
irrigation and debridement of the open fracture
and sagittal reconstructions (Fig. 11.2) was
(I&D) and temporary bridging external fixation.
obtained, which showed healing of the
He then required several washouts due to the
intra-articular component but a clear nonunion of
contamination. He subsequently underwent
the metaphyseal portion with varus collapse with
definitive ORIF approximately 2 weeks after the
pullout of screws as well as broken screws.
initial injury with a 7-hole LCP-LISS plate.
The patient underwent hardware removal with
Patient was followed by the original surgeon and
debridement of all fibrous tissue from the non-
then referred for a nonunion, with hardware
union site. The RIA system was utilized in a
failure at 5 months with AP and lateral radio-
retrograde fashion to obtain bone graft from the
graphs shown in Fig. 11.1.
femoral canal of the affected leg. A retrograde
The patient was evaluated and found only to
nail with a fixed angle blade component distally
have hepatitis C. The patient denied any history
was inserted and statically locked proximally
of wound problems or infections after the
with two screws. The RIA bone graft was packed
definitive procedure. The patient had not smoked
into the nonunion. The postoperative images are
for 30 years and quit drinking 10 years prior to
shown in Fig. 11.3.
presentation. Laboratory markers were all within
The patient went onto heal the nonunion by
normal limits for his white blood cell
7 months (Fig. 11.4). At this point, he underwent
(WBC) count, C-reactive protein (CRP) and
manipulation under anesthesia of his right knee,
erythrocyte sedimentation rate (ESR).
Fig. 11.1 a Anteroposterior

and b lateral radiographs of
the right knee showing
hardware failure, shortening
and varus
252 A. Agarwal, MD
Fig. 11.2 a Axial computed tomography (CT) image b coronal CT image showing the varus alignment, failure
showing healing of the intra-articular component; the gap of hardware and the metaphyseal nonunion; c sagittal
between the plate and the bone is well visualized; image also showing the nonunion
Fig. 11.3 Immediate postoperative images after retro- anatomical axis; b, c anteroposterior and lateral of the
grade intramedullar nailing and bone grafting. a Full right knee showing the nonunion site with bone graft
length right femur showing re-establishment of femoral
quadricepsplasty and an arthrotomy with lysis of arthritic-like symptoms and probably a degener-
adhesions for persistent poor knee motion (0° to ative medial meniscal tear (Fig. 11.5). He was
65°). The patient eventually achieved 110° of also having hardware symptoms distally at the
motion. lateral aspect of the knee. Arthroscopic debride-
The patient did well and returned to his ment along with hardware removal was dis-
activities, which included downhill skiing. cussed with the patient since the patient was
Patient returned 7 years later with complaints of going under anesthesia. The patient had arthro-
knee pain, which was felt to be consistent with scopic debridement of the knee. He was found to

the right knee at 7 months
showing consolidation of the
nonunion site
have Grade III medial tibial compartment disease laboratory studies were within normal limits. The
but only Grade I lateral compartment disease. nuclear medicine studies obtained were negative.
The nail was removed without difficulty Clinically, she did not have any evidence of
(Fig. 11.6). Patient returned to his snow skiing infection nor did she report ever having any
and has improved motion to 120° of flexion and wound problems or any other issues after the
has always maintained his extension. index procedure until 22 months later when she
noticed the sudden pain. A CT scan was obtained
Case 2 and confirmed the nonunion and hardware fail-
The patient is a 54-year-old Latin American ure. The joint was healed (Fig. 11.8).
female who sustained multiple injuries in an The patient underwent removal of the hard-
MVA in 2006. The patient was treated for a left ware, RIA of the femur for bone graft and
distal femur fracture with ORIF at an outside placement of a retrograde nail with a fixed angle
institution. The patient was followed for blade component distally. It was statically locked
approximately 17 months, after which she was proximally with two screws. Her postoperative
told she was healed and discharged. She appar- images are shown in Fig. 11.9.
ently was fully weight bearing. The patient was allowed to be immediately
She then presented 2 years out from the initial weight bearing and went on to heal by 6 months.
injury with hardware failure and a nonunion of (Figure 11.10). At her last follow-up of
the left distal femur (Fig. 11.7). The patient was 13 months, she was ambulating fully with the
unable to give details of the injury as to whether use of a cane for long distances. She was
or not it was an open fracture. The patient is pain-free with 0° to 95° of motion (Fig. 11.11).
morbidly obese. She has diabetes, hypertension
and a history of deep vein thrombosis. Her lab- Case 3
oratory evaluation showed a normal WBC but an The patient is a 38-year-old white male who was
elevated ESR of 74 and CRP of 21.3. Her other initially injured in an MVC while working out of
254 A. Agarwal, MD

the right femur at 7 years
showing well-healed femur
with stable hardware
town. He had sustained a left Grade IIIA open followed and felt to be progressively healing
distal femur fracture/dislocation. His operative (Fig. 11.13; 6 months). He was fully weight
report indicated that both the anterior cruciate bearing, but at 9 months he developed increased
ligament (ACL) and posterior cruciate ligament pain. The radiographs showed subsidence of the
(PCL) were out, but no indication regarding the hardware and some collapse (Fig. 11.14). The
status of his collaterals. He had an initial irriga- patient underwent a CT scan (Fig. 11.15),
tion and debridement with application of a tem- which showed a persistent nonunion of the
porary bridging external fixator. He subsequently metaphyseal area as well as part of the
underwent ORIF at the outside institution. Patient intra-articular region. The allograft bone had not
returned to the area and presented to our institu- been incorporated.
tion approximately 6 weeks out (Fig. 11.12). The patient underwent revision ORIF as
The patient is otherwise healthy. His physical opposed to nailing because of concern for a
examination at that time showed well-healed persistent intra-articular nonunion. The hardware
incisions and traumatic lacerations. He was was removed, and the allograft bone was

the right femur at 3 months
after hardware removal
nonviable and had not incorporated; it was deb- graft placed into the nonunion site with addi-
rided, resulting in the large void shown in tional bone graft extender (demineralized bone
Fig. 11.16. The intra-articular nonunion was matrix [DBM]) (Fig. 11.19). The final postoper-
stabilized with a screw (Fig. 11.17). Bone graft ative radiographs are shown in Fig. 11.20.
was obtained via the RIA system from the left Patient went onto heal the nonunion with
femur after the hardware was removed. It was abundant bone around the site and was func-
done retrograde through the nonunion site tioning well at 18 months. His range of motion
(Fig. 11.18). Revision ORIF with a variable was 0° to 115°. His last follow-up radiographs
angle locked plate was performed and the bone are shown in Fig. 11.21.
256 A. Agarwal, MD
Fig. 11.7 a Anteroposterior and b lateral radiographs of easily visualized; b the break in the plate is well
the left knee 2 years after the initial fixation showing visualized as well as the recurvatum deformity at the
loosening of hardware and nonunion. a The loose screw is nonunion site
Fig. 11.8 Computed tomography scan images showing the nonunion: a axial image showing lack of bone, b coronal
image showing the varus and nonunion, c sagittal image showing the recurvatum deformity and nonunion
Fig. 11.9 Immediate a anteroposterior and b lateral postoperative left femur radiographs showing stabilization of the
nonunion with a retrograde nail and bone grafting
Fig. 11.10 a Anteroposterior and b lateral left femur radiographs at 6 months showing consolidation across the
nonunion site
Case 4 radiographs showed bending of the plate and

The patient is a 51-year-old morbidly obese loosening of the screws distally. There was an
woman who is referred for a nonunion of her obvious nonunion of the meta-diaphyseal region
right distal femur. She is approximately one year (Fig. 11.22).
out from her initial injury, which was a right She reports no immediate complications after
grade III A open distal femur fracture. She was her initial surgery and denies any history of
managed with ORIF at an outside institution. The infection. Her only medical problem is morbid
258 A. Agarwal, MD
Fig. 11.11 Final follow-up a anteroposterior and b lateral left femur radiographs at 13 months showing a well-healed
femur without hardware complications
Fig. 11.12 a Anteroposterior and b lateral left knee radiographs at 6 weeks after open reduction internal fixation. The
fracture appears well reduced and restoration of the anatomical axis
Fig. 11.13 a Anteroposterior and b lateral left knee metaphyseal region, but some subsidence of the plate is
radiographs at 6 months. a There appears to be some seen with collapse at the fracture site but stable hardware;
consolidation at the medial cortex as well as in the b lateral shows increasing consolidation anteriorly
Fig. 11.14 a Anteroposterior and b lateral left knee radiographs at 9 months. a There appears to be further subsidence
of the plate and thus collapse at the fracture site; b lateral shows increasing consolidation posteriorly and intact plate
260 A. Agarwal, MD
Fig. 11.15 A computed tomography scan was obtained to b coronal image showing again the allograft bone and its
evaluate the fracture site. a Axial image showing the lack of incorporation as well as subtle varus collapse; c the
allograft bone still unincorporated and a lack of bridging; obvious nonunion is clearly visualized on the sagittal image
Fig. 11.16 Intra-operative fluoroscopic image after plate Fig. 11.17 Intra-operative fluoroscopic image showing
removal and debridement of the allograft showing the the additional partially threaded cancellous screw for lag
large void screw fixation of the articular nonunion
obesity (BMI 64). She is a smoker (half pack per evaluation showed her CRP to be 19.5, WBC 11.9
day). Smoking cessation was recommended. She and ESR of 22. Her 25-OH vitamin D was less than
had been previously prescribed an ultrasound unit 15. She was immediately started on vitamin D2 at
in an attempt to aid consolidation. Physical 50,000 units weekly. She responded to the dosing
examination showed well-healed surgical scars with her 25-OH vitamin D increasing to 39. Her
and lacerations without signs of infection. Her nuclear medicine studies showed uptake consis-
range of motion was 0° to 100° compared to 0° to tent with degenerative changes in the knee joint
120° on her unaffected side. Her laboratory but no evidence of infection.
managed with a shoe insert, and had regained her

full knee range of motion (ROM) (Fig. 11.25).
Case 5
The patient is a 48-year-old Latin American
female who presents with pain, discomfort and
inability to bear weight on her right lower
extremity. Radiographs obtained at the time
show a right distal femoral nonunion with hard-
ware failure (Fig. 11.26).
Her original injury was 4 years prior at which
time she was treated at an outside facility for a
right grade IIIA open distal femur fracture. She
Fig. 11.18 Intra-operative fluoroscopic image showing
reports having multiple surgeries (10) afterward
the reamer for the reamer–irrigator–aspirator (RIA) going for various reasons, including infection. She is
retrograde through the mobile nonunion site morbidly obese and has hypothyroidism (on
thyroid replacement). She does not smoke. Her
physical examination showed well-healed surgi-
cal incisions and lacerations. Her range of motion
was 5° to 35°, and it appeared that she had about
30° to 40° of malrotation. She had no signs of
infection.
Nonunion evaluation was performed.
Her WBC was 9.6, ESR 29 and a CRP 19.9. She
also had hypovitaminosis D. Nuclear medicine
studies were performed which showed positive
bone scan, indium scan but discordant uptake on
sulfur colloid scan, indicating a concern for
infection (Fig. 11.27). The CT scan showed an
obvious nonunion (Fig. 11.28).
Surgical options were discussed with the
Fig. 11.19 Intra-operative fluoroscopic image after sta- patient, including staging the definitive man-
bilization and bone grafting of the nonunion site
agement, if there was presence of an infection.
At the time of surgery, the nonunion site was
The patient underwent repair of her nonunion evaluated after all the hardware was removed.
with removal of all hardware with obvious The native bone appeared normal. There was a
motion seen at the nonunion site. The nonunion significant amount of allograft ‘croutons’ that
was stabilized with a retrograde nail after were loose and thus were debrided from the
obtaining bone graft using the RIA system. We nonunion site. Intra-operative cultures were sent
obtained 40 cc of bone graft, which was all as well as a stat Gram’s stain, which was neg-
placed into the nonunion site and supplemented ative for bacteria and only 2–3 polymorphonu-
by demineralized bone matrix. The immediate clear cells (PMN) per high-power field
postoperative images are shown in Fig. 11.23. (HPF) on frozen section of the tissue. There was
The patient was followed and felt to have no purulence, just the unincorporated bone
healed by 6 months with bridging bone graft. The decision was made to proceed with
(Fig. 11.24). At her last follow-up at four years, definitive management with the benign appear-
she was ambulating without assistive devices, ance of the nonunion site and the negative
had only a 5 mm leg length discrepancy intra-operative studies. She underwent
262 A. Agarwal, MD

postoperative
a anteroposterior and b lateral
left knee images showing the
final construct
correction of her deformity through the mobile cane on occasion. She reported only occasional
nonunion site, both angulation and rotation. discomfort with weather changes (Fig. 11.31).
There was a 50% circumferential defect for a
length of about 10-cm. Retrograde nailing was Case 6
performed after obtaining autogenous bone graft The patient is a 33-year-old Latin American
using the RIA system. Reaming to 16 mm was female who was involved in a head-on MVA and
performed, and a 15 mm diameter nail was sustained multiple bilateral lower extremity
placed. The defect was packed first around the injuries, including a left grade IIIA open distal
nail with calcium sulfate beads impregnated femur fracture with intra-articular involvement.
with vancomycin (off-label use) after which the In addition to damage control management of her
autograft was packed on top. The entire defect other injuries, she underwent I&D and temporary
was filled. Her knee was then manipulated after bridging external fixation across the left knee
closure. We were able to passively fully extend (Fig. 11.32). The patient returned to the ICU and
her and flex her to 95° (Fig. 11.29). her condition improved.
The patient went on to heal by 8 months as The patient underwent definitive ORIF of her
seen in Fig. 11.30. At her three-year follow-up, left distal femur fracture once she was stable.
she maintained her 95° of flexion, had a slight leg There was extensive comminution and bone loss
length discrepancy of 1.5 cm managed with a in the meta-diaphyseal region extending into the
shoe lift, and was ambulating with the use of a trochlear region. The patient also had calcium
Fig. 11.21 Follow-up

radiographs at 18 months after
the revision open reduction
internal fixation and bone
grafting of the left knee,
showing consolidation of the
nonunion site
Fig. 11.22 a Anteroposterior and b lateral radiographs of the right femur showing the bending of the original fixation
and loss of fixation distally. The large soft tissue density can also be appreciated
264 A. Agarwal, MD
Fig. 11.23 Immediate postoperative a anteroposterior and b lateral radiographs of the right femur showing
stabilization of the nonunion site with a retrograde nail and placement of the bone graft
Fig. 11.24 a Anteroposterior and b lateral radiographs of the right femur at 6 months showing increased consolidation
and bridging of the nonunion. Abundant bone formation is visualized within the marked areas
Fig. 11.25 Last follow-up a anteroposterior and b lateral radiographs of the right femur at 4 years showing resolution
of the nonunion and stable hardware. There has been further consolidation across the nonunion site
Fig. 11.26 a Anteroposterior and b lateral radiographs of the right femur showing failed hardware with significant
varus deformity, hardware failure and nonunion
266 A. Agarwal, MD
Fig. 11.27 Nuclear

medicine studies: a bone scan
showing increased uptake in
the entire distal half of the
right femur (circled);
b subtraction image of sulfur
colloid from indium showing
areas with increased activity
indicating discordant uptake
and suggestive of infection
Fig. 11.28 Computed

tomography scan images
showing the lack of bridging
bone and obvious nonunion;
a axial image; b coronal
image
sulfate impregnated with vancomycin and sup- sulfate slowly resorbed with some consolidation.
plemented by DBM, placed into this large defect At seven months, she was ambulating with a
(Fig. 11.33) cane but with some discomfort (Fig. 11.34). Due
The patient was followed and went on to heal to concern over incomplete healing, a CT scan
all her other fractures, which included a left tibial was obtained (Fig. 11.35). It revealed a large
plafond fracture and right patella fracture. Her anteromedial defect with healing of the lateral
femur continued to progress, and the calcium cortex only. The posterior cortex appeared to
Fig. 11.29 Immediate postoperative a anteroposterior calcium sulfate beads with vancomycin (off-label use) and
and b lateral right femur images showing correction of the the bone graft; b the layering of the bone graft on top of
deformity as well as stabilization of the nonunion with a the calcium sulfate is better delineated
retrograde nail. a The nonunion site is packed with the
Fig. 11.30 a Anteroposterior and b lateral radiographs of the right femur at 8 months showing complete bridging of all
4 cortices and stable hardware
268 A. Agarwal, MD
Fig. 11.31 Three-year

follow-up a anteroposterior
and b and lateral radiographs
of the right femur showing
continued stable hardware and
further consolidation of the
nonunion site
have some bridging. The intra-articular compo- She did have limited ROM to only 90° of
nent had healed completely. It was felt to be a flexion. Repair of the nonunion was discussed
meta-diaphyseal nonunion. and she underwent surgery. Multiple options
The patient never had any problems post- were discussed with the patient to include just
operatively in terms of infection and never autogenous bone grafting. She did not want
showed any signs of infection. All of her harvesting of bone from any other site. It was
laboratory studies were within normal limits. decided to remove the plate and screws and
Fig. 11.32 Initial

radiographs of the left knee
after irrigation and
debridement and bridging
external fixation

postoperative
images of the left femur after
open reduction internal
fixation and placement of
calcium sulfate bead with
vancomycin (off-label use)
and demineralized bone
matrix (DBM). A stainless
steel locking condylar plate
(LCP)—less invasive
stabilization system (LISS)
plate was used
270 A. Agarwal, MD
Fig. 11.34 At 7 months,

radiographs of the left femur
show complete absorption of
the calcium sulfate and
bridging laterally. The
hardware is stable
place a retrograde nail during which the RIA knee was manipulated and full flexion was
system would be used to obtain bone graft, obtained. The patient went on to heal by
which could then be placed into the defect 4.5 months (Fig. 11.37). At her last follow-up
(Fig. 11.36). An open lysis of adhesions was (13 months out from her nonunion repair), she
performed while the hardware was removed. had full ROM and was ambulating without
After the nail and bone graft was placed, the assistive devices (Fig. 11.38).
Fig. 11.35 Computed tomography scan images: a axial lateral cortex; c sagittal image shows the lack of bridging
image shows the lateral bridging but the central nonunion; bone anterior or posterior with central defect
b coronal image shows the defect centrally but the healed

postoperative
images of the left femur after
removal of hardware and
placement of a retrograde
intramedullary nail with the
bone graft
272 A. Agarwal, MD
Fig. 11.37 At 4.5 months, the a anteroposterior and b lateral images of the left femur showed complete bridging of the
nonunion site. It was felt that the patient was healed
Fig. 11.38 At 13 months, a anteroposterior of the left femur and close-up; b, c anteroposterior and lateral images of
the left knee showed solid consolidation of the nonunion site
References callus formation. J Orthop Trauma. 2010;24(3):156–

62.
15. Rodriguez EK, Boulton C, Weaver MJ, Herder LM,
1. Gangavali AK, Nwachuku CO. Management of Morgan JH, Chacko AT, et al. Predictive factors of
distal femur fractures in adults: an overview of distal femoral fracture nonunion after lateral locked
options. Orthop Clin N Am. 2016;47(1):85–96. plating: a retrospective multicenter case-control study
2. Ehlinger M, Ducrot G, Adam P, Bonnomet F. Distal of 283 fractures. Injury. 2014;45(3):554–9.
femur fractures. Surgical techniques and a review of 16. Rodriguez EK, Zurakowski D, Herder L, Hall A,
the literature. Orthop Traumatol. 2013;99(3):353–60. Walley KC, Weaver MJ, Appleton PT, Vrahas M.
3. Wu CC. Retrograde dynamic locked nailing for Mechanical construct characteristics predisposing to
aseptic nonunion of femoral supracondyle after non-union after locked lateral plating of distal femur
antegrade locked nailing. Arch Orthop Trauma Surg. fractures. J Orthop Trauma. 2016;30(8):403–8.
2011;131(4):513–7. 17. Canadian Orthopaedic Trauma Society. Are locking
4. Forster MC, Komarsamy B, Davison JN. Distal constructs in distal femoral fractures always best? A
femoral fractures: A review of fixation methods. prospective multicenter randomized controlled trial
Injury. 2006;37(2):97–108. comparing the less invasive stabilization system with
5. Weight M, Collinge C. Early results of the less the minimally invasive dynamic condylar screw
invasive stabilization system for mechanically unsta- system. J Orthop Trauma. 2016;30(1):e1–6.
ble fractures of the distal femur (AO/OTA types A2, 18. Vallier HA, Immler W. Comparison of the 95-degree
A3, C2, and C3). J Orthop Trauma. 2004;18(8): angled blade plate and the locking condylar plate for
503–8. the treatment of distal femoral fractures. J Orthop
6. Kregor PJ, Stannard JA, Zlowodzki M, Cole PA. Trauma. 2012;26(6):327–32.
Treatment of distal femur fractures using the less 19. Southeast Fracture Consortium. LCP versus LISS in
invasive stabilization system: surgical experience and the treatment of open and closed distal femur
early clinical results in 103 fractures. J Orthop fractures: does it make a difference? J Orthop
Trauma. 2004;18(8):509–20. Trauma. 2016;30(6):e212–6.
7. Hoffmann MF, Jones CB, Sietsema DL, Tornetta 20. Bottlang M, Doornink J, Lujan TJ, Fitzpatrick DC,
P 3rd, Koenig SJ. Clinical outcomes of locked Marsh JL, Augat P, et al. Effects of construct stiffness
plating of distal femoral fractures in a retrospective on healing of fractures stabilized with locking plates.
cohort. J Orthop Surg Res. 2013;8:43. J Bone Joint Surg Am. 2010;92(Suppl 2):12–22.
8. Henderson CE, Lujan TJ, Kuhl LL, Bottlang M, 21. Bottlang M, Fitzpatrick DC, Sheerin D, Kubiak E,
Fitzpatrick DC, Marsh JL. 2010 Mid-America Ortho- Gellman R, Vande Zandschulp C, et al. Dynamic
paedic Association physician in training award: fixation of distal femur fractures using far cortical
healing complications are common after locked locking screws: a prospective observational study.
plating for distal femur fractures. Clin Orthop Relat J Orthop Trauma. 2014;28(4):181–8.
Res. 2011;469(6):1757–65. 22. Ricci WM, Streubel PN, Morshed S, Collinge CA,
9. Henderson CE, Kuhl LL, Fitzpatrick DC, Marsh JL. Nork SE, Gardner MJ. Risk factors for failure of
Locking plates for distal femur fractures: is there a locked plate fixation of distal femur fractures: an
problem with fracture healing? J Orthop Trauma. analysis of 335 cases. J Orthop Trauma. 2014;28
2011;25(Suppl 1):S8–14. (2):83–9.
10. Moloney GB, Pan T, Van Eck CF, Patel D, Tarkin I. 23. Vallier HA, Hennessey TA, Sontich JK, Patter-
Geriatric distal femur fracture: are we underestimat- son BM. Failure of LCP condylar plate fixation in the
ing the rate of local and systemic complications. distal part of the femur: A report of six cases.
Injury. 2016;47(8):17326. J Bone J Surg Am. 2006;88(4):846–53.
11. Ebraheim NA, Martin A, Sochacki KR, 24. Barei DP, Beingessner DP. Open distal femur
Liu J. Nonunion of distal femoral fractures: a fractures treated with lateral locked implants: union,
systematic review. Orthop Surg. 2013;5(1):46–50. secondary bone grafting, and predictive parameters.
12. Gardner MJ, Toro-Arbelaez JB, Harrison M, Hier- Orthopedics. 2012;35(6):e843–6.
holzer Christian, Lorich DG, Helfet DL. Open 25. Kammerlander C, Riedmuller P, Gosch M, Zegg M,
reduction and internal fixation of distal femoral Kammerlander-Knauer U, Schmid R, Roth T. Func-
nonunions: long-term functional outcomes following tional outcome and mortality in geriatric distal
a treatment protocol. J Trauma. 2008;64(2):434–8. femoral fractures. Int J Care Injured. 2012;43
13. Henderson CE, Lujan T, Bottlang M, Fitzpatrick DC, (7):1096–101.
Madey SM, Marsh JL. Stabilization of distal femur 26. Wu CC. Retrograde dynamic locked nailing for
fractures with intramedullary nails and locking femoral supracondylar nonunions after plating.
plates: differences in callus formation. Iowa J Trauma. 2009;66(1):195–9.
Orthop J. 2010;30:61–8. 27. Chapman M, Finkemeier CG. Treatment of supra-
14. Lujan TJ, Henderson CE, Madey SM, Fitzpatrick DC, condylar nonunions of the femur with plate fixation
Marsh JL, Bottlang M. Locked plating of distal and bone graft. J. Bone Joint Surg. 1999;81(9):
femur fractures leads to inconsistent and asymmetric 1217–28.
274 A. Agarwal, MD
28. Bellabarba C, Ricci WM, Bolhofner BR. Indirect 33. Koval KJ, Seligson D, Rosen H, Fee K. Distal
reduction and plating of distal femoral nonunions. femoral non-union: treatment with retrograde
J Orthop Trauma. 2002;16(5):287–96. inserted locked intramedullary nail. J Orthop
29. Bellabarba C, Ricci WM, Bolhofner BR. Results of Trauma. 1995;9(4):285–91.
indirect reduction and plating of femoral shaft 34. Ali F, Saleh M. Treatment of distal femoral
nonunions after intramedullary nailing. J Orthop nonunions by external fixation with simultaneous
Trauma. 2001;15(4):254–63. length and alignment correction. Injury. 2002;33
30. Wang JW, Weng LH. Treatment of distal femoral (2):127–34.
nonunion with internal fixation, cortical allograft 35. Haidukewych GJ, Springer BD, Jacofsky DJ,
struts, and autogenous bone-grafting. J Bone Joint Berry DJ. Total knee arthroplasty for salvage of
Surg Am. 2003;85A(3):436–40. failed internal fixation or nonunion of the distal
31. Amorosa LF, Jayaram PR, Wellman DS, Lorich DG, femur. J Arthroplasty. 2005;20(3):344–9.
Helfet DL. The use of the 95-degree-angled blade 36. Davila J, Malkani A, Paiso JM. Supracondylar distal
plate in femoral nonunion surgery. Eur J Orthop Surg femoral nonunions treated with a megaprosthesis in
Traumatol. 2014;24(6):953–60. elderly patients: a report of two cases. J Orthop
32. Holzman MA, Hanus BD, Munz JW, O’Connor DP, Trauma. 2001;15(8):574–8.
Brinker MR. Addition of a medial locking plate to an 37. Vaishya R, Singh AP, Hasija R, Singh AP. Treatment
in situ lateral locking plate results in healing of distal of resistant nonunion of supracondylar fractures
femoral nonunions. Clin Orthop Relat Res. 2016;474 femur by megaprosthesis. Knee Surg Sports Trau-
(6):1498–505. matol Arthrosc. 2011;19(7):1137–40.
Nonunions of the Tibial Plateau
and Proximal Tibial Metaphysis 12
Thomas F. Higgins, MD
be worthy of a case report [6]. In a large series of

12.1 Epidemiology
open reduction and internal fixation (ORIF) with
plate and screw fixation, Barei et al. [7] reported
The risk of nonunion after operative treatment of
one case out of 83 bicondylar fractures that went
intra-articular tibial plateau fracture is considered
onto nonunion. Egol’s series of 36 plateau frac-
to be very low [1, 2]. Perhaps the reason for the
tures treated with a LISS (Less Invasive Stabi-
very low rate of nonunion is the rich extraoss-
lization System) plate featured two cases that
eous blood supply of the metadiaphyseal region,
were “prophylactically bone grafted” at the two
as thoroughly elucidated by Borrelli et al. [3].
month mark for metaphyseal bone loss [8]. Phi-
Branches of the popliteal artery perfuse this
sitkul’s [9] series of minimally invasive plating
metaphyseal zone from the anterior tibial artery
of plateau fractures demonstrated very high
laterally and from the posterior tibial artery
infection (22%) and malunion (22%) rates in
medially. Given the relative paucity of cases in
their series of mostly 41-C fractures, but they did
the literature on this topic, the best way to
not report on nonunions. Similarly, Rademaakers
examine the incidence is by separating articular
et al. [10] reported long-term follow-up on a
from extra-articular injuries, and further dividing
series of 109 fractures treated with ORIF in
by method of treatment.
which 4% were reported as malunion, but no
Historically, nonunion of intra-articular prox-
nonunions were described. However, in stark
imal tibial fractures is considered quite rare.
contrast to most papers on ORIF, a 2014 paper
Shatzker et al. [4] published a 1979 series of 94
by Ruffolo et al. [11] showed a 10% nonunion
plateau fractures with various treatments and all
rate in their series of 138 bicondylar plateau
united. In 1987, Moore and colleagues from the
fractures treated with plates and screws .
University of Southern California described
Skinny wire (or circular) external fixation also
between one- and ten-year follow-up on 399
seems to demonstrate a low rate of nonunion.
plateau fractures treated with or without opera-
A series of 24 plateau fractures from University
tion, and reported no nonunions [5]. Nonunion of
of California at Los Angeles treated with Ilizarov
articular injuries after nonoperative treatment
technique had no nonunions [12]. Similarly, the
was felt to be so rare that in 1991 King and
Canadian Orthopedic Trauma Society random-
Schatzker deemed the report of this occurrence to
ized 83 patients to circular frames or ORIF for
their plateau fractures, and reported no non-
unions [13]. Watson and Coufal [14] reported a
T.F. Higgins (&) series of 14 unicondylar plateau fractures treated
Department of Orthopedics, University Orthopedic
with circular frames, and all went on to union.
Center, University of Utah Health Sciences, 590
Wakara Way, Salt Lake City, UT 84108, USA Weiner’s series of 50 proximal tibial fractures
e-mail: Thomas.Higgins@hsc.utah.edu treated with “hybrid” external fixation also

276 T.F. Higgins, MD
demonstrated a 100% union rate [15]. Ahern also declare a fracture nonunited. The development of
described a series of 54 plateau fractures treated sclerotic margins along the intra- or
with either circular frames or locking plates, and extra-articular portion of the fracture on consec-
all fractures went onto union [16] . utive radiographic images may be indicative of
A number of papers address extra-articular an impending or established nonunion. Definitive
proximal tibia fractures, and various methods diagnosis is often aided by a computed tomog-
used to address these injuries. Martinez and raphy (CT) scan, which is also helpful in plan-
coauthors reported on their series of 108 proximal ning for repair if the diagnosis is confirmed.
tibia extra-articular fractures treated closed with a Although rare, intra-articular nonunion may be
functional brace [17]. The resultant three non- very difficult to definitively establish on plain
unions yielded a nonunion rate of 2.7%. Naik radiographs due to overlying hardware, and so a
et al. [18] reported on percutaneous locked plating CT scan is practically mandatory to establish this
for extra-articular proximal tibia fractures in 2013. diagnosis with confidence.
Similar to the Phisitkul study, which concerned After establishing the diagnosis, etiology must
mostly articular injuries, these 49 fractures had a be evaluated before treatment can be pursued.
high malunion rate (20%), but the three nonunions Patient behavioral and metabolic factors must be
in the series (6%) were all in infected cases. evaluated as potential causative agents. (For the
A large amount of the data on extra-articular tibia etiology of nonunions, see Chap. 1, “Nonunions:
fractures is based on efforts to use intramedullary Diagnosis, Evaluation and Management.”)
fixation to address proximal tibial metaphyseal For all open fractures and all fractures that
injuries. Ricci’s series of eleven intramedullary have been operatively repaired, infection must be
nails with blocking screws for extra-articular entertained as a potential etiological factor for
proximal tibia produced one nonunion [19]. nonunion. Workup for all previously operated or
Freedman and Johnson [20] reported malalign- open fractures should include complete blood
ment in 7 of 12 nailed proximal tibias, but did not count, erythrocyte sedimentation rate, and
report on nonunion . Krettek’s series of 21 tibia C-reactive protein. Any persistent drainage
fractures treated with Poller screws mentions one should be considered evidence of infection until
case that required eventual autograft [21]. Tor- proven otherwise. A patient history of intermit-
netta’s series of 25 proximal tibial fractures trea- tent prescription of oral antibiotic to suppress
ted with a nail, using a semi-extended technique, drainage should also be considered highly sug-
reported zero subsequent nonunions [22]. In gestive of infection. Nuclear medicine examina-
contrast, Lang and coauthors found in their series tions such as tagged white blood cell scan have
of proximal third tibia fractures treated with a nail been recommended if the presence or absence of
that 9 of 32 (28%) required exchange nailing or infection cannot be definitively established, but
bone grafting to unite, and two failed to unite the results of these tests are frequently ambigu-
entirely [23]. A 2012 series by Kulkarni et al. [24] ous or confusing. Magnetic resonance imaging
showed 5 nonunions out of 75 proximal tibial with gadolinium contrast has also been suggested
fractures (7%) treated with Poller screws and to evaluate infection, but in the presence of
intramedullary nailing. hardware from prior surgery, this is often not
feasible or helpful. If any of these investigations
indicate infection, a staged approach to treat-
12.2 Diagnosis ment, including open biopsy before definitive
fixation, is recommended.
Diagnosis of proximal tibia nonunion is based on The final etiologic contributors to nonunion
persistent discomfort with ambulation, radio- may be grouped under the heading of “surgeon
graphic evidence of nonunion, or progressive factors.” These include malalignment and
deformity of the proximal tibia over time. It is excessive stripping of the surrounding soft tis-
not clear how far after repair it is reasonable to sues at the time of index surgery. Malalignment
12 Nonunions of the Tibial Plateau and Proximal Tibial Metaphysis 277
of the proximal tibial metaphysis may preclude the affected site. If the defect is sizable or results
union, due to alterations in the mechanical axis, in instability, a cement spacer may be placed to
and may be suggested on either physical exam- fill this defect, with the potential secondary
ination or radiographic evaluation. Patient his- benefit of enhancing stability. If the debridement
tory may include subjective progressive loss of results in substantial increase in instability, tem-
limb alignment, which may be confirmed with a porary external fixation may need to be consid-
full-length weight-bearing film of the bilateral ered. In the event of positive cultures, the patient
lower extremities to evaluate side-to-side differ- is treated with a minimum of 6 weeks (or longer,
ence and limb length discrepancy. Rotational if recommended by infectious disease consulta-
malalignment may be harder to discern definition) of antibiotics. The requirements to resume
tively, but must be assessed either with gunsight definitive treatment of the nonunion after treat-
CT examination or prone physical examination ment of infection are as follows:
(thigh-foot axis) to uncover any tibial rotational
asymmetry. Malalignment must be addressed at 1. A dry healed incision with no drainage or
the time of definitive fixation in order to make warmth
union possible. 2. Absence of fevers or other constitutional
Surgical insult to the soft tissues and blood symptoms of infection
supply may be evidenced radiographically with a 3. Two weeks off antibiotics without recurrence
“wide surgical footprint.” The placement of of drainage or symptoms
multiple plates on multiple aspects of the tibial 4. Improvement, if not complete normalization,
metaphysis through a single incision may indi- of laboratory inflammatory parameters.
cate that there was extensive stripping of the
blood supply at the time of index surgery. In A more persistent or chronic infection in the
addition to stable fixation, nonunions that show setting of nonunion may influence the choice of
radiographic signs of having been previously definitive fixation, weighing in favor of circular
aggressively devascularized most often require frame methods.
autograft to address the biologic deficit. Intra-articular nonunion, although rare,
requires a very thorough understanding of the
fragment morphology before repair can be
12.3 Treatment undertaken. This is most often achieved with a
CT scan. In some instances, part of an
Assuming a staged approach for cases where intra-articular fragment may be united, and part
infection is suspected, treatment will focus on not. Because articular congruity and absolutely
repair of the aseptic nonunion. stability are the primary goals in order to pre-
In cases of suspected or confirmed infection, a serve the joint surface and achieve healing, open
staged approach is fairly standard. Briefly, this reduction of the joint stabilized with plate and
consists of standard laboratory values, to include screw fixation are the treatments of choice. The
complete blood count with differential, erythro- nonunited fragments are usually partial articular
cyte sedimentation rate, and C-reactive protein. (essentially “B-type” patterns), so lag screws and
These will serve as baseline levels. Prior to the nonlocking buttress fixation are the standard
index debridement and cultures, the patient is methods of fixation.
given a one-week holiday from antibiotics, pre- However, prior to rigidly repairing the non-
sumably to maximize the sensitivity of tissue union site, the metaphyseal extension of the
cultures. The goal of debridement is twofold: to fragment must be thoroughly debrided. This will
eliminate infected tissue and to obtain accurate often entail removing some healed bony tissue
cultures. If debridement of nonviable bone where the fragment is partially united at the
results in a defect, antibiotic cement beads may metaphysis. But if the fibrous tissue and callus are
be placed locally to deliver treatment directly to not completely removed from the nonunion site,
the fragment cannot be anatomically reduced to cancellous bone. When operating on the lower
the remaining tibia. This may often be quite time extremity, the proximal tibial metaphysis is a
consuming and tedious, but obtaining correct convenient source of autograft, but this may be
reduction without this step is not possible. Dis- difficult if the proximal tibia is the focal point of
section of the cortical side of the nonunited the established nonunion. However, when
fragment should be avoided to minimize the working on a lateral articular nonunion, the
vascular insult to this site, but the fragment must medial metaphysis may be used as a source, and
be adequately mobilized to achieve correct when working on lateral side, medial graft may
alignment of the fragment (Fig. 12.1). An excel- be collected. Once placed within the metaphyseal
lent description of the technique of René Marti, as nonunion site, the bone graft may be so bulky as
applied with an opening wedge osteotomy for to be an impediment to anatomic reduction of a
valgus malunion after plateau fracture, can be partial articular fragment, and in this case, the
found in a paper by Kerkhoffs and coauthors [25]. graft must be packed into a vacancy in the
Autologous bone grafting may or may not be metaphysis, or not utilized.
necessary in this setting. If needed, iliac crest has Once anatomic reduction is achieved, clamps
been utilized as the conventional harvest site for may be utilized to provisionally stabilize the
Fig. 12.1 a Knee radiographs demonstrate varus defor- through a single posteromedial approach restores articular
mity and apparent intra-articular nonunion. b Full-length congruity and alignment. e Films at one-year follow-up
weight-bearing view of the bilateral lower extremities demonstrate good healing with restoration of alignment
confirms varus malalignment as well as the relative limb and articular congruity. Long-standing films actually
shortening that results from varus. c Computed tomogra- demonstrate some valgus malalignment following surgi-
phy scan better defines the morphology of the nonunited cal correction. Patient reported markedly improved func-
fragment and the partial union of the metaphysis. tion clinically
d Correction with compression and buttress plating
nonunion and provide compression. Buttress applied through the application jig of many nail
fixation then relies on a well-contoured, or even systems. Another method for compression over a
slightly “under-contoured” plate, with the initial rod is dynamic locking, but this may not be
point of fixation being just distal to the apex of recommended in the setting of a potentially less
the nonunited fragment. If the patient is espe- stable metaphyseal nonunion.
cially osteopenic, due to age, disease, or pro- Smaller articular-epiphyseal segments may be
longed nonweight-bearing, cortical purchase of better captured and compressed with either
the buttress screws may be compromised, locking plate constructs or with Ilizarov tech-
requiring the use of locking fixation. Otherwise, niques. The degree of deformity may drive
nonlocking fixation should be ideal for this choice of fixation based on (1) suitability of
indication, and lag screws may be used, either implants to the corrected deformity or (2) the
through the plate or around the plate, for further pliability of soft tissue coverage at the concavity
rigid fixation of the nonunited articular fragment of the nonunion. For difficult soft tissue coverage
(see Fig. 12.1d). situations, circular frame correction over time
Extra-articular nonunion of the proximal tibia may allow for gradual correction and slow
presents a wider array of treatment options. expansion of the soft tissue envelope. This
Given the location, both hypertrophic and olig- technique may also be helpful in the setting of
otrophic nonunions would likely benefit from questionable patient compliance, as a circular
compression in this area, but this can be achieved frame may permit immediate weight bearing
with nailing, plate and screw constructs, or cir- (Fig. 12.4). For small peri-articular blocks, the
cular frames. author prefers dual locking plates applied sepa-
As discussed earlier, correction of alignment rately through two incisions, but familiarity with
is of the highest priority in addressing proximal Ilizarov techniques is required if soft tissues will
tibial nonunions. Extra-articular nonunions will not tolerate immediate on-table complete cor-
usually feature deformity in one or more of these rection of alignment.
planes (coronal, sagittal, rotational). In the To utilize locking plates for small
author’s experience, autogenous bone graft is peri-articular blocks, start with adequate
routinely utilized for almost all of these debridement of the metaphyseal nonunion site
nonunions. and correction of alignment. Alignment correc-
The two most important variables in deter- tion may be aided with an AO (Arbeitsgemein-
mining the ultimate method of fixation are (1) the schaft für Osteosynthesefragen/Association for
size of the proximal segment and (2) the degree the Study of Internal Fixation) universal dis-
of deformity. For larger proximal segments, tractor, or by leveraging the plates that will
where adequate fixation may be achieved with an ultimately be used for fixation. The author pre-
intramedullary device, this method may be prefers to establish proximal locking fixation both
ferred (Figs. 12.2 and 12.3). Blocking screws medially and laterally before imparting com-
may be employed to maintain correct alignment pression and final fine-tuning of alignment.
and enhance stability, and a variety of methods Balanced compression of each plate may then be
may be used to achieve compression with an used to fine-tune coronal plane alignment
intramedullary rod. Blocking screws are gener- (Figs. 12.5 and 12.6). Intraoperatively, fluoro-
ally left in place at the conclusion of the opera- scopy may not be able to adequately confirm
tion (as opposed to removing them after rodding) coronal plane alignment, and this may require an
to help maintain alignment as the fracture heals. intraoperative plain radiograph before finalizing
Compression may be achieved by locking the rod the distal fixation (see Fig. 12.6c).
distally first, and then “backslapping” the intra- On rare occasions of substantial articular bone
medullary rod, or via compression devices loss, or multiple failed attempts to get a small
Fig. 12.2 a Presenting films show frank nonunion of the compression through the rod. Immediate weight bearing
proximal tibial metaphysis and varus/procurvatum defor- was permitted. c The return of pain at the nonunion site
mity. b Correction was performed in an open manner with and broken screws after six months demonstrated persis-
autogenous cancellous bone graft and osteotomy of tent nonunion. d Medial and lateral compression plating
previously healed fibular fracture. The repair was fixed with bone grafting ultimately achieved osseous union
with an intramedullary rod utilizing blocking screws and
epiphyseal segment to heal to the metaphysis address the underlying problem (nonunion) but
without success, more extreme measures may be rather bypasses it in exchange for a return to
employed. Arthrodesis has been discussed as an function and diminished discomfort.
option, but there are no series in the literature
detailing this approach for proximal tibial non-
union. In addition to not addressing the under- 12.4 Outcomes
lying problem of metaphyseal nonunion,
arthrodesis comes with very significant func- There are scarce hard data on outcomes after
tional limitations. This should only be considered repair of proximal tibial nonunions. Toro-
as an option of last resort. Arbelaez et al. [26] reported a series of five
Arthroplasty has also been employed with intra-articular nonunions treated with correction
substantial articular loss or especially small epi- of deformity, lag screws, and a buttress plate.
physeal segments. This mandates a stemmed Four of five returned to pre-injury function, but
tibial component to bypass the nonunited site. two of five required total knee arthroplasty after
This is demonstrated in Fig. 12.7, where a very achievement of union. From the same institution,
poor ORIF of a bicondylar plateau fracture led to Michael Gardner et al. reported sixteen
a nonunited metaphysis and degenerative joint. extra-articular nonunions treated with deformity
Like arthrodesis, this does not necessarily correction, bone grafting, lateral plating, and
Fig. 12.3 a Presenting radiographs 4 months after frac- fracture would benefit from further stability but minimal
ture demonstrate acceptable alignment and some effort to insult to the local biology, a closed nailing was performed
heal the fracture, despite the nonunion. b Assuming the with resulting union and good function
compression. All healed within four months,

within five degrees of anatomical alignment [27]. 12.5 Summary
Wu and colleagues reported a series of 28 bi-
condylar plateau fractures initially treated with Nonunion of the proximal tibia is relatively rare.
lateral plating only that went onto either non- The evaluation requires thorough assessment of
union or malunion. They described open repair deformity in all planes, evaluation for infection,
of deformity using an angled blade plate and and appraisal of metabolic causes of nonunion.
compression with a minimum one-year follow-up There is little definitive outcomes data, but sur-
on 25 patients. All had healed by four months, gical intervention that includes correction of
and 22 of 25 showed improved function after alignment, compression, and bone grafting will
intervention [28]. most often yield satisfactory functional results.
Fig. 12.4 a Nonunion with ring fixator applied. b Healed nonunion after removal of frame
Case 2
12.6 Case Discussions A 38-year-old female seven years removed from
being struck by a bus. Her leg buckles when she
Case 1 attempts to walk. She has surprisingly good knee
A very active 65-year-old physicist presenting motion and no drainage. Laboratories were neg-
one year after open treatment of bicondylar tibial ative for infection or metabolic causes of
plateau fracture. nonunion.
Fig. 12.5 a Original postoperative films show restora- lateral locked plating is applied in some valgus. Subse-
tion of correct alignment. b Four months postoperatively, quent application of a medially based articulated tension-
deformity is present and fracture is not healed. ing device provides compression and correction of
c Full-length weight-bearing view of the bilateral lower alignment to anatomic. e Three-month follow-up films
extremities confirms varus alignment. d After grafting, already demonstrate solid union
Case 3 testosterone, which were treated. Compliance

A 57-year-old female with significant COPD on concerns led to selection of ring fixator and
corticosteroids failed closed treatment of this immediate weight bearing.
proximal tibial fracture in the setting of signifi-
cant osteopenia.
Case 5
Case 4 A 28-year-old male with unstable proximal tibia
A 22-year-old male with proximal metaphyseal metaphyseal fracture treated with lateral locked
tibial nonunion after closed treatment. Metabolic plate. Patient is referred at 4 months with pro-
workup revealed low vitamin D and low gressive deformity and nonunion.
Fig. 12.6 a Presenting films demonstrate varus defor- screws distally. c Intraoperative plain films are taken on
mity that was progressive since the time of his initial 17″ 14″ cassettes to confirm coronal plane alignment
stabilization in near-anatomic alignment. Sagittal plane with a larger field of view than offered by the fluoroscopy.
alignment appears good. b After establishing locking d After 4 months, osseous union is confirmed radio-
fixation in the proximal segment medially and laterally, graphically. e Full-length weight-bearing view of the
compression and fine-tuning of alignment were achieved bilateral lower extremities confirms restoration of
with medial and lateral Verbrugge clamps and push-pull alignment
Case 6 An active 82-year-old male with proxi- Case 7 An active 72-year-old female with bi-
mal tibial fracture treated with lateral locked condylar tibial plateau fracture treated initially
plating, presented four months after index treat- with spanning external fixation and secondary
ment with progressive deformity. Preoperative ORIF. At one year, patient is referred with non-
workup revealed vitamin D deficiency, which union, varus deformity, and lateral compartment
was treated. arthrosis.
Fig. 12.7 a Presentation films one year after injury show metaphysis. (Case example and image courtesy of
varus malreduction with lateral joint arthrosis and meta- Christopher L. Peters, MD, University of Utah, Salt Lake
physeal nonunion. b Stemmed total knee arthroplasty is City, Utah, USA)
used to correct alignment and bypass nonunited
fractures utilizing a two-incision technique. J Orthop

References Trauma. 2004;18(10):649–57.
8. Egol KA, Su E, Tejwani NC, Sims SH, Kummer FJ,
1. Choo KJ, Morshed S. Postoperative complications Koval KJ. Treatment of complex tibial plateau
after repair of tibial plateau fractures. J Knee Surg. fractures using the less invasive stabilization system
2014;27(1):11–9. plate: clinical experience and a laboratory comparison
2. Papagelopoulos PJ, Partsinevelos AA, Themisto- with double plating. J Trauma. 2004;57(2):340–6.
cleous GS, Mavrogenis AF, Korres DS, Soucacos PN. 9. Phisitkul P, McKinley TO, Nepola JV, Marsh JL.
Complications after tibia plateau fracture surgery. Complications of locking plate fixation in complex
Injury. 2006;37(6):475–84. proximal tibia injuries. J Orthop Trauma. 2007;21
3. Borrelli J Jr, Prickett W, Song E, Becker D, Ricci W. (2):83–91.
Extraosseous blood supply of the tibia and the effects 10. Rademakers MV, Kerkhoffs GM, Sierevelt IN, Raay-
of different plating techniques: a human cadaveric makers EL, Marti RK. Operative treatment of 109
study. J Orthop Trauma. 2002;16(10):691–5. tibial plateau fractures: five- to 27-year follow-up
4. Schatzker J, McBroom R, Bruce D. The tibial plateau results. J Orthop Trauma. 2007;21(1):5–10.
fracture. The Toronto experience 1968–1975. Clin 11. Ruffolo MR, Gettys FK, Montijo HE, Seymour RB,
Orthop Relat Res. 1979;(138):94–104. Karunakar MA. Complications of high-energy
5. Moore TM, Patzakis MJ, Harvey JP. Tibial plateau bicondylar tibial plateau fractures treated with dual
fractures: definition, demographics, treatment ratio- plating through two incisions. J Orthop Trauma.
nale, and long-term results of closed traction man- 2015;29(2):85–90.
agement or operative reduction. J Orthop Trauma. 12. Mikulak SA, Gold SM, Zinar DM. Small wire
1987;1(2):97–119. external fixation of high energy tibial plateau frac-
6. King GJ, Schatzker J. Nonunion of a complex tibial tures. Clin Orthop Relat Res. 1998;356:230–8.
plateau fracture. J Orthop Trauma. 1991;5(2): 13. Watson JT, Coufal C. Treatment of complex lateral
209–12. plateau fractures using Ilizarov techniques. Clin
7. Barei DP, Nork SE, Mills WJ, Henley MB, Orthop Relat Res. 1998;353:97–106.
Benirschke SK. Complications associated with inter- 14. Canadian Orthopaedic Trauma Society. Open reduc-
nal fixation of high-energy bicondylar tibial plateau tion and internal fixation compared with circular
fixator application for bicondylar tibial plateau with small diameter intramedullary nails. J Bone
fractures. Results of a multicenter, prospective ran- Joint Surg Br. 1999;81(6):963–8.
domized clinical trial. J Bone Joint Surg Am. 22. Tornetta P 3rd, Collins E. Semiextended position of
2006;88(12):2613–23. intramedullary nailing of the proximal tibia. Clin
15. Weiner LS, Kelley M, Yang E, Steuer J, Watnick N, Orthop Relat Res. 1996;328:185–9.
Evans M, Bergman M. The use of combination 23. Lang GJ, Cohen BE, Bosse MJ, Kellam JF. Proximal
internal fixation and hybrid external fixation in severe third tibial shaft fractures. Should they be nailed?
proximal tibia fractures. J Orthop Trauma. 1995;9 Clin Orthop Relat Res. 1995;315:64–74.
(3):244–50. 24. Kulkarni SG, Varshneya A, Kulkarni S, Kulkarni GS,
16. Ahearn N, Oppy A, Halliday R, Rowett-Harris J, Kulkarni MG, Kulkarni VS, Kulkarni RM. Intrame-
Morris SA, Chesser TJ, Livingstone JA. The out- dullary nailing supplemented with Poller screws for
come following fixation of bicondylar tibial plateau proximal tibial fractures. J Orthop Surg (Hong
fractures. Bone Joint J. 2014;96-b(7):956–62. Kong). 2012;20(3):307–11.
17. Martinez A, Sarmiento A, Latta LL. Closed fractures 25. Kerkhoffs GM, Rademakers MV, Altena M,
of the proximal tibia treated with a functional brace. Marti RK. Combined intra-articular and varus open-
Clin Orthop Relat Res. 2003;417:293–302. ing wedge osteotomy for lateral depression and
18. Naik MA, Arora G, Tripathy SK, Sujir P, Rao SK. valgus malunion of the proximal part of the tibia.
Clinical and radiological outcome of percutaneous J Bone Joint Surg Am. 2008;90(6):1252–7.
plating in extra-articular proximal tibia fractures: a 26. Toro-Arbelaez JB, Gardner MJ, Shindle MK,
prospective study. Injury. 2013;44(8):1081–6. Cabas JM, Lorich DG, Helfet DL. Open reduction
19. Ricci WM, Rudzki JR, Borrelli J Jr. Treatment of and internal fixation of intraarticular tibial plateau
complex proximal tibia fractures with the less nonunions. Injury. 2007;38(3):378–83.
invasive skeletal stabilization system. J Orthop 27. Gardner MJ, Toro-Arbelaez JB, Hansen M, Bora-
Trauma. 2004;18(8):521–7. iah S, Lorich DG, Helfet DL. Surgical treatment and
20. Freedman EL, Johnson EE. Radiographic analysis of outcomes of extraarticular proximal tibial nonunions.
tibial fracture malalignment following intramedullary Arch Orthop Trauma Surg. 2008;128(8):833–9.
nailing. Clin Orthop Relat Res. 1995;315:25–33. 28. Wu CC. Salvage of proximal tibial malunion or
21. Krettek C, Stephan C, Schandelmaier P, Richter M, nonunion with the use of angled blade plate. Arch
Pape HC, Miclau T. The use of Poller screws as Orthop Trauma Surg. 2006;126(2):82–7.
blocking screws in stabilising tibial fractures treated
Tibial Nonunions
13
Samuel E. Galle, MD and David P. Zamorano, MD
13.1 Introduction 13.1.2 Incidence of Nonunion
13.1.1 Tibial Shaft Fractures Nonunions are defined by the US Food and Drug
Administration as the failure of fracture union by
9 months after injury or the lack of radiographic
Tibial shaft fractures are universal orthopedic
injuries, as they are the most common long bone progress toward union over 3 consecutive
months [5]. This definition is debated in the lit-
diaphyseal fracture [1]. According to the US
erature. Many authors, among them the senior
National Center for Health Statistics, fractures of
tibia, fibula, and ankle together have an annual author of this chapter, define nonunions simply
as those fractures which have failed to progress
incidence of 492,000 [2]. A significant portion of
to union from prior clinic visits or have little to
these injuries is attributed to road traffic injuries,
falls, and various sporting activities, and the no potential for further healing without additional
treatment or intervention. Clinically, they present
majority of these athletic injuries being attributed
with persistent pain, motion, and residual swel-
to soccer. Once sustained, tibia and fibula frac-
tures result in 77,000 hospitalizations annually, ling about the nonunion site. As tibial fractures
are the most common long bone injury, tibial
accounting for 569,000 hospital days, 825,000
nonunions occur more frequently than any other
physician office visits, and considerable direct
and indirect healthcare costs [3, 4]. These long bone [5, 6]. This is probably due to
high-energy mechanisms that are often associ-
fractures are commonly treated with cast man-
ated with tibial fractures and the limited soft
agement, intramedullary fixation, plate os-
teosynthesis, or external fixation, depending on tissue envelope. Additionally, the tibia is prone
to significant soft tissue injury as well as damage
the initial presentation and severity of the frac-
to the underlying vascular supply. The incidence
ture. Various authors have reported tibial fracture
union in approximately 17 weeks time; never- of delayed unions and nonunions of the tibia has
been reported as high as 48%. Heckman et al.
theless, some patients suffer from long-term
reported an average incidence of 25% for all
functional impairment regardless of union
status [1]. tibial delayed union and nonunions [7, 8].
Various authors have retrospectively reviewed
tibial fracture cases for criteria that might define
those most likely to progress to nonunion in the
future. Fong et al. described fracture character-
S.E. Galle D.P. Zamorano (&) istics associated with tibial nonunions and
Orthopedic Trauma Service, Orthopedic Trauma &
found that fractures with less than 25% cortical
Fracture Clinic, Saint Alphonsus—Boise Campus,
901 Curtis Road, Suite 501, Boise, ID 83706, USA contact were the most predictive of nonunion
e-mail: David.Zamorano@saintalphonsus.org and subsequent re-operation. The authors also

288 S.E. Galle, MD and D.P. Zamorano, MD
acknowledged that open fractures, presence of doubled that of those who did not have a non-
comminution, and oblique or segmental fractures union ($25,555.97 vs. $11,686.24, P < 0.001).
were also associated with tibial nonunions as
well [9]. Teraa et al. [10] found significantly
more delayed unions and nonunions in segmental 13.2 Causes of Nonunion
fractures than nonsegmental fractures of the tibia.
Nicoll [11] analyzed treatments of 705 patients Tibial fractures that progress to nonunions can be
with tibial fractures and concluded that the most grouped into one of two categories: (1) those that
significant factors affecting healing were initial lack stability and (2) those that lack biology
displacement, comminution, soft tissue wounds, (Table 13.1). This is an important distinction as
infection, and distraction. In 2013, Yang et al. it guides the surgeon’s treatment.
developed questionnaires applied to 56 patients
with tibial fractures treated with intramedullary
nailing and asked three fellowship-trained trauma 13.2.1 Mechanical
surgeons to assess for likelihood of progression
to nonunion at 3 months postoperatively. The Instability may cause excessive motion at the
authors found diagnostic accuracy of 74% with a fracture site, which encourages stem cells to
sensitivity of 62% and specificity of 77%, indi- differentiate into fibroblasts, resulting in the for-
cating that nonunion prediction is improving mation of fibrous tissue formation and delayed
based on the clinical criteria as described by union or nonunion. Factors that promote insta-
Fong, Teraa, and Nicoll but clearly this is an area bility at the fracture site include bony com-
of future clinical research to improve the accu- minution, inadequate or poor plate fixation, small
racy of nonunion prediction [12]. diameter nails, poorly constructed external fixa-
tion constructs, and inadequate bony contact.
Comminuted tibial shaft fractures are typically
13.1.3 Ramifications of Nonunion treated with relative stability techniques includ-
ing intramedullary nailing, bridge plating, and
Tibial nonunions can have a catastrophic effect external fixation. When treating these fractures, it
on a patient. Limb length inequality, rotational is important for the surgeon to choose the con-
and angular deformity, pain and instability can struct based on appropriate stiffness. This can be
all result in the inability to work, placing further a difficult task and requires much experience with
economic burdens on the patient [13, 14]. Brin- tibial nonunions and the fixation constructs
ker et al. retrospectively reviewed 243 tibial shaft available. In general, the use of larger diameter
nonunions in 237 consecutive patients and per- nails, stiffer plates, and multi-planer external
formed quality of life outcome measurements. fixation constructs can decrease the risk of non-
The most commonly affected tibial segment was union in comminuted tibial shaft fractures.
the distal third (49%), and infection was present Bony contact is also an important factor in
in 18% of all cases. They noted that “the impact providing stability to the fracture. As bony con-
of tibial shaft fracture nonunion on physical tact decreases, fracture stability will also
health was comparable with the reported impact decrease, predisposing to hypertrophic nonunion.
of end-stage hip arthrosis and worse than that of
congestive heart failure” [15]. The authors con-
cluded that tibial nonunions are devastating in 13.2.2 Biological
nature and have a significant adverse effect on
quality of life. In terms of economic cost, Ano- There are many factors that contribute to “bio-
tonova et al. [16] reported on the economic logic” nonunions. A common cause includes
burdens of tibial nonunions and found median poor blood supply to the fracture often secondary
total care cost for tibial nonunion patients almost to the soft tissue injury, surgical technique, or a
13 Tibial Nonunions 289
Table 13.1 Biological risks for nonunion inhibitory effects on the formation of
Open fracture fibroblast-rich granulation tissue leading to
Age impaired healing [21]. Smoke inhalation leads to
Smoking
low concentration of antioxidant vitamins and
reactive oxygen species that cause cellular dam-
Alcohol
age, particularly in osteoblasts, fibroblasts, and
NSAIDs
macrophages. Nicotine has been shown to
Nutritional deficiency increase platelet aggregation, to inhibit fibroblast
Prior radiation function, and to decrease blood flow to extrem-
Endocrinopathies ities due to increased peripheral vasoconstriction
Infection [22].
A large number of studies document the
effects of smoking and nicotine in various animal
combination of the two. Bishop et al. [17] models. In the rabbit model, Donigan et al.
described patient-related contributors of non- studied the effects of transdermal nicotine on
unions and included medical comorbidities, fracture healing in 22 mid-shaft tibial osteo-
advancing age, smoking, alcohol abuse, nons- tomies treated with plate fixation. They noted
teroidal anti-inflammatories, nutritional defi- that, although the nicotine-treated rabbits had
ciency, prior radiation treatment, genetic similar areas of periosteal callus formation, these
disorders, and various metabolic diseases. rabbits had significantly less torsional resistance
and stiffness at 21 days postoperatively and three
13.2.2.1 Open Fracture rabbits had gross nonunions [23]. Similar results
The tibial shaft maintains a subcutaneous reported by Raikin et al. [24] showed that
anatomical location for a substantial portion of its nicotine-exposed rabbits had tibial healing that
length especially along its medial border. Poor was 26% weaker resistant to three-point bending
soft tissue coverage of the tibia has long been than those not exposed. In humans, the majority
associated with a higher incidence of open frac- of research confirmed similar associations as that
tures, which have a higher likelihood of non- of the animal models; however, this has been
union. Rosenthal et al. retrospectively reviewed mostly retrospective reviews rather than the
104 open tibial fractures for the relationship understandably difficult prospective, randomized
between initial wound presentation and potential study. Castillo et al. as part of their prospective
for healing. Records were analyzed for some 71 lower extremity assessment project of 268 tibial
patients: all Gustilo type I fractures united, two fracture patients revealed that current smokers
patients in type II continued to nonunion, and 13 had a higher incidence of nonunion at 24 months
patients in the type III fracture classification went after injury compared to nonsmokers (24.1 vs.
on to nonunion. The authors concluded that there 9.9%, respectively). Smokers were also more
was a strong association between fractures that than twice as likely to develop infection and 3.7
suffered nonunions and extensive soft tissue loss times as likely to develop osteomyelitis [25]. In a
[18]. retrospective study, Adams et al. [26] compared
140 smoking and 133 nonsmoking patients.
13.2.2.2 Smoking Mean time to union was 32 weeks compared to
While surgeons cannot always choose their 28 weeks, respectively. Clearly, there is an
patients in trauma, care should be taken in association with smoking and delayed fracture
selecting patients for elective nonunion surgery. healing, but further research is necessary to
Cigarette smoking and nicotine have been identify the exact molecular pathway and possi-
implicated in inhibiting fracture healing and ble therapeutics to counteract its effects. Prior to
increasing the risk of delayed union or nonunion any surgical intervention, smoking cessation
[19, 20]. The effects of smoking are related to its should be emphasized to enhance the likelihood
of healing. Urine and/or blood screening for abnormalities have all been implicated in the for-
nicotine and cotinine can be used to confirm mation of nonunions. Brinker et al. [33] analyzed
patient’s smoking status. Clinical experience has 37 prescreened nonunion patients with the
shown that blood levels of nicotine will return to hypothesis that these idiopathic nonunions iden-
normal within 2 weeks of cessation while urine tified actually had underlying endocrine and
will be positive for several weeks. metabolic abnormalities. They found that 83.8%
of the 37 patients had some type of endocrinopathy
13.2.2.3 Medications with the most common being vitamin D defi-
Nonsteroidal anti-inflammatory drugs (NSAIDs) ciency. These authors proposed a diagnostic
are commonly cited in the literature as being algorithm for identifying these patients for further
associated with delayed unions and nonunions, workup by an endocrinologist as part of their
while controversy remains regarding their effect study. Additional research may further elucidate
on fracture repair [27]. The exact biochemical the causal nature of various endocrinopathies and
pathway is an area of further research, but many metabolic disorders and their relationship to non-
authors have theorized that these medications unions, as well as potential medical treatments.
inhibit cyclooxygenase leading to less pros-
taglandin E2, which leads to less bone formation 13.2.2.5 Infection
by osteoblasts [28, 29]. Zhang et al. [30] pro- Infected tibial nonunions pose a complex clinical
posed a schematic model for cyclooxygenase-2, problem for surgeons and can lead to significant
(COX-2) effects on bone repair after fracture morbidity. In the setting of tibial fractures,
using COX knock out mice, whereby decreased infections are propagated from open wounds or
levels lead to decreased production of pros- introduced during surgical management. Sta-
taglandin E2, which may lead to low levels of phylococcus aureus is the most commonly
(bone morphogenic) protein. Simon and implicated organism and has been found in 65–
O’Connor expanded on this murine model and 70% of patients with long bone infections [34].
administered various doses of celecoxib, a On the microscopic level, bacteria will form a
selective COX-2 inhibitor, to explore the biofilm or glycocalyx that significantly inhibits
dose-dependent and time-dependent effects of ability of the immune system to clear the infec-
this NSAID. The authors found impaired healing tion. This leads to involucrum formation, which
with increasing dosage radiographically, in tor- is reactive bone, as the body attempts to limit the
sional stability, and overall increased formation spread of the infection. Shortly following is
of nonunion [31]. Giannoudis et al. [32] retro- sequestrum formation, or necrotic bone, indicat-
spectively reviewed the effects of NSAIDs on ing a chronic infection with little ability to heal
femoral nonunions in 32 patients and noted a without intervention.
strong correlation. While this association has not
been proven definitively in humans with a
prospective randomized control trial, caution 13.3 Evaluation and Diagnosis
should be used when prescribing NSAIDs in the
setting of tibial fractures, especially in those 13.3.1 History
patients with impaired healing, e.g., smokers,
diabetics, etc. Of the utmost importance in defining the scope of
the problem is the history of the tibial fracture
13.2.2.4 Endocrinopathies and prior treatment modalities that have failed to
Patients who present with a tibial nonunion with- obtain fracture union. This includes mechanism
out an obvious cause should be worked up for an of injury, prior open wounds, pain with weight
endocrinopathy. Vitamin D, vitamin C, calcium, bearing, feelings of instability, and any delayed
thyroid hormone, and parathyroid hormone wound healing. Patients who present with tibial
nonunions have often had an extensive treatment hormone, phosphorus, and alkaline phosphatase
history at multiple institutions. Previous records, levels.
including operative notes, injury and postopera-
tive radiographs, and any pertinent laboratory
values, should be obtained from all previously 13.3.4 Radiographs
treating physicians. Questions specific to
infectious etiology are particularly important, Radiographs on initial presentation should
covering wound drainage, prior cellulitis, con- include the standard anteroposterior and lateral of
stitutional symptoms, pertinent cultures/ the tibia/fibula to document the characteristics of
sensitivities, and previous antibiotic treatment the nonunion. Forty-degree internal and external
regimens. A complete account of the patient’s oblique views and stress views may also be
chronic illnesses is also important and will help useful to better characterize the nonunion.
to guide treatment algorithms. This should
include the patient’s nutritional status, smoking
history, constitutional symptoms, and prior his- 13.3.5 CT/MRI
tory of nonunion.
Computed tomography (CT) and magnetic res-
onance imaging (MRI) of nonunions are
13.3.2 Physical Exam important tools for defining the three dimen-
sional extent of tibial pathology. CT scans can
The physical examination of all tibial nonunions provide useful information regarding the number
begins with observation of the lower extremity of cortices that have healed across a tibial frac-
for prior wounds, surgical incisions, erythema, ture site with bridging callus formation. One
gross deformity, and the general state of the study of CT scans for the presence of tibial
surrounding soft tissue. Tenderness to palpation nonunion found 100% sensitivity and 62%
about the nonunion site should be noted and specificity [35]. MRI of tibial nonunions may
gross motion may be found as well. The surgeon delineate soft tissue infections from underlying
should document the limb vascularity, limb osteomyelitis. Osteomyelitis appears as a low
lengths, and range of motion of the knee and signal intensity of T1-weighted images and high
ankle joints, as contractures may have occurred. signal intensity of T2-weighted sequences
(Fig. 13.1a, b).
13.3.3 Laboratories
13.3.6 Nuclear Imaging
Important laboratory markers in the evaluation of
tibial nonunions that can help guide the surgeon’s Various modalities are available and are used to
treatment include erythrocyte sedimentation rate evaluate infection as an etiology of the nonunion.
(ESR), C-reactive protein (CRP), and white blood The most commonly used nuclear medicine
cell (WBC) count. Unfortunately, many authors scans of nonunions include technetium-99 m,
note that negative laboratory markers do not gallium-67 citrate, and indium-111-labeled
completely rule out indolent infection. In non- leukocyte. Madsen recently reported a case
unions with reasonable stabilization, laboratory report on the use of bone SPECT/CT imaging to
evaluation for metabolic and endocrine disorders detect sequestrum formation in a chronically
should be obtained in consultation with an infected tibial nonunion [36]. Further research is
endocrinologist, as previously discussed. These required to determine the clinical applicability of
markers include serum calcium, serum using SPECT/CT in the setting of tibial
25-hydroxy-vitamin D, thyroid-stimulating nonunions.
Fig. 13.1 Demonstration of osteomyelitis on magnetic Animesh Agarwal, MD, Department of Orthopedics,
resonance imaging. T1-weighted image are low signal and University of Texas Health Science Center, San Antonio,
appear dark (a) while T2-weighted images show high Texas, USA
signal (b). Images courtesy of Animesh Agarwal, MD
consider factors such as the patient’s functional

13.4 Treatment level, occupation, and expectations when devel-
oping a treatment plan to ensure its eventual
The most important aspect of treating tibial non- success for both the surgeon and the patient.
unions is identifying and correcting the underly- Various treatment modalities will be discussed
ing cause of the nonunion. This may be a systemic and the surgeon should select one based on their
issue, such as an endocrinopathy, or a localized training, comfort level, and prior experience.
pathology, for example an infection. Once the
etiology of the nonunion has been addressed, the
surgeon can continue his or her plan to repair the 13.4.1 Based on Nonunion Type
nonunion. Basic treatment modalities include
improving fracture mechanics, restoring the local 13.4.1.1 Hypertrophic
biology of the fracture, providing electrical or In most cases of hypertrophic nonunion, fracture
ultrasonic stimulation, and various combinations stabilization is the fundamental management
of these modalities. The surgeon should also concept and bone grafting is generally not
13.4.1.2 Atrophic
Unlike hypertrophic nonunions, atrophic tibial
nonunions present with poor callus formation,
indicating little to no attempt at fracture healing
(Figs. 13.4 and 13.5). Classically, these non-
unions are thought to be poorly vascularized, but
recent research has elucidated a more compli-
cated understanding of atrophic nonunions.
Matuszewski and Mehta recently described a
case report of a 30-year-old patient who sus-
tained a type IIIC tibial shaft fracture initially
treated with vascular repair, soft tissue coverage
and plating and yet unfortunately progressed to
an aseptic, atrophic nonunion. The treating team
noticed pallor of the lower extremity and angio-
gram revealed stenosis of both the anastomosis
sites. After angioplasty, the patient planned on
further intervention but was delayed secondary to
Fig. 13.2 Hypertrophic nonunion. Broken distal inter-
lock (red arrow) consistent with excessive motion at the pregnancy. When she returned to clinic 5 months
fracture site. Note the abundant callus formation which is later, 15 months after initial treatment, the frac-
a hallmark of a hypertrophic nonunion ture site was radiographically healed without
further intervention, implying the importance of
vascular supply in the setting of atrophic non-
necessary. Hypertrophic tibial nonunions present unions [37]. Brownlow et al. analyzed 16 rabbits
with callus formation about the fracture ends, with atrophic nonunions at various time points to
leading to the very characteristic flared ends as document the vascularity at the fracture site
the bone attempts to unite (Fig. 13.2). These compared with controls. The authors found that
nonunions tend to be well vascularized but are at 1 week the control fracture sites were vascu-
thought to lack the requisite mechanical stability larized and the experimental fracture sites were
for bone formation. Once bony stability has been nonvascularized, but this difference resolved by 8
restored, motion at the fracture site is decreased and 16 weeks [38].
and allows for capillary ingrowth with eventual Treatment goals should be focused on the
enchondral ossification. Often this case presents underlying etiology of the nonunion, stimulating
as a tibial shaft that had been previously treated a healing response, and providing stable fixation
with an intramedullary nail. In such a case, if needed. Bone grafting with autogenous graft
exchange nailing with a larger nail can provide remains the gold standard but adjuncts such as
the necessary stability to promote bone healing. bone morphogenic protein-2 and parathyroid
In cases where the nonunion had been treated hormone (PTH) can also be useful. Atrophic
nonoperatively or with an unstable external fixa- nonunions are most likely multifactorial and
tor, intramedullary nailing or compression plating present an area for further research.
is all that is required to obtain fracture healing.
Figure 13.3 demonstrates a hypertrophic non- 13.4.1.3 Oligotrophic
union due to a lack of mechanical stability. Oligotrophic nonunions are those that have
A typical large callus has formed in the attempt to characteristics of both atrophic and hypertrophic
increase fracture stability. The nonunion was nonunions, as previously discussed. Management
treated with exchange intramedullary nailing with options follow those for hypertrophic nonunions
a larger diameter nail. No bone grafting or other as well as examination for possible causes of a
biological adjunct was used to achieve union. biological lack of bony healing.
Fig. 13.3 32-year-old male with a closed tibial shaft fracture (a). Closed intramedullary nailing with a 10 mm nail
resulted in a hypertrophic nonunion (b). Exchange nailing was successfully performed using a 11.5 mm nail (c, d)
13.4.1.4 Infected pain, erythema, swelling, and possibly a draining

Infected nonunions of the tibia can be a chal- sinus. Laboratory markers include ESR, CRP
lenging problem. Multiple surgeries are usually levels, and a WBC count may be elevated and
required to get adequate debridement and even- can be used to diagnose and demonstrate clinical
tually restore the function of the limb. Patzakis improvement after treatment. Computed tomog-
and Zalavras [39] summarized the basic tenets of raphy or MRI are more useful than plain radio-
care, which includes surgical debridement, graphs in identifying affected areas, periosteal
antibiotics, fracture stabilization, adequate soft reactions, or abscess formations in the preoper-
tissue coverage, and eventual restoration of bone ative setting. Erdman et al. [41] demonstrated
defects. that MRI of patients suspected of having
Cierny et al. [40] described the basic classi- osteomyelitis had a sensitivity of 98% and a
fication schema of osteomyelitis based on ana- specificity of 75%.
tomic types, e.g., medullary, superficial, Intra-operative cultures from the sinus tract,
localized, and diffuse, and patient characteristics purulent discharge, soft tissue, and
based on underlying comorbidities (Fig. 13.6). curetted/debrided bone are imperative, as they
Clinically, these patients present with ongoing can help determine a proper antibiotic regimen.
of the patient. In either case, a good support

system for the patient is imperative for a suc-
cessful outcome.
Fracture stability is of the upmost importance
in treating infected tibial nonunions. The senior
author was taught as a resident, and this still
stands true that “an infected STABLE nonunion
is better than an infected UNSTABLE non-
union.” Patzakis and Zalavras [39] have similarly
recommended retaining the implants in infected
nonunions in certain clinical situations, e.g.,
early diagnosis, known bacterial species, antibi-
otic sensitive species, etc. Implants that may be
colonized or have failed should be replaced with
either external fixation or intramedullary nailing;
however, plate fixation may also be reasonable in
certain settings. Megas et al. [43] treated nine
patients with infected tibial nonunions and bone
defects of 2–12 cm after intramedullary nailing
Fig. 13.4 Classic atrophic nonunion. Notice there is no with Ilizarov external fixation and reported a
evidence of bone healing. These types of nonunions
typically require a biological stimulus to promote healing
100% union rate with a mean external fixation
time of 187.4 days. Consideration should also be
given to the placement of poly(methyl
Perry et al. [42] noted that superficial wound methacrylate) (PMMA) beads impregnated with
cultures and needle aspirations were insufficient heat-stable antibiotics such as tobramycin and
to rule out infection perioperatively. vancomycin. Holtom and Patzakis [44] recom-
For the medullary and superficial cases of mended approximately 2.4–4.8 g of tobramycin,
osteomyelitis (Cierny-Mader type I and II), or vancomycin, 2–4 g, per 40 g of PMMA
general consensus treatment includes removal of cement to achieve local bactericidal conditions.
metal implants and radical debridement of all Adequate soft tissue coverage should also be
involved bone and soft tissue. It is important that obtained during the wound debridement of
the surgeon does not sacrifice a thorough infected tibial nonunions. This is usually
debridement for the hope of an easier recon- accomplished with a rotational muscle flap or a
struction. While reconstructing large soft tissue free muscle transfer, depending on the integrity
and bony defects can be difficult, an inadequate of the local tissues and the size and location of
debridement will be doomed to failure. For the the soft tissue defect. Muscle transfers are par-
localized and diffuse cases (Cierny-Mader type ticularly useful for providing a new influx of
III and IV), nonviable bone must be debrided vascular supply, which improves antibiotic dis-
fully and may require reconstruction at a later persal and host immune system, preventing fur-
date. The complicated diffuse cases with exten- ther microbial seeding. Figure 13.7 demonstrates
sive bony and soft tissue defects may result in the usefulness of muscle transfers to help eradi-
amputation as the only viable treatment option, cate osteomyelitis and heal a nonunion. A thor-
especially when presenting in patients with sev- ough debridement of the infected area is vital to a
ere comorbidities. Amputation versus limb sal- good outcome. Once the surgical debridement is
vage is a clinical judgment based on patient complete, an antibiotic cement depot is inserted
comorbidities, soft tissue defects, bony involve- and the patient receives intravenous antibiotics
ment, neurovascular assessment, and the desires tailored to the specific bacteria that is cultured.
Fig. 13.5 22-year-old male with a type IIIC open tibia reamer-irrigator/aspirator technique. Nonunion eventually
(a). Free flap with intramedullary nailing and cement healed with small anteromedial defect (c, d)
spacer (b) that was eventually bone grafted using a
After 6 weeks of antibiotics, the flap is elevated, provide the best implant fixation. Treatment
the cement spacer is removed, and autogenous regimens include fixed-angle plating, external
bone graft is placed in the defect. fixation, and at times intramedullary nailing.
Harvey et al. [45] reported on 17 proximal and
13.4.1.5 Nonunion Location 13 distal tibial nonunions using customized
Peri-articular nonunions are a relatively uncom- blade plate fixation, with 29 unions and five
mon occurrence but they can be difficult clinical persistent nonunions after initial blade plate fix-
problems to treat. Metaphyseal bone of the ation. These authors found that blade plate fix-
peri-articular region is well vascularized which ation was able to achieve eventual union in
provides the basic components for fracture 97.2% of peri-articular fractures. Gardner et al.
healing. Nevertheless, nonunions do occur in this performed a retrospective clinical study on 16
area and the soft metaphyseal bone may not patients with proximal tibial nonunions treated
Fig. 13.6 Basic classification schema of osteomyelitis based on anatomic types. Modified from Cierny et al. [40], with
permission
with deformity correction, bone grafting, and 13.4.1.6 Segmental Defects

lateral plating. The authors found that all non- Segmental tibial nonunions are a clinically
unions healed at an average of 4 months, Knee challenging problem to manage, as the body’s
Society function scores improved significantly, natural ability to fill in bony defects is fairly
and 88% were able to return to their prior limited [49]. Surgical options include acute
activities [46]. Reed and Mormino reviewed shortening with possible future bone lengthening,
functional outcomes after distal tibial metaphy- autologous cancellous bone grafting, vascular-
seal nonunion fixation with blade plates. The ized fibula cortical bone graft, and bone transport
authors found all 11 patients had healed and with an Ilizarov frame or over an intramedullary
AOFAS scores improved from average scores of nail. Although there is no formal consensus on
29–89. Pain scores also improved from an treating segmental tibial nonunions, many sur-
average preoperative score of 14–36 postopera- geons approach them with treatment guided by
tively [47]. Alternatively, Richmond et al. [48] the size of the defect. Bone loss of less than 2 cm
reported on 32 patients with distal tibial non- can be effectively treated with autologous bone
unions treated with intramedullary nailing and grafting. Defects between 2 and 6 cm may be
noted 91% union rate at an average of treated with large autologous bone grafting, such
3.5 months as long as there was enough room as the Masquelet technique, or bone transport.
distally for two interlocking screws. The Masquelet technique involves a two-stage
Fig. 13.7 28-year-old female with a Gustilo-type IIIB thickness skin graft (d). The patient would later go on to
open tibia and a segmental bone defect. Intramedullary have removal of the spacer with autogenous bone grafting
nailing was performed with insertion of an antibiotic 6 weeks later. (Images courtesy of Garrett A. Wirth, MD,
impregnated cement spacer (red arrow, a, b). Harvesting University of California–Irvine, Orange, California, USA)
of a free gracilis muscle (c). Final coverage with split
procedure starting with radical debridement and technique for bone defects from 1 to 25 cm in
cement spacer placement, which induces an length. At the final clinical follow-up at
osteoinductive membrane, followed by autolo- approximately 1 year postoperatively, 17 of the
gous bone grafting after removal of the cement nonunions had achieved clinical union. Kundu
spacer [50]. This technique can be quite power- et al. reported results of the Huntington’s pro-
ful. Our institution has had success with defects cedure, a tibialization of the fibula, for bone
up to 9 cm in the tibia and even larger in the defects over 6 cm in size in 22 patients. The
femur. Bone defects larger than 6 cm are often authors described clinical unions in 21 of the 22
treated with bone transport, free vascularized patients with full unprotected weight bearing at
fibular transfer or amputation. The Ilizarov bone an average of 16 months [52]. Figure 13.8
transport technique is a useful tool and allows for demonstrates the Masquelet technique in a
bifocal or trifocal correction of large segmental 38-year-old female who presented from an out-
defects. Sala et al. [51] reviewed results from 12 side institution 6 months out with a draining
patients with post-infectious segmental tibial sinus over her anterior tibia. She underwent a
nonunions treated with Ilizarov bone transport staged procedure with irrigation and removal of
(Taylor Spatial Frame) in a bifocal or trifocal all dead and infected bone, placement of an
method and noted 100% union rate in an average antibiotic spacer, and a free gracilis muscle
external fixation time of 418 days (range 300– transfer. She was placed on intravenous antibi-
600 days). otics for 6 weeks. After a 2 week antibiotic
Stafford et al. [49] retrospectively reviewed 19 holiday, infectious laboratory markers were
segmental tibial nonunions treated with the drawn and were normal. She was taken back to
reamer-irrigator/aspirator (RIA) system (Synthes, the operating room for removal of the antibiotic
Paoli, PA, USA) and a two-stage Masquelet spacer and autologous bone grafting using RIA
Fig. 13.8 38-year-old female presented with a infected performed at 8 weeks from ipsalateral femur. Bone
nonunion (a, b). Underwent debridement and placement of grafting of the tibial defect (e). AP and lateral radiographs
antibiotic cement spacer (c, d). Reamer-irrigator/aspirator of tibia at 2 weeks show position of graft (f, g)
on her ipsalateral femur. A robust pseudomem- nailing. The mean time to union, which included
brane was generated and preserved (see clinical and radiographic parameters, averaged
Fig. 13.8d). 27.6 weeks (range 11–59), and all patients
eventually healed.
13.4.1.7 Prior Failed Treatment
Recalcitrant nonunions of the tibia are those that
have undergone multiple operations and fail to 13.4.2 Methods of Treatment
achieve union. The etiology of these resistant
nonunions may be multifactorial and are more Tibial nonunions have a variety of nonsurgical
difficult to treat. The soft tissue envelope and, and surgical treatment modalities. Regarding
therefore, the blood supply, of recalcitrant non- nonsurgical options, surgeons may pursue
unions are often tenuous due to the initial trauma, observation, functional bracing, external electri-
chronic inflammation, and multiple surgical cal or ultrasonic stimulation, or extracorporeal
procedures. When treating these nonunions, it is shock wave therapy [54]. Surgical management
important to consider the type of procedure pre- includes fibular osteotomy, compression plating,
viously performed, as there is little value of intramedullary nailing, external fixation, bone
repeating a properly done surgery that has grafting, and distraction osteogenesis.
already failed. Instead, the surgeon must consider
possibly using a combination of different 13.4.2.1 Functional Bracing
modalities to achieve successful union. Desai In a small subset of tibial nonunions, the lack of
et al. [53] reported on nine recalcitrant tibial clinical healing may be augmented by a
nonunions that had previously undergone at least weight-bearing cast or brace. Sarmiento and
four operations and treated them with the RIA Latta suggested that weight-bearing causes a
bone graft, human recombinant bone mor- change in the fracture site biology that stimulates
phogenic protein 2 (BMP-2), and intramedullary callus formation [55]. This is especially
important with regard to patients who have particular importance in patients who have had
undergone a nonweight-bearing period to main- prior surgeries and those with prior insult to the
tain reduction but were then referred when the surrounding soft tissue. Piriou et al. [57] reported
fracture did not heal promptly. Functional brac- on 18 patients with tibial nonunions treated with
ing is indicated in the setting of stable fracture decortication and medial plating and achieved
patterns, acceptable alignment, and a compliant approximately 94% union rate at an average of
patient. Sarmiento et al. also reported on 67 108 days. Helfet et al. [58] presented 33 patients
delayed and nonunited tibial fractures that were with tibial nonunions treated with tension-band
treated with functional bracing alone in nine plating, augmented with lag screw fixation and
patients, with fibular ostectomy in 48 patients, autogenous bone grafting, and reported 100%
and with fibular ostectomy with bone grafting in healing rate at an average of 4 months postop-
ten patients. Of the 67, 91% went on to union at eratively. Lastly, Wiss et al. [59] found a 92%
an average of 3–4 months. In the group treated healing rate at 7 months with compression plat-
exclusively with the functional brace, only one of ing augmented by bone grafting on his 50
nine failed to heal [56]. Overall this method tends patients. Compression plating remains a valuable
to harbor low risk for the patient and allows for tool in the armamentarium of any orthopedic
future treatment options should they be surgeon when treating tibial nonunions.
necessary. Intramedullary Nailing. Intramedullary
nailing of tibial nonunions provides an ideal
13.4.2.2 Mechanical combination of augmenting mechanical stability,
Compression Plating. Achieving compression dispersing of growth factors, and allows early
across the nonunion site is paramount for proper active rehabilitation. Indications for this proce-
healing, and the most basic tool to achieve it is dure continue to broaden and are similar to the
the compression plate. Compression plating of indications for acute fracture treatment, including
tibial fractures continues to be a widely used prior closed and open diaphyseal injuries, seg-
technique for treating tibial nonunions, as it mental nonunions, and nonunions with signifi-
allows for immediate correction of any defor- cant bone loss.
mity, debridement of the nonunion, placement of With exchange nailing, reaming the medullary
bone graft, and direct compression. Compression canal can promote angiogenesis and allow for a
with a plate can be achieved in many ways, larger, more rigid, nail to be placed. This is
including using an articulated tensioning device, especially important in hypertrophic nonunions
compressing with a verbrugge clamp against a where mechanical stability is lacking. Johnson
screw placed outside the plate, and even a and Marder discussed 20 patients, of whom 11
two-screw technique with a farbeuf clamp. Cau- had hypertrophic nonunions, treated with intra-
tion should be used when relying on the plate medullary nailing and found a healing rate of
compression holes to achieve compression as 100% with callus formation at an average of
they do not generally achieve enough compres- 11 weeks in the hypertrophic patients alone [60].
sion for nonunion repair. Bone grafting is also Compression with a nail can be achieved with
possible to augment this technique and the direct “back slapping” or reverse impaction, utilizing
exposure to the nonunion site allows for simul- the compression screw in many nail designs and
taneous formal debridement. Postoperatively, the by early weight bearing depending on the
patient is able to begin early mobilization and amount of cortical contact in the repaired non-
rehabilitation. Figure 13.9 shows an atrophic union. Clinical union rates range from 76 to
nonunion in a patient with a history of neurofi- 100% [61–63]. Zelle et al. retrospectively
bromatosis. Direct compression plating and bone reviewed the results from 38 aseptic nonunions
grafting were performed. As with all open tech- after initial unreamed nailing that underwent
niques, plate osteosynthesis presents increased exchange reamed intramedullary nailing. The
risk of wound complications, which is of clinical union rate, as defined by painless weight
Fig. 13.9 41-year-old male with a history of neurofi- with bone morphogenic protein (BMP)-2. Eventually
bromatosis and a 1 year history of aseptic nonunion (a, healed 2 months after surgery (c, d)
b). Underwent compression plating and bone grafting
bearing and bridging callus on three cortices, was cases of tibial nonunions treated with
95% at an average time of 29 weeks with only mono-lateral external fixation and either
three complications noted, one deep vein corticotomy/bone transport for cases with >3 cm
thrombosis and two hardware failures [64]. of shortening or compression/distraction for
External Fixation. External fixation in the cases with <3 cm of shortening. These authors
setting of tibial nonunions is a useful tool and is reported at an average of 8 months follow-up
used primarily to manage infections and defor- approximately 96.7% of cases achieved union in
mity in patients with a poor soft tissue envelope. the tibial group alone [65]. Menon et al. [66]
However, external fixation can be used for the used circular fixators to achieve compression in
same indications as intramedullary nailing and nine tibial nonunion patients and retrospectively
compression plating as well. Ring fixators also found 100% union rate at an average of
have the advantage of immediate postoperative 6.2 months. García-Cimbrelo and Martí-Gonzá-
weight bearing. Harshwal et al. reported on 30 lez [67] presented 82 patients treated with
Fig. 13.10 12-year-old male with a type IIIB open tibia removed (d). A ring external fixator was eventually
fracture (a). Initially treated with a free muscle transfer placed (e) and the patient went on to heal his infected
and open reduction internal fixation (b). Developed a deep nonunion (f)
infection (c), which was debrided, and the implants were
circular external fixators for patients with tibial used in conjunction with compression plating,
nonunions with and without bone loss and noted exchange nailing, or external fixation in order to
an overall 93% healing rate. External fixation allow the force of weight bearing to be trans-
remains a powerful technique for the correction mitted preferentially through the tibia.
of tibial nonunions with many different charac-
teristics, e.g., infected, bone loss, angular defor- 13.4.2.3 Biological
mity, etc. Figure 13.10 demonstrates the Bone Grafting. Iliac crest autograft is considered
usefulness of a ringed external fixator for man- the gold standard for bone grafting of tibial
agement of tibial nonunions. The patient had an nonunions across a variety of locations given its
infected nonunion with a free muscle transfer. osteoconductive, osteoinductive, and osteogenic
Fibular Osteotomy. Historically, fibular properties. The RIA has made possible the har-
osteotomy has been used to increase the load vesting of large quantities of cancellous bone
transmitted through the tibial fracture site. The from the endosteal surface of the femur. This
theory is based on the intact or healed fibula technique is appropriate for nonunions with up to
acting as a distracting strut that prevents tibia 6 cm of bone loss and has a reported success rate
fracture apposition. The procedure is oftentimes of 88–95%.
Surgical approaches to the nonunion site Administration solely under the “humanitarian
include posterolateral or anterolateral, both of device exemption” [70]. Friedlaender et al. [71]
which provide adequate soft tissue coverage for reported the results of a prospective clinical trial
graft incorporation. Takemoto et al. [49] per- of tibial nonunions treated with intramedullary
formed analysis of different bone graft sites, nailing with autologous bone graft or BMP-7 and
including iliac crest, proximal humerus, and found similar rates of union at 9 and 24 months.
proximal tibia, for various BMPs and their The authors concluded that BMP-7 was at least
receptors and found no significant difference at as effective as autologous bone grafts in the
any of the three sites for all variables [68]. setting of tibial nonunions treated with intrame-
Bone Graft Substitutes. Bone graft substi- dullary nailing. Calori et al. performed retro-
tutes are widely available today and are osteo- spective cost analysis of 54 patients treated with
conductive and osteoinductive in nature. They are BMP-7 or autologous bone grafting and noted
designed to avoid harvest site morbidity; how- clinical union in 89.3 and 76.9%, respectively,
ever, they are relatively expensive and have not P = 0.22. The authors noted significantly higher
been proven as clinically effective as cancellous costs associated with BMP-7 treatment and total
bone grafting. Demineralized bone matrix costs incurred during BMP-7 usage [72].
(DBM) consists of allograft derived bone graft Dehabreh et al. performed economic analysis of
with osteoconductive properties. While DBM nonunion treatment across various anatomic
may hold osteoconductive properties by provid- locations prior to and after application of BMP-7
ing a scaffold for bony ingrowth, its osteoinduc- to augment fracture healing. The authors con-
tive characteristics have been inconsistent. DBM cluded that BMP-7 resulted in a 47% reduction in
is available as putties, pastes, gels, and granules costs compared to persistent fracture nonunions
from a variety of different companies and is best treated otherwise [73]. BMP is not routinely used
used as a graft extender in conjunction with as a first-line treatment for nonunions at our
autograft bone. Ceramic bone substitutes include institution; currently, BMP is reserved for recal-
calcium sulfate and calcium phosphate, which citrant nonunions that have failed previous
similarly have osteoconductive properties. attempts at union and for certain patients who are
Both DBM and ceramics have been used to fill not candidates for autogenous bone grafting.
defects after nonunion formation. McKee et al. Further research is necessary to identify the
utilized medical grade alpha-hemihydrate cal- patient population that will benefit the most from
cium sulfate impregnated with tobramycin this costly adjunctive procedure.
(Osteoset-T, Wright Medical Technology,
Arlington, TN, USA) pellets to address voids in 13.4.2.4 Adjunct Treatments
25 patients with femoral, tibial, ulna, or humeral Distraction Osteogenesis. Modern distraction
osteomyelitis. Each patient was treated with the osteogenesis is based on the principles set forth
pellets in addition to external and/or internal fix- by Gavriil Ilizarov. Ilizarov is credited for
ation, and the authors achieved 92% rate of inventing a circular-type external fixator that
infection eradication. By 6 months, postopera- allowed correction of deformity with six degrees
tively all pellets were noted to be resorbed and of freedom [74]. His external fixation device or
71% of those patients treated with bone substitute Ilizarov frame became the basis for bone regen-
without autograft achieved union [69]. eration from tension stress applied to a cortico-
Growth Factors. Bone morphogenic pro- tomy site. With distraction osteogenesis,
teins, specifically, recombinant human BMP-2 nonunions undergo increased perfusion, inflam-
and BMP-7, have been utilized to augment tibial matory response, and gain the ability to ward off
healing in a variety of studies. rhBMP-7, also infection (Fig. 13.11). Ilizarov argued treating
known as osteogenic protein-1, has been studied infected nonunions with his method as “infection
specifically for tibial nonunions and has been burns in the fire of regeneration.” Ring et al.
approved by the US Food and Drug retrospectively reviewed 27 patients comparing
Fig. 13.11 33-year-old-male status post falls from horse. utilized (e, f) until final healing (g, h). Images courtesy of
Patient suffered a Gustilo-type IIIB open fracture. Animesh Agarwal, MD Animesh Agarwal, MD, Depart-
Presented with an infected nonunion (a, b). Debridement ment of Orthopedics, University of Texas Health Science
and a monolateral external fixator wa placed along with Center, San Antonio, Texas, USA
proximal corticotomy (c, d). Distraction osteogenesis was
autologous bone grafting with the aforemen- injury. All patients were cured of the infection.
tioned Ilizarov technique. All patients in the bone As to results achieved, 14 were considered
grafting group went on to heal and nine of the ten excellent, eight good, one fair, and five poor [76].
patients treated with the Ilizarov technique went Structural Grafts. In the treatment of tibial
on to clinical healing. The authors concluded that nonunions, the Huntington procedure is
the Ilizarov technique was best suited for proxi- well-described technique for tibial defects greater
mal or distal metaphyseal nonunions, and for than 12 cm. The first successful such procedure
those patients with large bone defects [75]. performed by Huntington in 1903 described
Dendrinos et al. reviewed 28 cases of infected ipsilateral transfer of the fibula to the tibia in a
tibial nonunions treated with distraction osteo- young boy [77]. Kassab et al. retrospectively
genesis at a mean of 16 months after the original reviewed the results of 11 tibial nonunions
treated with the Huntington procedure and LIPUS generates high-frequency sound waves
observed healing in eight of 11 patients at an that induce mechanical pressure on bone to
average of 10.5 months. Of the three persistent augment the healing process. These ultrasonic
nonunions, one patient had a 22 cm defect after waves have been shown in animal models to
tumor removal and the other two developed increase callus formation and accelerate bone
infections that led to subsequent amputations healing. Randomized control studies have been
[78]. performed in acute fractures of the tibia, but they
are lacking in the setting of established non-
13.4.2.5 Electrical unions [82]. Of those cohort studies performed
Electrical techniques include a variety of appli- without controls, LIPUS union rate ranged from
cations from invasive to noninvasive that all 81 to 100% with the caveat that some tibial
provide electromagnetic field stimulation to nonunions may spontaneously heal without any
induce fracture healing. The exact mechanism of intervention [83–89]. Given the current lack of
augmenting bone healing is not completely good quality randomized studies, Watanabe et al.
understood but proposed theories include [89] concluded that while there are many
increased mineralization and angiogenesis, potential benefits of LIPUS, including lack of
increased DNA synthesis, and changes in osteo- harmful effects, ease of use, and possible use in
blast intracellular calcium levels [79]. These compromised patients, further research is neces-
electrical treatments are applied with concurrent sary to fully understand the indications and
bony stabilization in the form of splinting or benefits of LIPUS.
external fixation. Electromagnetic stimulation Extracorporeal shock wave therapy is thought
may be considered especially useful in those to increase bone mass, augment angiogenesis, and
patients who are poor surgical candidates sec- enhance mesenchymal stem cells differentiation
ondary to medical comorbidities or local soft to osteogenic cells lines. Elster et al. retrospec-
tissue pathology that may not heal the postsur- tively reviewed 192 tibial nonunions treated with
gical incision. Anglen [80] noted that electro- ECSW therapy and found 80.2% healing rate in
magnetic techniques are not effective and an average of 4.8 months. The authors concluded
contraindicated in the setting of fracture gaps that ESCW therapy was indeed a safe and reliable
greater than half the diameter of bone, synovial treatment modality for tibial nonunions to be used
pseudarthrosis and unacceptable deformity. in conjunction with fracture immobilization and
Prospective randomized clinical trials demon- stabilization [90]. Further prospective research
strating efficacy of these modalities is limited despite into the effects of ECSW therapy is necessary to
a body of literature supporting the technique in vitro. make this adjunctive treatment option part of the
A comprehensive review performed by Griffen et al. standard of care for tibial nonunion management.
[81] as part of the Cochrane Library concluded that
the current body of research is inconclusive and
insufficient to determine clinical practice. Addi- References
tionally, there is no ability to correct rotation,
angulation, or limb-shortening deformities. 1. Trafton PG. Tibial shaft fractures. In: Browner BD,
Jupiter JB, Levine AM, Trafton PG, editors. Skeletal
trauma: basic science, management, and reconstruc-
13.4.2.6 Ultrasound tion. 4th ed. Philadelphia: WB Saunders; 2009.
Ultrasound techniques use high-frequency sound 2. Praemer A, Furner S, Rice DP. Musculoskeletal
waves and include low-intensity pulsed ultra- conditions in the United States. Park Ridge, IL:
sound (LIPUS), high-intensity focused ultra- American Academy of Orthopedic Surgeons; 1992.
3. AHRQ (Agency for Health Care Research and Quality).
sound (HIFUS) and extracorporeal shockwave Introduction to the healthcare cost and utilization project
therapies (ECSW) . All methods rely on bone (HCUP) state inpatient databases (SID). http://www.
growth in response to mechanical forces across hcup-us. ahrq.gov/db/state/siddist/Introduction_to_SID.
the fracture site to encourage fracture union. pdf. Accessed 22 April 2016.
4. Schmidt AH, Finkemeier CG, Tornetta P 3rd. Treat- hybrid external fixation of distal tibial fractures: a
ment of closed tibial fractures. Instr Course Lect. review of 47 cases. J Trauma. 2007;62(1):174–83.
2003;52:607–22. 21. Sloan A, Hussain I, Maqsood M, Eremin O,
5. Brinker MR, O’Connor DP. Nonunions: evaluation El-Sheemy M. The effects of smoking on fracture
and treatment. In: Browner BD, Jupiter JB, healing. Surgeon. 2010;8(2):111–6.
Levine AM, Trafton PG, editors. Skeletal trauma: 22. Mosely LH, Finseth F. Cigarette smoking: impair-
basic science, management, and reconstruction. 4th ment of digital blood flow and wound healing in the
ed. Philadelphia: W.B. Saunders; 2009. p. 615–708. hand. Hand. 1977;9(2):97–101.
6. Court-Brown CM, Keating JF, Christie J, 23. Donigan JA, Fredericks DC, Nepola JV,
McQueen MM. Exchange intramedullary nailing. Smucker JD. The effect of transdermal nicotine on
Its use in aseptic tibial nonunion. J Bone Joint Surg fracture healing in a rabbit model. J Orthop Trauma.
Br. 1995;77(3):407–11. 2012;26(12):724–7.
7. Canadian Institute for Health information (CIHI). 24. Raikin SM, Landsman JC, Alexander VA, Froim-
National trauma registry: hospital injury admissions. son MI, Plaxton NA. Effect of nicotine on the rate of
Ottawa: Canadian Institute for Health Information; 2003. strength of long bone fracture healing. Clin Orthop
8. Heckman JD, Sarasohn-Kahn J. The economics of Relat Res. 1998;353:231–7.
treating tibia fractures: the cost of delayed unions. 25. Castillo RC, Bosse MJ, MacKenzie EJ. Patterson
Bull Hosp Jt Dis. 1997;56(1):63–72. BM; LEAP Study Group. Impact of smoking on
9. Fong K, Truong V, Foote CJ, Petrisor B, Williams D, fracture healing and risk of complications in
Ristevski B, et al. Predictors of nonunion and limb-threatening open tibia fractures. J Orthop
reoperation in patients with fractures of the tibia: an Trauma. 2005;19(3):151–7.
observational study. BMC Musculoskeletal Disord. 26. Adams CI, Keating JF, Court-Brown CM. Cigarette
2013;14:103. smoking and open tibial fractures. Injury. 2001;32
10. Teraa M, Blokhuis TJ, Tang L, Leenen LPH. Seg- (1):61–5.
mental tibial fractures: an infrequent but demanding 27. Bo J, Sudmann E, Marton PF. Effect of indomethacin
injury. Clin Orthop Relat Res. 2013;471(9):2790–6. on fracture healing in rats. Acta Orthop Scand.
11. Nicoll EA. Fractures of the tibial shaft: a survey of 1976;47(6):588–99.
705 cases. J Bone Joint Surg Br 1964;46-B:373–87. 28. Murnaghan M, Li G, Marsh DR. Nonsteroidal
12. Yang JS, Otero J, McAndrew CM, Ricci WM, anti-inflammatory drug-induced fracture nonunion:
Gardner MJ. Can tibial nonunion be predicted at an inhibition of angiogenesis? J Bone Joint Surg Am.
three months after intramedullary nailing? J Orthop 2006;88(Suppl 3):140–7.
Trauma. 2013;27(11):599–606. 29. Karachalios T, Boursinos L, Poultsides L, Khaldi L,
13. Alt V, Donell ST, Chhabra A, Bentley A, Eicher A, Malizos KN. The effects of the short-term adminis-
Schnettler R. A health economic analysis of the use tration of low therapeutic doses of anti-COX-2 agents
of rhBMP-2 in Gustilo-Anderson grade III open tibial on the healing of fractures: an experimental study in
fractures for the UK, Germany, and France. Injury. rabbits. J Bone Joint Surg Br. 2007;89(9):1253–60.
2009;40(12):1269–75. 30. Zhang X, Schwarz EM, Young DA, Puzas JE,
14. Hak DJ, Saleh K. Socioeconomic burden of trau- Rosier RN. O’Keefe. Cyclooxygenase-2 regulates
matic tibial fractures: nonunion or delayed union. mesenchymal cell differentiation into the osteoblast
Mescape; 2001. http://www.medscape.org/ lineage and is critically involved in bone repair.
viewarticle/418523. Accessed 22 April 2016. J Clin Invest. 2002;109(11):1405–15.
15. Brinker MR, Hanus BD, Sen M, O’Connor DP. The 31. Simon AM, O’Connor JP. Dose and time-dependent
devastating effects of tibial nonunion on health effects of cyclooxygenase-2 inhibition on fracture-
related quality of life. J Bone Joint Surg Am. healing. J Bone Joint Surg Am. 2007;89(3):500–11.
2013;95(24):2170–6. 32. Giannoudis PV, MacDonald DA, Matthews SJ,
16. Antonova E, Le TK, Burge R, Mershon J. Tibia shaft Smith RM, Furlong AJ, De Boer P. Nonunion of
fractures: costly burden of nonunions. BMC Muscu- the femoral diaphysis. J Bone Joint Surg Br. 2000;82
loskeletal Disorders. 2013;14:42. (5):655–8.
17. Bishop JA, Palanca AA, Bellino MJ, Lowenberg DQ. 33. Brinker MR, O’Connor DP, Monla YT, Earth-
Assessment of compromised fracture healing. J Am man TP. Metabolic and endocrine abnormalities in
Acad Orthop Surg. 2012;20:273–82. patients with nonunions. J Orthop Trauma. 2007;21
18. Rosenthal RE, MacPhail JA, Ortiz JE. Non-union in (8):557–70.
open tibial fractures. J Bone Joint Surg. 1977;59 34. Patzakis MJ, Wilkins J, Kumar J, Holtom P, Green-
(2):244–8. baum B, Ressler R. Comparison of the results of
19. Hernigou J, Schuind F. Smoking as a predictor of bacterial cultures from multiple sites in chronic
negative outcome in diaphyseal fracture healing. Int osteomyelitis of long bones: a prospective study.
Ortho. 2013;37(5):883–7. J Bone Joint Surg Am. 1994;76(5):664–6.
20. Ristiniemi J, Flinkkilä T, Hyvönen P, Lakovaara M, 35. Bhattacharyya T, Bouchard KA, Phadke A,
Pakarinen H, Biancari F, Jalovaara P. Two-ring Meigs JB, Kassarjian A, Salamipour H. The accuracy
of computed tomography for the diagnosis of tibial defects with a combined Ilizarov/Taylor spatial frame
nonunion. J Bone Joint Surg Am. 2006;88(4):692–7. technique. J Orthop Trauma. 2011;25(3):162–8.
36. Madsen J. Bone SPECT/CT detection of a seques- 52. Kundu ZS, Gupta V, Sangwan SS, Kamboj P. Gap
trum in chronic-infected nonunion of the tibia. Clin nonunion of tibia treated by Huntington’s procedure.
Nucl Med. 2008;33(10):700–1. Indian J Orthop. 2012;16(6):653–8.
37. Matuszewski PE, Mehta S. Fracture consolidation in 53. Desai PP1, Bell AJ, Suk M. Treatment of recalcitrant,
a tibial nonunion after revascularization: a case multiply operated tibial nonunions with the RIA graft
report. J Orthop Trauma. 2011;25(2):e15–20. and rh-BMP2 using intramedullary nails. Injury.
38. Brownlow HC, Simpson AH, Kenwright JK. The 2010;41(Suppl 2):S69–71.
vascularity of atrophic non-union gaps. J Bone Joint 54. Anglen J. Update on the management of nonunion.
Surg [Br]. 1999;81-B(Suppl. III):331. American Academy of Orthopaedic Surgeons Web
39. Patzakis MJ, Zalavras CG. Chronic posttraumatic Site: Orthopaedic Knowledge Online 2012;10(7).
osteomyelitis and infected nonunion of the tibia: http://orthoportal.aaos.org/oko/article.aspx?article=
current management concepts. J Am Acad Orthop OKO_TRA039#article. Accessed 30 Mar 2016.
Surg. 2005;13(6):417–27. 55. Sarmiento A, Latta LL. Closed functional treatment
40. Cierny G 3rd, Mader JT, Penninck JJ. A clinical of fractures. Berlin: Springer; 1981. p. 1–58.
staging system for adult osteomyelitis. Contemp 56. Sarmiento A, Burkhalter WE, Latta LL. Functional
Orthop. 1985;10:17–37. bracing in the treatment of delayed union and
41. Erdman WA, Tamburro F, Jayson HT, Weatherall PT, nonunion of the tibia. Int Ortho. 2003;27(1):26–9.
Ferry KB, Peshock RM. Osteomyelitis: characteris- 57. Piriou P, Martin JR, de Loubresse CG, Judet C.
tics and pitfalls of diagnosis with MR imaging. Tibial nonunion after intramedullary nailing for
Radiology. 1991;180(2):533–9. fracture: decortication and osteosynthesis by medial
42. Perry CR, Pearson RL, Miller GA. Accuracy of plating. Rev Chir Orthop Reparatrice Appar Mot.
cultures of material from swabbing of the superficial 2005;91(3):222–31.
aspect of the wound and needle biopsy in the 58. Helfet DL, Jupiter JB, Gasser S. Indirect reduction
preoperative assessment of osteomyelitis. J Bone and tension-band plating of tibial non-union with
Joint Surg. 1991;73(5):745–9. deformity. J Bone Joint Surg Am. 1992;74(9):1286–
43. Megas P, Saridis A, Kouzelis A, Kallivokas A, 97.
Mylonas S, Tyllianakis M. The treatment of infected 59. Wiss D, Johnson DL, Miao M. Compression plating
nonunion of the tibia following intramedullary nail- for non-union after failed external fixation of open
ing by the Ilizarov method. Injury. 2010;41(3):294– tibial fractures. J Bone Joint Surg Am. 1992;74
9. (9):1279–85.
44. Holtom PD, Patzakis MJ. Newer methods of antimi- 60. Johnson EE, Marder RA. Open intramedullary
crobial delivery for bone and joint infections. Instr nailing and bone-grafting for non-union of tibial
Course Lect. 2003;52:745–9. diaphyseal fracture. J Bone Joint Surg Am. 1987;69
45. Harvey EJ, Henley MB, Siontkowski MF, Agel J, (3):375–80.
Benirschke SK. The use of a locking custom 61. Brinker MR, O’Connor DP. Exchange nailing of
contoured blade plate for peri-articular nonunions. ununited fractures. J Bone Joint Surg Am. 2007;89
Injury. 2003;34(2):111–6. (1):177–88.
46. Gardner MJ, Toro-Arbelaez JB, Boraiah S, 62. Hsiao CW, Wu CC, Su CY, Fan KF, Tseng IC,
Lorich DG, Helfet DL. Surgical treatment and Lee PC. Exchange nailing for aseptic tibial shaft
outcomes of extraarticular proximal tibial nonunions. nonunion: emphasis on the influence of a concomi-
Arch Orthop Trauma Surg. 2008;128(8):833–9. tant fibulotomy. Chang Gung Med J. 2006;29
47. Reed LK, Mormino MA. Functional outcome after (3):283–90.
blade plate reconstruction of distal tibia metaphyseal 63. Court-Brown CM, Keating JF, Christie J,
nonunions. J Orthop Trauma. 2004;18(2):81–6. McQueen MM. Exchange intramedullary nailing.
48. Richmond J, Colleran K, Borens O, Kloen P, Its use in aseptic tibial nonunion. J Bone Joint Surg
Helfet DL. Nonunions of the distal tibia treated with Br. 1995;77(3):407–11.
reamed intramedullary nailing. J Orthop Trauma. 64. Zelle BA, Gruen GS, Klatt B, Haemmerle MF,
2004;18(9):603–10. Rosenblum WJ, Prayson MJ. Exchange reamed
49. Stafford PR, Norris BL. Reamer-irrigator-aspirator nailing for aseptic nonunion of the tibia. J Trauma.
bone graft and bi Masquelet technique for segmental 57(5):1053–9.
bone defect nonunions: a review of 25 cases. Injury. 65. Harshwal RK, Sankhala SS, Jalan D. Management of
2010;41(Suppl 2):S72–7. nonunion of lower-extremity long bones using
50. Masquelet AC, Begue T. The concept of induced mono-lateral external fixator— report of 37 cases.
membrane for reconstruction of long bone defects. Injury. 2014;45(3):560–7.
Orthop Clin North Am. 2010;41(1):27–37. 66. Menon DK, Dougall TW, Pool RD, Simonis RB.
51. Sala F, Thabet AM, Castelli F, Miller AN, Capi- Augmentative Ilizarov external fixation after failure
tani D, Lovisetti G, Talamonti T, Singh S. Bone of diaphyseal union with intramedullary nailing.
transport for postinfectious segmental tibial bone J Orthop Trauma. 2002;16(7):491–7.
67. García-Cimbrelo E, Martí-González JC. Circular 79. Bassett CA. Fundamental and practical aspects of
external fixation in tibial nonunions. Clin therapeutic uses of pulsed electromagnetic fields
Orthop. 2004;419:65–70. (PEMFs). Crit Rev Biomed Eng. 1989;17(5):451–
68. Takemoto RC, Fajardo M, Kirsch T, Egol KA. 529.
Quantitative assessment of the bone morphogenetic 80. Anglen J. The clinical use of bone stimulators.
protein expression from alternate bone graft harvest- J South Orthop Assoc. 2003;12(2):46–54.
ing sites. J Orthop Trauma. 2010;24(9):564–6. 81. Griffen XL, Costa ML, Parsons N, Smith N. Elec-
69. McKee MD, Wild LM, Schemitsch EH, Wad- tromagnetic field stimulation for treating delayed
dell JP. The use of an antibiotic-impregnated osteo- union or non-union of long bone fractures in adults.
conductive bioabsorbable bone substitute in the Cochrane Database Syst Rev. 2011 Apr 13;(4):
treatment of infected long bone defects: early results CD008471.
of a prospective trial. J Orthop Trauma. 2002;16 82. Heckman JD, Ryaby JP, McCabe J, Frey JJ, Kil-
(9):622–7. coyne RF. Acceleration of tibial fracture-healing by
70. Schultz, DG. HP19992 OP-1 (TM) Implant. Letter to non-invasive, low-intensity pulsed ultrasound.
LaForte AJ. Office of Device Evaluation, Food and J Bone Joint Surg Am. 1994;76(1):26–34.
Drug Administration, Department of Health & 83. Mayr E, Frankel V, Rüter A. Ultrasound—an
Human Services, Rockville, MD. 17 Oct 2001. alternative healing method for nonunions? Arch
http://www.accessdata.fda.gov/cdrh_docs/pdf/ Orthop Trauma Surg. 2000;120(1–2):1–8.
h010002a.pdf. Accessed 29 Mar 2016. 84. Rubin C, Bolander M, Ryaby JP, Hadjiargyrou M.
71. Friedlaender GE, Perry CR, Cole JD, Cook SD, The use of low-intensity ultrasound to accelerate the
Cierny G, Muschler GF, et al. Osteogenic protein-1 healing of fractures. J Bone Joint Surg Am.
(bone morphogenetic protein-7) in the treatment of 2001;83-A(2):259–70.
tibial nonunions. J Bone Joint Surg Am. 2001;83-A 85. Mizuno K, Yamano Y, Itoman M, et al. Effects of
(Suppl 1Pt 2):S151–8. low-intensity pulsed ultrasound therapy for delayed
72. Calori MG, Capanna R, Colombo M, Biase PD, unions and nonunions: a multi-center clinical study.
O’Sullivan C, Cartareggia V, Conti C. Cost effec- Orthopaedic Surg Traumatol. 2003;46:757–65.
tiveness of tibial nonunion treatment: a comparison 86. Nolte PA, van der Krans A, Patka P, Janssen IM,
between rhBMP-7 and autologous bone graft in two Ryaby JP, Albers GH. Low-intensity pulsed ultra-
Italian centers. Injury. 2013;44(12):1871–9. sound in the treatment of nonunions. J Trauma.
73. Dahabreh Z, Dimitriou R, Giannoudis PV. Health 2001;51(4):693–702.
economics: a cost analysis of treatment of persistent 87. Pigozzi F, Moneta MR, Giombini A, Giannini S, Di
fracture nonunions using bone morphogenetic Cesare A, Fagnani F, Mariani PP. Low-intensity
protein-7. Injury. 2007;38(3):371–7. pulsed ultrasound in the conservative treatment of
74. Kanellopoulos AD, Soucacos P. Management of pseudoarthrosis. J Sports Med Phys Fitness. 2004;44
nonunion with distraction osteogenesis. Injury. (2):173–8.
2006;37(Suppl 1):S51–5. 88. Rutten S, Nolte PA, Guit GL, Bouman DE,
75. Ring D, Jupiter JB, Gan BS, Israeli R, Yarem- Albers GH. Use of low-intensity pulsed ultrasound
chuk MJ. Infected nonunion of the tibia. Clin for posttraumatic nonunions of the tibia: a review of
Orthop. 1999;369:302–11. patients treated in the Netherlands. J Trauma.
76. Dendrinos GK, Kontos S, Lyritsis E. Use of the 2007;62(4):902–8.
Ilizarov technique for treatment of non-union of the 89. Watanabe Y, Matsushita T, Bhandari M, Zdero R,
tibia associated with infection. J Bone Joint Surg Schemitsch EH. Ultrasound for fracture healing:
Am. 1995;77(6):835–46. current evidence. J Ortho Trauma. 2010;24(Suppl 1):
77. Huntington TWVI. Case of bone transference: use of S56–61.
a segment of fibula to supply a defect in the tibia. 90. Elster EA, Stojadinovic A, Forsberg J, Shawen S,
Ann Surg. 1905;41(2):249–51. Andersen RC, Schaden W. Extracorporeal shock
78. Kassab M, Samaha C, Saillant G. Ipsilateral fibular wave therapy for nonunion of the tibia. J Orthop
transposition in tibial nonunion using Huntington’s Trauma. 2010;24(3):133–41.
procedure: a 12 year follow up study. Injury.
2003;34(10):770–5.
Distal Tibia and Ankle Nonunions
14
Kevin J. Pugh, MD
involving several stages, is devised by following

14.1 Introduction
a set of principles as illustrated throughout this
book. The principles of treating nonunions about
Normal gait is dependent on normal function of
the ankle will be illustrated in this chapter.
the ankle joint [1, 2]. Disruption of the ankle
Injury to bone, whether it be due to a fracture,
mortise by fracture of the plafond, malleoli, or
osteotomy, or attempted arthrodesis, initiates a
syndesmosis can lead to ankle instability or lack
predictable series of events that under normal
of articular congruity, both of which can lead to
circumstances should proceed to a union.
degenerative arthritis. Anatomic reduction and
Although a full discussion of the physiology of
stable fixation of fractures about the ankle is
bone healing can be found in Chap. 2, to briefly
required to restore function and to obtain union.
summarize, the initial inflammatory process leads
Nonunions of the ankle lead to stiffness, dys-
to cell recruitment and differentiation, the laying
functional gait, prolonged morbidity, and pro-
down of new bone matrix, and eventually to
longed social stress on the patient. In addition,
ossification, boney union, and remodeling of the
malreduced nonunions affect the mechanical axis
immature construct. However, not all fractures
of the limb, leading to dysfunction of the adja-
are created equally. They differ in terms of the
cent knee and subtalar joint.
energy imparted, the size of the zone of injury,
whether they are open or closed, the fracture
pattern and comminution, the bone involved as
14.2 Analysis of the Fracture
well as the region on that bone, and the presence
of articular involvement. Injuries to the ankle
There is no one method of treatment for a non-
involve more than just the boney structures. The
union. Instead, it is an analysis and correction of
bone around the ankle is generally subcutaneous
the cause of the nonunion and the unique features
and often has poor soft tissue coverage. Thus,
of each specific case. The fact that there is no one
due to these characteristics, fractures about the
technique or implant to treat nonunions is what
ankle, unfortunately, often have significant
makes them both challenging and satisfying to
associated soft tissue injuries. Perhaps the most
treat. The treatment plan, whether simple or
important variable, and the least predictable, is
the host. The medical and social history of the
patient and his compliance can have a major
effect on the success or failure of the treatment
K.J. Pugh (&) plan. A successful outcome requires a balance
Orthopedic Trauma and Reconstructive Surgery,
between the stability of the fracture treatment
Grant Medical Center, 285 E. State St. #500,
Columbus, OH 43215, USA construct and the biological viability of the
e-mail: kjportho@mac.com fracture site.

310 K.J. Pugh, MD
Unlike muscle, cartilage, and other connective defines a nonunion as being “established when a
tissue, bone does not heal with scar tissue. It is minimum of nine months has elapsed since
one of the few tissues in the body that heals with injury and the fracture site shows no visibly
the same material—bone. Fractures, osteotomies, progressive signs of healing for a minimum of
and arthrodesis constructs are thus repaired by a three months” [4]. This definition is not prag-
process of boney regeneration. A boney union is matic and leads to prolonged morbidity, long
one that is repaired to the degree that it is periods of work-related impairment or socioe-
mechanically able to function like denovo bone. conomic stress, and the potential for narcotic
The patient experiences no pain, and there is abuse. As surgeons, we know that there are
clinical stability at the fracture site. Clinical injuries that are at risk for nonunion based on the
fracture unions are accompanied by radiographic injury (open fracture, comminution, or bone
signs of healing. In order to confirm a true union, loss), anatomic location (distal tibia and 5th
both radiographic and clinical signs should be metatarsal), or host (diabetic, smoker, and cancer
present. A delayed union is a fracture that, patient). It has thus become more acceptable to
though making continual progress toward union, label fractures as delayed or nonunions when the
has not healed in the usual amount of time for a surgeon believes the fracture has little or no
similar fracture. A nonunion is a fracture that will ability to heal. Delaying intervention for an
not heal. It has sustained an arrest of the repair arbitrary length of time before calling a fracture a
process and has not shown radiographic or clin- nonunion results in more disability, more time off
ical progress toward healing for months. Non- work, and greater psychological stress for the
unions may have some clinical stability, as they patient. As soon as slow healing is identified,
will have cartilage or fibrous interposition instead there should be a frank discussion with the
of bone. Others will be atrophic, with little patient about the possibility of nonunion and the
healing tissue, and have no clinical stability. need for further future treatment. Most patients
Though nonunions cannot be predicted, some will opt for early intervention if it means an
fractures are destined to go on to nonunion from earlier return to work or recreational activities.
the beginning of treatment. The causes of nonunion are multiple, and if
identified should be addressed during treatment
[4–8]. An inappropriate fracture treatment con-
14.3 Determination of Delayed struct, whether unstable or too stiff, malposi-
or True Nonunion tioned or distracted, will lead to poor results. The
injury itself may result in comminution, bone
The first issue in treatment is determining whe- loss, soft tissue injury, or stripping and avascu-
ther the fracture is merely delayed or a true larity. Infection can result in bone death with
nonunion [3]. A delayed union may go on to a formation of a sequestrum, as well as osteolysis
successful outcome if given more time, while a with loosening of implants and eventual insta-
true nonunion will require intervention to bility at the fracture site. Host factors such as age,
achieve union. This is not a trivial question to nutritional status, metabolic abnormalities [9],
answer for the patient. Though most nonunions chronic disease, medicines, and smoking all play
will be diagnosed if the surgeon waits long a role. In some patients, the cause is not identi-
enough, it is imperative to identify fractures that fiable, and thus idiopathic.
are falling behind as soon as possible in order to
shorten overall treatment time and to restore the
patient back to full function. Government payers 14.4 Classification of Nonunions
and many private insurance companies subscribe
to clinical guidelines that incorporate a time Unlike acute fractures, there is no single defini-
factor into the definition of a nonunion. The tive classification system for nonunions. Non-
United States Food and Drug Administration unions can be classified on the basis of their
14 Distal Tibia and Ankle Nonunions 311
anatomy, the presence or absence of infection, period of months. Failed soft tissue coverage,
their biological potential, or their stiffness. Often failure to eradicate the infection, or failure to
more than one method of describing the nonunion obtain boney union may lead to eventual
will be helpful in determining a treatment plan. amputation.
Nonunions can be classified by their anatomic Weber and Čech [10] classified nonunions
location. Diaphyseal nonunions have relatively radiographically based on their biological
less biological potential as they involve cortical potential. Hypertrophic nonunions are charac-
bone, but are amenable to a wide variety of terized by abundant bone formation and are often
treatment methods, including nails, compression referred to as having the appearance of an ele-
plating, and external fixation. The goal in this phant foot. In general, they are stiff and relatively
instance is to restore length and axial alignment stable. Patients are often able to weight bear with
while achieving fracture union. As the nonunion pain on a hypertrophic nonunion. They have
reaches the metaphyseal region, the goals remain excellent blood supply and biological potential,
the same. The potential for bone growth and often require only the addition of stability for
improves in the metaphysis, but the options for the fracture to unite (Fig. 14.1). Atrophic non-
fixation are more limited. Fractures of the meta- unions, on the other hand, have little biological
diaphyseal region of the tibia are particularly potential. Atrophic nonunions are often the result
problematic. Peri-articular nonunions may also of open fractures or previous surgical procedures
be associated with stiff, contracted, or arthritic that have caused a disruption of the normal
joints that must be accounted for in the preop- vascular supply to the bone. They have had a
erative plan. Nonunions of the malleoli, with cessation of the regeneration process, resorption
their ligamentous attachments, can lead to joint of the bone ends, and in many instances, closure
instability. Nonunions of the articular surface are of the endosteal canal of the bone. These non-
particularly challenging. Defining the extent of unions are mobile; patients usually are unable to
the non-united segment may require multiple bear weight and may require external immobi-
radiographs and computed tomography scans. lization for comfort. A special case of the
Step-offs, gaps, and injury to the joint surface atrophic nonunion is a true pseudarthrosis in
may lead to local or global articular arthritis. In which a false joint has been created between the
the ankle, treatment may consist of open reduc- two ends of the bone. These fractures need bio-
tion and rigid fixation or arthrodesis. logical stimulation in addition to skeletal stabil-
Nonunions may be aseptic or infected. ity. Bone grafting and other adjuvants often play
Though many authors have shown that bone a role in their treatment. Oligotrophic nonunions
constructs with adequate stability can heal in the are somewhere in between these two extremes.
face of infection, the general goal is to convert an They have very little callous formation, but the
infected nonunion into a non-infected nonunion, bone ends are vital. They often require both
and then proceed with treatment of the fracture. biological and mechanical augmentation.
Though many infected nonunions will have skin Ilizarov described nonunions as stiff or mobile
breakdown, open wounds, and drainage, the [11–13]. Stiff nonunions, those with less than 7
diagnosis is not always obvious. Laboratory degrees of motion at the fracture site, are thought
studies can be helpful, as can nuclear medicine to be biologically viable and thus equivalent to
studies. The patient should be counseled that the hypertrophic nonunion. They require stability
treatment might take several staged procedures to heal, are biologically viable, and can be a
for hardware removal, debridement of dead bone, source of new bone formation if distracted. In
soft tissue coverage, and stabilization. A period many instances, they can be treated closed. Lax
of intravenous antibiotics-based thorough deep nonunions have more than 7 degrees of motion
cultures is followed by definitive reconstruction. and are equivalent to the atrophic nonunion of
Depending on the extent of the infection and the Weber. They not only require stability to heal, but
amount of bone resected, this may require a also require open debridement of the nonunion
312 K.J. Pugh, MD
Fig. 14.1 13-year-old girl hit by a light pole while c Application of a multiplanar external fixator to add
waiting for the school bus. a, b Anteroposterior (AP) and stability, no fracture exposure, deformity correction. d, e
lateral of the hypertrophic nonunion with deformity. AP and lateral of fracture union after frame removal
site to stimulate the bone and the addition of bone amount of support from family or friends, their
graft or other biological stimulus. They are not a financial resources, the location of their home,
source of new bone with pure distraction, but and what type of dwelling in which they reside is
instead require compression. helpful in planning successful aftercare.
The evaluation of the patient with a nonunion, A thorough musculoskeletal examination is
just as with an acute injury, requires a thorough also mandatory. Examination of the patient’s
look at more than just the fracture pattern and the other extremities will provide clues as to other
radiographs. One must determine the “personal- disabilities that may play a role in mobility and
ity of the fracture,” as coined by Schatzker and later rehabilitation. Examination of the
Tile [14]. This involves a complete history of the non-united segment includes an inspection for
events of the injury, the fracture, the host, the gross deformity and overall limb alignment.
treating physician, and the institution at which Gross limb length can be checked, and if the
the treatment will occur. Only with this kind of patient is ambulatory, the gait pattern should be
analysis can one do proper preoperative planning examined. The fracture site should be checked
and optimize the chance for success. for pain to manual stress, as well as the presence
of gross or subtle motion. The stability and
motion of adjacent joints should be examined.
14.5 Clinical Evaluation Ligamentous instability may require reconstruc-
tion as part of the treatment plan. If there is joint
A comprehensive history is essential, as a com- contracture or subluxation present, it should be
plete picture of the fracture and the host must be determined if it is due to soft tissue contracture,
obtained. Was the initial injury open or closed? heterotopic ossification, joint ankylosis, or a
Was there a high-energy mechanism such as a combination of factors.
motorcycle accident or a lower energy trip and The skin should be inspected for the presence,
fall? Were there any neurovascular issues at the location, and healing status of previous open
time of initial injury or after treatment? A wounds and incisions. Adherent skin, especially
determination of the type and number of previous in areas with subcutaneous bone such as the
surgeries is essential, as is the presence and medial face of the tibia, the distal fibula, and the
treatment of previous infection. If there is calcaneus, can be an obstacle. The presence or
retained hardware at the fracture, old operative absence of lymphedema or venous stasis should
notes can be helpful in identifying the hardware be noted, as it may influence the choice of sur-
type and manufacturer for planned removal. gical approach. If previous external fixators have
Have previous fractures healed in a timely fash- been in place, the condition of the old pin sites
ion? Patients with recreational drug habits or should be examined for signs of previous infec-
other substance abuse may have compliance tion. A complete neurovascular examination
issues. Smokers are at risk because of the should be carried out. Patients with suspected
well-documented relationship between nicotine dysvascular limbs should be sent for more thor-
use and delayed healing. Patients using nicotine ough testing, including transcutaneous oxygen
gum are not immune to this problem. The tension and ankle-brachial indices. Existing
occupation of the patient is important, as treat- nerve deficits can be examined tested by elec-
ment that requires a non-weight bearing gait will tromyography to determine the likelihood of
cause a longer period off work for a laborer than recovery.
for a patient with a more sedentary occupation. Radiographic evaluation includes true
The knowledge of the avocations and hobbies of anteroposterior and lateral films of the problem
your patient are also important, as it rounds out limb segment, orthogonal to the “normal” portion
the level of activity to which the patient must of the limb. Radiographic signs of a nonunion
return. Hospital discharge planning often begins can be subtle but include the absence of bridging
before surgery. The patient’s living situation, trabeculae, sclerotic fracture edges, persistent
314 K.J. Pugh, MD
fracture lines, and broken or displaced hardware. At the end of the evaluation, the surgeon
If deformity or limb length issues are suspected, should create a complete problem list in antici-
additional work-up is required. Standard pation of preoperative planning [15]. An attempt
full-length alignment films should be obtained, as should be made to define the cause of the non-
well as alignment films centered on each area in union and reverse it. Soft tissue defects, either
question, i.e., tibia or ankle. Deformities must be existing or anticipated, must be covered. The
fully characterized in all six axes so that correc- consults required should be listed and obtained.
tion can be planned. Comparison films of the Infected nonunions require debridement, tempo-
contralateral leg are helpful in determining the rary stabilization, and conversion to a
normal alignment of the patient, and population non-infected nonunion, with eventual staged
normals can be used if the problem is bilateral. reconstruction. Constructs with mechanical
Computed tomography scans with reconstruc- instability should be made stable; those with a
tions can be helpful in analyzing subtle non- gap require strategies to restore bone loss; those
unions, but can be hard to interpret with fracture with deformity require a better reduction or
fixation devices in place. Plain tomography can length; and those with vascularity require a better
be very helpful in these instances, but is soft tissue environment and biological
increasingly unavailable. If infection is sus- stimulation.
pected, a combined bone scan and tagged white Using this problem list, a detailed preopera-
cell study can help differentiate bone turnover tive plan should be drawn out in detail in all but
from active infection. Magnetic resonance the simplest of conditions. Putting the case on
imaging can be helpful in evaluating a bone for paper, often with multiple methods or implants,
infection, or looking at ligaments in adjacent allows one to foresee possible obstacles to suc-
joints, but are not commonly used in the evalu- cess, to define the sequential steps in the opera-
ation of nonunions. tion, to select the appropriate patient positioning
Laboratory studies can round out the clinical as well as to ensure the availability of equipment
picture of the patient. In addition to routine and implants, and to make the procedure in the
preoperative chemistries and blood counts, operating room the execution of a plan instead of
patients suspected of infection should have their a surgical adventure (Tables 14.1 and 14.2).
erythrocyte sedimentation rate and C-reactive Some nonunions need no treatment at all.
protein checked. Patients suspected of malnutri- Patients with normal alignment, normal function,
tion should have a complete nutritional panel and no pain may not require surgical treatment.
drawn, including liver enzymes, total protein and This is most common with small fractures of the
albumin levels, and calcium, phosphate and posterior malleolus or at the tips of the medial or
vitamin D levels. lateral malleolus. Nonunions of the metaphysis
The last part of developing the “personality of are usually painful or involve deformity, while
the fracture” is a critical self-examination of the those of the articular surface predispose the
surgeon and the treating facility. Surgeons should patient to arthritis and require treatment. In all
honestly examine whether they have the training, cases, surgery is contraindicated where the mor-
skill, patience, and experience necessary to treat bidity of the treatment exceeds the expected
a complex nonunion. Even the most gifted sur- benefit in function.
geon requires help, and the appropriate consul- Hypertrophic and stiff nonunions require only
tants must be available from plastic and vascular stability to promote union [11, 16]. Stable con-
surgery, internal medicine, and infectious dis- structs minimize motion, allow compression, and
ease. The hospital is the final element. Is the minimize shearing at the fracture site. A stable
correct equipment in the house or available to be fracture construct, made up of the patients bone
brought in? Is experienced nursing and surgical and the fixation device, allows stable vascular
assistance available? Can the anesthesia staff care ingrowth and the progression of fracture healing.
for the needs of the patient? This may involve the addition of blocking screws
Table 14.1 Treatment suggestions: distal tibia and ankle nonunions

Classification Objective Treatment Suggestions Problems
Hypertrophic Provide stability Plate, external Does not require grafting Must provide
nonunion fixation, nail with adequate
polar screws stability
Atrophic/oligotrophic Provide stability Bone graft or Thorough debridement or Failure to
nonunion and biological appropriate excision of nonunion provide biology
stimulus substitute, stable and stability
fixation
Nonunion with Treat nonunion Deformity correction, Formal deformity analysis Failure to
deformity and deformity stability and biology restore
mechanical axis
Metaphyseal Maintain axial Plate, external Provide adequate fixation, Prevention of
nonunion alignment fixation, nail with build external fixator to foot deformity
polar screws if needed
Malleolar nonunion Restore joint AO techniques Restore ankle mortise, Failure to
stability stress views in OR restore joint
stability
Articular nonunion Restore articular Rigid internal Arthrodesis if surface is not Cartilage injury,
surface fixation reconstructable poor prognosis
Table 14.2 Treatment strategy: distal tibia and ankle nonunions

Treatment method Clinical indication
Plate and screw fixation Metaphyseal, malleolar, or articular nonunion, no infection, adequate soft tissue
Intramedullary nail Metaphyseal location, may require polar screws for stability, no infection
Multiplanar external Larger deformity, leg length deficiency, infection, bone defect, poor soft tissue, joint
fixation subluxation
Acute correction Small or no deformity, no lengthening, adequate soft tissues, nonunion requires open
approach
Gradual correction Larger deformity, leg length deficiency, infection, bone defect, poor soft tissue, joint
subluxation
to improve the stability of a nailing construct in plastic surgery consults may be necessary to
the distal tibial metaphysis, plate fixation, or correct soft tissue problems. These fractures must
compression with an external fixator . be opened and the bone ends debrided back to
healthy viable tissue. All nonviable scar tissue
must be removed, and the endosteal canal of the
14.6 Atrophic Nonunions bone must be opened, either with a curette or a
drill bit.
Atrophic and oligotrophic nonunions require Atrophic fractures also require the addition of
both stability and biology. Patients with bone boney stimulus in the form of bone grafting [17–
loss and infection fall into this category as well 21]. Though the gold standard remains autoge-
(Fig. 14.2). These fractures require the most nous cancellous bone from the iliac crest, there
preoperative planning, often involving several are many other methods available to the ortho-
stages, as the physiological environment is pedic surgeon. Autogenous graft may come in the
inadequate to promote healing. Medical prob- form of local bone from the proximal tibia or
lems must be treated, while vascular surgery and calcaneus and endosteal bone harvested from the
316 K.J. Pugh, MD
Fig. 14.2 30-year-old man after high-velocity gunshot loss c Anteroposterior after bone grafting with reamer
wound. a Initial injury. b Three months after initial irrigator aspirator. d Fracture union 3 months after
percutaneous plating, impending nonunion due to bone grafting
femur or the tibia with a reamer-irrigator-aspirator Fixation) or AO group. Plating can be used to
or with other patient-derived material such as provide minimally invasive bridging fixation and
bone marrow aspirate. There are many commer- stability to a well-aligned hypertrophic nonunion
cially available bone graft substitutes in the in a previously non-operatively treated distal
marketplace from osteoconductive ceramics to fibula or tibial metaphysis. However, this method
osteoinductive growth factors, each with a generally requires judicious fracture exposure for
specific use. Some patients will require more of removal of hardware and direct reduction of the
an osteoconductive scaffold, while others will fragments. In nonunions with a questionable soft
require a true osteogenic graft substitute. The tissue envelope, this requirement can limit the
challenge to the surgeon is to cut through the available surgical approaches and fracture access.
extensive marketing noise and to select the pro- AO technique allows for rigid fixation of artic-
duct, combination of products, or method that ular fragments and is thus required in nonunions
will solve unique clinical needs of each specific involving the articular surface as well as the
patient. malleoli and syndesmosis. Plating also allows for
absolute or relative stability constructs of the
metadiaphysis [25]. Locking screw technology
14.7 Infected Nonunions and anatomically contoured plates allow fixation
of increasing smaller segments of the distal tibia.
Infected nonunions (Fig. 14.3) require a thorough Plating techniques are suitable for use in non-
debridement with intraoperative cultures. Exist- unions with smaller and acutely correctable
ing colonized hardware must be removed. Dead deformities, as well as in those with only small
bone and devitalized soft tissue must be debrided length discrepancies. As deformities become
back to a healthy fracture bed. The endosteal larger, especially in length, they are stiffer and
canal must be opened to improve blood supply. more difficult to correct; plating becomes more of
In some cases, this means resecting the infected a challenge.
nonunion as if it were a malignancy. Dead space
and areas of bone loss may be filled with
antibiotic impregnated methacrylate beads or 14.9 External Fixation
block spacers. These methods preserve space for and Intramedullary Nailing
bone grafting, elute local antibiotics, and allow
the formation of vascularized membranes as Simple uniplanar external fixation is useful for
popularized by Masquelet [22, 23]. Most infected the temporary stabilization of an infected non-
nonunions are treated in a staged fashion. The union in preparation for definitive reconstruction.
first stage is to convert the infected nonunion to Uniplanar constructs, usually with two tibial half
an aseptic atrophic nonunion by creating a viable pins and a transfixation pin through the calcaneus
fracture environment, obtaining deep cultures, joined by bars in a delta configuration, provide
managing dead space, and providing temporary relative stability at a distance after removal of
stability. The second stage is definitive treatment hardware and fracture debridement in an infected
of the atrophic fracture. case, or one that may require soft tissue cover-
age. In preparation for the second stage of
treatment, patients are more comfortable, the
14.8 Traditional Plating Techniques fracture is grossly aligned, and the skin is
accessible to examine wounds or flaps. These
Nonunions of the distal tibia and ankle are very frames are generally removed at the time of
amenable to treatment with traditional plating definitive reconstruction.
techniques [24], following the principles advo- Multiplanar external fixation [26, 27] is
cated by the Arbeitsgemeinschaft für Osteosyn- generally reserved for definitive fixation of
thesefragen (Association for the Study of Internal nonunions of the distal tibial metaphysis. These
318 K.J. Pugh, MD
Fig. 14.3 40-year-old diabetic 8 weeks after pilon external fixator, and antibiotic spacer. c After flap
fracture, grossly infected. a Mortise on presentation, coverage, nail, reamer-irrigator-aspirator graft. Salvage
medial wound visible. b After debridement of infected of joint with arthrodesis. d Final union
distal tibia, hardware removal, temporary spanning
Fig. 14.4 40-year-old patient eight weeks after treatment fixator for joint distraction and reduction e after correc-
for a trimalleolar ankle fracture a mortise, b lateral and tion of joint subluxation f intraoperative view of repair of
c computed tomography of malunited medial malleolus, posterior malleolus g twenty-four month follow up with
non-united lateral and posterior malleolus with posterior joint reduction and fracture union
ankle subluxation d application of multiplanar external
320 K.J. Pugh, MD
Fig. 14.5 70-year-old man after distal tibia fracture while external fixation. c Antibiotic spacer removal, bone graft-
on cruise (repaired in the Middle East). a Fracture at ing. d Nonunion in external fixator. External fixator
presentation. Erythrocyte sedimentation rate and C-reactive removed, frame holiday. e Salvage of metaphyseal
protein elevated. b Hardware removed, antibiotic spacer, nonunion with intramedullary nailing. f Fracture union
devices are available in a unibody design with bearing during treatment. The major advantage
fixation to the patient using half pins in the to these devices in distal tibia nonunions is the
tibia and foot, or a circular ring configuration ability to simultaneously correct multi-axis
using thin wires and half pins for fixation. deformity, lengthen the bone if necessary, and
Traditional circular frames are built specifically modulate the fracture dynamically in compres-
for each patient, but may require multiple sion and distraction. They can also be used to
revisions in patients requiring step-wise cor- span the ankle joint, allowing simultaneous
rection of accompanying deformity. Newer correction of foot deformities, ankle subluxa-
fixator designs incorporate a Web-based com- tion, or contracture (Fig. 14.4). In many
puter program that may decrease the need for instances, they can be combined with rigid
multiple rebuilds. Multiplanar external fixators screw fixation of the joint to solve multi-focal
are very stable and can allow earlier weight problems.
Intramedullary nailing is generally not a pri- 3. Bhandari M, Fong K, Sprague S, Williams D,

mary method in the treatment of distal tibia and Petrisor B. Variability in the definition and perceived
causes of delayed unions and nonunions: a
ankle nonunions [28]. Metadiaphyseal fractures cross-sectional, multinational survey of orthopaedic
of the tibia treated initially with a nail and which surgeons. J Bone Joint Surg Am. 2012;94(15):
go on to nonunion are often revised to another e1091–6.
form of fixation to add stability. Because of the 4. Haverstock BD, Mandracchia VJ. Cigarette smoking
and bone healing: implications in foot and ankle
size mismatch between the canal diameter and surgery. J Foot Ankle Surg. 1998;37(1):69–74;
the nail, nailing does not increase stability unless discussion 8.
used in combination with blocking screws or 5. Niikura T, Lee SY, Sakai Y, Nishida K, Kuroda R,
supplementary plates. A common application of Kurosaka M. Causative factors of fracture nonunion:
the proportions of mechanical, biological,
a nail in this setting is in the salvage of the patient-dependent, and patient-independent factors.
unreconstructable ankle with a pantalar fusion. J Orthop Sci. 2014;19(1):120–4.
Nails can also be used to accelerate frame 6. Hernigou J, Schuind F. Smoking as a predictor of
removal in fractures with slow healing regenerate negative outcome in diaphyseal fracture healing. Int
Orthop. 2013;37(5):883–7.
or bone graft (Fig. 14.5). 7. McKee MD, DiPasquale DJ, Wild LM, Stephen DJ,
Arthodesis is a viable option in ankle and Kreder HJ, Schemitsch EH. The effect of smoking on
distal tibia nonunions in patients with joint clinical outcome and complication rates following
stiffness, an articular surface that is not recon- Ilizarov reconstruction. J Orthop Trauma. 2003;17
(10):663–7.
structable, or with a poor capacity to heal. Lag 8. Schmitz MA, Finnegan M, Natarajan R,
screw fixation, blade plating, and external fixa- Champine J. Effect of smoking on tibial shaft fracture
tion can all be utilized successfully to achieve a healing. Clin Orthop Relat Res. 1999;365:184–200.
9. Brinker MR, O’Connor DP, Monla YT, Earth-
solid fusion. If the subtalar joint remains mobile,
man TP. Metabolic and endocrine abnormalities in
it may be spared. If it is stiff, painful, or con- patients with nonunions. J Orthop Trauma. 2007;21
tracted, the ankle may be salvaged with a fusion (8):557–70.
of both ankle and subtalar joints. 10. Weber BG, Čech O. Pseudoarthrosis: pathology,
biomechanics, therapy, results. Berne, Switzerland:
Amputation is always part of the informed
Hans Huber Medical Publisher; 1976.
consent process, especially in patients who have 11. Ilizarov GA. The principles of the Ilizarov method.
had multiple procedures for stability and soft tissue Bull Hosp Jt Dis Orthop Inst. 1988;48(1):1–11.
coverage and are still not united. Some of these 12. Ilizarov GA. The tension-stress effect on the genesis
and growth of tissues. Part I. The influence of
patients will have been undergoing limb recon-
stability of fixation and soft-tissue preservation. Clin
struction for several months without result and can Orthop Relat Res. 1989;238:249–81.
view this discussion as an admission of failure. 13. Ilizarov GA. The principles of the Ilizarov method.
These patients often have severe economic and 1988. Bull Hosp Jt Dis. 1997;56(1):49–53.
14. Schatzker J, Tile M. The rationale of operative
social pressures and pain control issues, and are
fracture care. 3rd ed. Berlin: Springer; 2005.
facing the prospect of months of continued treat- 15. Wiss DA, Stetson WB. Tibial nonunion: treatment
ment. An amputation is not a failure for the patient, alternatives. J Am Acad Orthop Surg. 1996;4
but a way to cut their losses and to get on with life. (5):249–57.
16. Müller ME. Treatment of nonunions by compression.
A well-done below knee amputation will remove
Clin Orthop. 1965;43:83.
the problem and allows the patient to be on their feet 17. Dawson J, Kiner D, Gardner W 2nd, Swafford R,
in a number of weeks instead of months. Nowotarski PJ. The reamer-irrigator- aspirator as a
device for harvesting bone graft compared with iliac
crest bone graft: union rates and complications.
J Orthop Trauma. 2014;28(10):584–90.
References 18. Masquelet AC, Benko PE, Mathevon H, Han-
nouche D, Obert L. French Society of Orthopaedics
1. Mann RA. Biomechanics of the foot. Instr Course and Traumatic Surgery. Harvest of cortico-cancellous
Lect. 1982;31:167–80. intramedullary femoral bone graft using the
2. Kotwick JE. Biomechanics of the foot and ankle. reamer-irrigator-aspirator (RIA). Orthop Traumatol
Clin Sports Med. 1982;1(1):19–34. Surg Res. 2012;98(2):227–232.
322 K.J. Pugh, MD
19. Pape HC, Evans A, Kobbe P. Autologous bone graft: 24. Khurana S, Karia R, Egol KA. Operative treatment of
properties and techniques. J Orthop Trauma. 2010;24 nonunion following distal fibula and medial malle-
(Suppl 1):S36–40. olar ankle fractures. Foot Ankle Int. 2013;34(3):
20. Quintero AJ, Tarkin IS, Pape HC. Technical tricks 365–71.
when using the reamer irrigator aspirator technique 25. Carpenter CA, Jupiter JB. Blade plate reconstruction
for autologous bone graft harvesting. J Orthop of metaphyseal nonunion of the tibia. Clin Orthop
Trauma. 2010;24(1):42–5. Relat Res. 1996;332:23–8.
21. Schmidmaier G, Herrmann S, Green J, Weber T, 26. Kabata T, Tsuchiya H, Sakurakichi K, Yamashiro T,
Scharfenberger A, Haas NP, et al. Quantitative assess- Watanabe K, Tomita K. Reconstruction with distrac-
ment of growth factors in reaming aspirate, iliac crest, tion osteogenesis for juxta-articular nonunions with
and platelet preparation. Bone. 2006;39(5):1156–63. bone loss. J Trauma. 2005;58(6):1213–22.
22. Masquelet AC, Begue T. The concept of induced 27. Feldman DS, Shin SS, Madan S, Koval KJ. Correction
membrane for reconstruction of long bone defects. of tibial malunion and nonunion with six-axis
Orthop Clin N Am. 2010;41(1):27–37; table of analysis deformity correction using the Taylor spatial
contents. frame. J Orthop Trauma. 2003;17(8):549–54.
23. Pelissier P, Masquelet AC, Bareille R, Pelissier SM, 28. Richmond J, Colleran K, Borens O, Kloen P,
Amedee J. Induced membranes secrete growth Helfet DL. Nonunions of the distal tibia treated by
factors including vascular and osteoinductive factors reamed intramedullary nailing. J Orthop Trauma.
and could stimulate bone regeneration. J Orthop Res. 2004;18(9):603–10.
2004;22(1):73–9.
Special Techniques for Nonunions
Associated with Traumatic Bone Loss 15
Mark A. Lee, MD and Michael P. Leslie, DO
Therefore, in large defects, specialized approa-

15.1 Introduction
ches are required. The most common classic
techniques are vascularized free bone transfer and
The reconstruction of long bone defects is often a
Ilizarov bone transport. Both of these techniques
major challenge in limb salvage regardless of the
require specialized training or equipment and a
etiology of bone loss. There is limited high-grade
high level of surgical expertise in conjunction
evidence on the efficacy of multiple techniques for
with postsurgical resources and support. Signifi-
bone regeneration and especially on comparative
cant patient compliance and cooperation are
outcomes of different management techniques.
required, and very large defects require protracted
The literature is rife with limited case series and
treatment times. Despite these limitations, these
single surgeon experiences that do not provide
are powerful techniques for bone regeneration,
evidence-based treatment recommendations.
and the results can be remarkable. In many situa-
Nonetheless, bone loss remains a common prob-
tions, neither vascularized transfer nor bone
lem for clinicians, and multiple approaches are
transport is optimal or available, and novel tech-
utilized depending on surgeon experience and
niques are being utilized. The induced membrane
resources. To a certain extent, defect size dictates
technique (Masquelet) is increasingly utilized in
many of the treatment approaches. Small, stabi-
bone defects to extend the application of cancel-
lized defects (2–3 cm) are frequently treated with
lous grafting to larger defect sizes. This technique
acute cancellous autograft application. While this
has been well reviewed in the literature [2–4] and
may be effective in favorable (i.e., well vascular-
is utilized with increasing frequency for massive
ized) healing sites, this is not typically used for
defects. Even more novel approaches to defects
larger defects (greater than 4 cm). In large defects,
include the use of spinal cages (for graft contain-
the healing is unpredictable, and significantly,
ment and structure support) and noncustom por-
larger bone graft volumes are necessary which
ous tantalum implants (for structural support and
leads to concerns about graft absorption [1].
defect substitution). These more unique approa-
ches await systematic evaluation but do provide
solutions in recalcitrant cases.
M.A. Lee (&)
Department of Orthopedic Surgery, University of
California, Davis, 4860 Y Street, Suite 3800, 15.2 Distraction Osteogenesis
Sacramento, CA 95817, USA
e-mail: mark.lee@ucdmc.ucdavis.edu
The concept of distraction osteogenesis as a
M.P. Leslie method of skeletal reconstruction can be traced
Department of Orthopedics and Rehabilitation,
Yale School of Medicine, New Haven back to as early as 1905. This has been done with
CT 06520, USA a variety of methods including osteotomy and

324 M.A. Lee. MD and M.P. Leslie, DO
immediate traction, external fixation, internal use of fine wire attached to circular frames. This
fixation with either intramedullary rods or extra- imparted stability and allowed for limb salvage
medullary osteosynthesis, and fine wire fixation. for many limbs that would have otherwise
Significant bone loss occurs in a minority of undergone amputation (Fig. 15.1). This was done
fractures (0.4%) but that is significantly higher in out of necessity as he faced a community of
cases of open fracture and when planned inter- patients in Siberia where antibiotics were scarce,
vention demands resection of large segments of osteomyelitis was common, and amputation led
bone [5]. Each case of bone loss carries an indi- to poor outcomes [6, 9]. The additional capacity
vidual character that is comprised of the patient to correct deformity while concurrently treating
demographics and comorbidities along with the bone loss remains unparalleled; however, the
injury itself. In the presence of an acute open technical challenges for the surgeon and the
fracture, the key concepts include soft tissue practical difficulties for the patient continue to
compromise and instability. In the case of tumor, limit the use of this technique. The current section
the demands of cancer treatment and medical discusses the use of fine wire circular fixation,
compromise of the patient have significant impact unilateral rail distraction, distraction over intra-
on the planned intervention for bone loss. Infec- medullary devices, and distraction with plate
tion and nonunion include both concepts of soft osteosynthesis as unique treatments for bone loss.
tissue compromise along with inflammation and
bone loss. The ultimate treatment for any of these
situations would include resection and bone 15.2.1 Fine Wire Circular Fixation
grafting with a source that provides cortical sta-
bility and rapid integration into the human External fixation has distinct advantages with
skeleton without risk of infection or rejection. respect to the ability to avoid direct instrumen-
Unfortunately, there are no current interventions tation at sites of infected nonunions and also with
that can achieve these idealistic goals. However, the ability to slowly correct deformity, which
the use of distraction osteogenesis remains the potentially can limit the risk of injury to struc-
closest to this potential gold standard. With the tures at risk. Fine wire circular fixation remains a
use of corticotomy and distraction, the donor powerful tool for both the correction of deformity
source risk is minimized as the graft is a similar and the application of distraction forces that
shape and has a robust soft tissue envelope pro- allow for deposition of new bone. The most
viding biology to the healing environment. critical components linked to this remain to be
Distraction osteogenesis refers to the forma- the handling of the soft tissues during treatment
tion of new bone between two ends of vascular- (Fig. 15.2). The surgeon might choose the use of
ized bone that are gradually distracted [6]. This fine wire circular fixation in the setting of a
can be accomplished through multiple methods of nonunion that involves bone loss and angular
distraction. This concept was first introduced in deformity. All external fixator systems allow for
1905 by Codivilla, who performed the first suc- multiple planes of freedom, but the use of fine
cessful limb lengthening by osteotomy and wire circular fixation is the only system that
immediate transcalcaneal traction [7]. In 1913, allows for both elastic control and dynamic
Ombredanne reported the fist use of external control that respect bone biology. When an
fixation for distraction. This was improved upon in-line or even multiplanar fixator is utilized with
by Putti in 1921 who utilized a monolateral fix- half pin fixation alone, there is not just control of
ator at a rate of 2–3 mm/day as opposed to the length imparted but a distinct lack of control of
5 mm/day used by the sentinel author [8]. These angulation. This lack of control is considered
concepts have been applied to many different “parasitic” to bony healing as it is uneven and
clinical scenarios with variable success and nonbiologic. With the use of fine wire fixation,
complication. After World War II, Ilizarov began the stability that is imparted will allow for heal-
to develop the concept of distraction through the ing by secondary intention and callous formation
15 Special Techniques for Nonunions Associated … 325
Fig. 15.1 31-year-old male who suffered a moped technique with a proximal corticotomy in a multiplanar
accident with an isolated complex open intraarticular external fixator. The patient went on to consolidate the
distal tibia and fibula fracture. He underwent staged regenerate and heal the docking site without need for bone
management and with debridement and spanning external grafting, despite severe noncompliance with care. He
fixation, followed by open reduction internal fixation of currently walks without assistive device and has since had
the articular block and application of antibiotic impreg- his distal tibial hardware removed due to a late infection
nated beads until he healed a free latissimus flap. 5 cm of due to shoewear breakdown of the free flap
bone loss was then healed using a distraction osteogenesis
but will at the same time limit the “parasitic” lack

of control of angulation [10].
The more popularized understanding of fine
wire fixation is that it can be used in conjunction
with independent distraction–compression devi-
ces that will allow for multiplanar correction of
deformity by application of compression in one
plane and distraction in another.
The use of fine wire circular fixation has been
successfully utilized in many clinical series as
outlined above to achieve restoration of skeletal
alignment and length. The cost and complexity
Fig. 15.2 Clinical photograph of a 38-year-old male
associated with these types of systems can,
who suffered a motorcycle collision with a complex
Gustilo and Anderson type IIIB open proximal tibia however, be burdensome and has lead many
fracture with 10 cm of proximal tibial bone loss. This surgeons to unilateral frames due to the ability to
patient required careful debridement, open reduction achieve skeletal success and simplify the process
internal fixation, and massive autologous bone grafting for both the surgeon and the patient.
using a Masquelet technique after a free flap successfully
healed. He ambulates without assistive device at 2 years In this technique, the nonunion site is debrided
post-reconstruction of all nonviable tissue and bone after the removal of
any preexisting internal fixation devices. A unilat- union, no angular deformity, a moderate amount
eral frame can then be applied in a monofocal or of pin site infections, and a 0.4 external fixation
bifocal method. In the monofocal method, com- index (number of months external fixator system
pression and distraction is initiated at the fracture worn divided by centimeters of distraction) [13].
site to stimulate osteogenesis. Distraction can then
also be done at the nonunion site to restore leg 15.2.2.2 Telescopic Intramedullary
length. If a bifocal method is done, the distraction is Restoration of Length
achieved outside of the nonunion site. The problems associated with lengthening over an
This is a widely used technique in all long intramedullary nail are consistent with external
bones. Harshwal et al. recently presented a series fixation problems in general. These include
of 37 patients (7 femur and 30 tibias) all treated pin-tract infections, scarring, pain, and patient
for nonunion within the first 8 months of the comfort. In order to obviate these problems, several
injury. Rate of union was reported at 91%. Min- entirely intramedullary devices have been devel-
imal complications were noted, primarily those of oped with the goal of using an internal lengthening
pin-tract infections. These results are consistent mechanism to provide distraction osteogenesis. The
with those reported by other authors [4, 11, 12]. intramedullary skeletal kinetic distractor (ISKD,
Orthofix Inc., McKinney, TX, USA, and the
PRECICE intramedullary nail (Ellipse Technolo-
15.2.2 Distraction Over gies, Irvine, CA, USA) utilize novel techniques of
Intramedullary Nails lengthening from within the canal (Fig. 15.3).
The ISKD Nail utilizes two internal rotating
15.2.2.1 Intramedullary Device Plus clutches to advance a threaded rod within the nail
External Fixation that is attached to the distal segment beyond an
Given the technical difficulties of controlling osteotomy with interlocking bolts. This provides
transport segments during distraction osteogene- distraction that is based on typical activities of
sis with purely external fixation, fine wire, or daily living that provide stimulus through 3–9
Schanz pin devices, the idea of guidance of the degrees of rotation through the osteotomy site.
transport over intramedullary devices has become There have been many challenges with this
appealing. In addition, the angular deformities device including a lack of absolute control of
introduced by the use of a unilateral rail fixator distraction. This can be due to variable activities
alone, in conjunction with the inability to be fully of patients, but can lead to a rate of distraction that
weight bearing, have demanded the ability to is suboptimal, either too fast or slow [14, 15].
guide a correction over an intramedullary device. The PRECICE nail uses an externally applied
In a recent series, Gulabi altered the original magnetic device to control the lengthening. The
descriptions of other authors to utilize acute proposed advantages to this include the ability to
compression and distraction osteogenesis. These not only monitor the lengthening but also change
patients were all tibial diaphyseal fractures with the prescription of lengthening based on optimal
bone loss. Custom intramedullary nails were conditions and the regenerate response time.
utilized with multiple locking hole options. In There is less clinical evidence regarding this
this technique, the bone loss site is cleared and a device but results appear similar to the ISKD
distant metaphyseal corticotomy is made that with unique difficulties encountered [16, 17].
liberates a transport segment. The bone loss With respect to critical cortical defects and non-
segment is shortened up to 5 cm, and the corti- union, these devices can be utilized for either com-
cotomy site is compressed. The transport then pression of a fracture site or distraction osteogenesis.
proceeds at 2 mm/day, and when docking is If a defect is predicted, this can be used to compress
achieved, the site is bone grafted from the iliac the fracture and then to perform an osteotomy and
crest. Their results demonstrated radiographic distract healthy bone to attain regenerate.
Fig. 15.3 55-year-old male who underwent en bloc intramedullary nail extraction and application of an
resection of the femur for malignant fibrous histiocytoma intramedullary telescopic nail with proximal corticotomy
20 years prior to presentation. The intercalary allograft through native metaphysis. At 6 months, post-op patient
femur had healed with limb foreshortening that lead to was pain free at the upper thigh and underwent a total hip
extensive back pain and hip arthritis. Staged management arthroplasty with concomitant removal of hardware at
included restoration of standing balance with 1 year
15.2.3 Distraction with Plate the time of publication. The primary advantage is
Osteosynthesis the ability to stabilize the transport segment and
remove the external fixator despite a lack of
The use of intramedullary nails in conjunction radiographic union. The disadvantage is theo-
with external fixator distraction can be compli- retically the lack of loadbearing the plate can
cated by pin site infection that can develop into contribute. However, the advantages of being
an intramedullary infection due to the proximity able to apply this technique to skeletally imma-
of the pins and the nail. It is also limited by the ture patients, large amount of bone available for
ability to apply transport to a proximal or distal placement of external fixation, and decreased
fracture. Oh et al. [18] recently reported the use time to removal of external fixation can outweigh
of locking plate stabilization with external fixator these disadvantages (Fig. 15.4).
generated distraction osteogenesis. In their series
of ten patients, a similar technique of cortico-
tomy is performed, and after a latency period, 15.3 Masquelet Technique
distraction proceeded with 1 mm/day. When the
docking site is achieved, the transport segment is The induced membrane technique is a unique
stabilized with screw fixation through the plate, alternative to acute bulk grafting. This technique
the docking site is grafted, and the external fix- was originally utilized for regeneration of dia-
ator is removed. All patients achieved radio- physeal defects, but use has been expanded to
graphic union, and complications involve pin site metaphyseal defects as well. Professor Masquelet
infections only. Theoretically, these patients developed the technique in early 1984 and soon
might be at higher risk for fracture of regenerate after initiated a clinical study to demonstrate its
bone, although this has not occurred for them at efficacy [2].
Fig. 15.4 14-year-old male who underwent resection for distally. Allowed for 4.6 cm of distraction in 63 days of
osteosarcoma with limb foreshortening and flexion con- external fixation. Consolidation of regenerate noted to be
tracture of the knee. Distraction with plate osteosynthesis complete by 4 months on the right. (Courtesy of
utilized with proximal tibial corticotomy highlighted Chang-Wug Oh, MD, Kyungpook National University
(yellow arrow) to the left. External fixator removed at Hospital, Daegu, Korea)
7 weeks and consolidate locked into the plate construct
Key Features success. Extensive animal evaluations in both

small and medium animal models have demon-
• A bioactive membrane is created by place- strated the membrane is made of a type I
ment of a Poly(methyl methacrylate) (PMMA) collagen-heavy matrix and fibroblastic cells. The
block into a clean, debrided defect (Fig. 15.5). membrane itself has tissue level organization
• The blood supply around the induced mem- with an inner aspect of epithelial-like fibroblasts
brane is left intact or optimized by free tissue and collagen bundles that run parallel to the
transfer. surface of the membrane. This tissue is well
• The induced membrane is incised, and the vascularized and contains a high concentration of
PMMA block is carefully removed, leaving vascular endothelial growth factor. Typically, a
the membrane intact as a protective and solid block of PMMA is used to produce the
supportive grafting bed. spacer; this induces a mild foreign-body inflam-
• Slow consolidation is observed, and weight matory response with giant cells and macro-
bearing is restricted until union [2]. phages. The inflammatory response slowly
decreases over time following spacer implanta-
tion may disappear by 6 months following bone
grafting. Tissue from these membranes has been
15.3.1 Membrane analyzed using molecular techniques including
immunohistochemistry, and these studies
The induced membrane is believed to be a unique demonstrate expression of proteins associated
property of this technique and critical to its with induction of new bone formation. Thus,
can be managed with negative pressure therapy

or bead pouch depending on the individual
patient need. Once the soft tissue bed is clean and
mature, the definitive solid spacer can be placed
with simultaneous muscle coverage.
When feasible, intramedullary reaming is
performed ot aid in the debridement of the
intramedullary canal and to stimulate an endo-
steal healing response. For optimum membrane
induction and better stability of the construct, the
cement should be placed inside the canal (when
feasible) and over the edges of the native bone
(wrapping) and should fill the space of defect.
While external fixation was utilized in the orig-
inal technique, more stable forms of internal
fixation are typically utilized, even intramedul-
lary nails. Use of intramedullary nails can
decrease required graft volumes and provide
long-term stability in these slowly healing con-
Fig. 15.5 The forceps are holding the induced mem- structs. Finally, optimal soft tissue blood supply
brane which has been opened longitudinally and provides is requisite around the induced membrane zone.
vascularized pouch for graft material Free tissue transfer is far optimal to a tight pri-
mary wound closure especially in the
many feel that these membranes are bioactive. In mid-to-distal tibia.
addition, the induced membrane also acts to
eliminate soft tissue interposition into defects and
created a protective cavity to accept bone graft. 15.3.3 Outcomes
The shape and size of the healed bone graft are
defined by the membrane [2, 19–22]. The original Masquelet series of 35 patients
with upper and lower extremity segmental
defects that measured 4–25 cm in length
15.3.2 Technique reported a 100% healing rate. Most of these
were treated with external fixation and many
By definition, this is a two-stage technique. The had free flaps. The mean time to full weight
first stage is akin to a tumor debridement with bearing was 8.5 months [23]. While this series
aggressive removal of nonviable bone, scar, and is impressive, it likely does not represent con-
any damaged or nonviable local soft tissues. The temporary use of the technique. Subsequent
bone debridement cannot be limited since fre- reports have included the use of bone morpho-
quently bone necrosis at the fracture edges has genetic protein (BMP), reamed intramedullary
progressed significantly proximal to the defect. grafts, and multiple modes of internal fixation—
After debridement/resection, the remaining bone for most of these techniques, ultimate union
ends should be healthy with a viable bleeding rates hover around 90% [2, 24, 25]. While
bed (Fig. 15.6). In the setting of a severe soft many of these publications report good results,
tissue deficit or wound problem, standard dead overall the level of evidence for this technique
space management techniques using PMMA remains low since these are mostly retrospective
bead strands can be used, while the preliminary case series or small prospective noncomparative
wound management is performed. Open wounds studies.
Fig. 15.6 Diaphyseal infection undergoes aggressive over wrapped with PMMA. At 8 weeks, the wound is
resection and debridement. a, b The defect is filled with filled with cancellous autograft and BMP and formal plate
PMMA, and preliminary stabilization is achieved with fixation is utilized. e, f At 6 months, the regenerate is
external fixation. c, d Classically, the ends of the bone are completely healed and the patient is weight bearing
15.3.4 New Considerations of the membrane and may suggest a role for
earlier secondary grafting. So in the absence of
The timing of bone grafting into the membrane definitive evidence for specific timing, grafting
has been recently evaluated [26]. While con- can be safely performed as soon as the wounds
temporary approaches demonstrate large vari- have healed well without evidence of residual
ability in timing of secondary cancellous grafting infection and systemic antibiotic therapy is near
into the membrane bed, most surgeons delay complete (4–6 weeks). There is likely little ben-
6 weeks or more after placement of the spacer. efit to protracted delays (greater than 8 weeks) to
A closer evaluation of one of the original animal secondary graft application.
studies demonstrated the time course of growth
factor expression from induced membrane sam-
ples with quantitative and qualitative immuno- 15.4 Cage Technique
histochemistry [20]. Maximum BMP-2 levels
were seen at 4 weeks post-procedure with In 2002, Ostermann published the first reports of
decrease over subsequent weeks. These data may extending the indication for use of titanium mesh
suggest that the optimal time of membrane cages to restore bony continuity [27]. These
bioactivity is earlier than suspected. Samples of devices are routinely utilized in spine surgery to
human induced membrane tissue were assayed augment the use of nonstructural allograft. They
for multiple time points. One-month-old mem- have demonstrated adequate ability to achieve
brane samples had the highest expression of bony union in conjunction with bone graft [28,
VEGF, IL-6, and Col-1, whereas two-month-old 29]. The goal of utilizing the titanium cage is that
membranes expressed <40% of the levels of the cancellous allograft and demineralized bone
one-month-old membranes [26]. This study matrix products offer advantages of no donor-site
suggests a time-dependent decrease in bioactivity morbidity and ease of application. The difficulty
in utilization of nonstructural allograft bone is 15.5.1 Material

that it does not reliably lead to bony union in
gaps greater than 3 cm, those of critical cortical Tantalum is a transition metal (atomic number
defects. The addition of the titanium mesh cage 73; atomic weight 180.05) that remains relatively
extends the application of the allograft material inert in vivo. Porous tantalum is an open-cell
by imparting additional stability. tantalum structure of repeating dodecahedrons
The technique involves either plate or intra- with an appearance similar to cancellous bone
medullary nail stabilization. It can be performed has been developed for clinical applications.
either acutely, on a delayed basis or in a non- (Zimmer-Biomet, Trabecular Metal Technology,
union setting. In each case, the cage is premea- Inc., Parsippany, NJ, USA). The basic structure
sured in accordance with the diameter of the of this porous tantalum metal yields a high vol-
bone and also the length of the defect to be umetric porosity, a low modulus of elasticity, and
spanned. The cage is prepared with a packing relatively high frictional characteristics [33]. This
that consists of cancellous bone graft, and if an frictional characteristic makes immediate stable
intramedullary nail is to be used, the guide wire interfaces with bone feasible and allows the
is passed through the middle to ensure that there potential for early or immediate weight bearing
is no mechanical blockade to passage. Standard (Fig. 15.8) [34].
intramedullary nailing techniques can then be Porous tantalum structures utilized for ortho-
utilized including reaming over a guide wire pedic implants have a porosity of 75–85%
(Fig. 15.7). Ostermann, Attias, and Cobos all compared to CoCr sintered beads (30–35%) [35].
reported success in small series with minimal The rigidity of porous tantalum increases with
complication, most notably in leg length dis- decreasing porosity. Current tantalum implants
crepancy [27, 30–32]. maintain a rigidity similar to the human fibula
In some situations, plate osteosynthesis might [36]. These characteristics optimize the biocom-
be the preferred method. Attias recommended patibility of these implants.
plate osteosynthesis in the setting of nerve In addition to its high biocompatibility, the
exploration or when intramedullary nailing might frictional characteristics and the rigidity similar-
be suboptimal such as a proximal or distal ity to native bone make porous tantalum an
metaphyseal segments. The same methodology intriguing candidate for defect management.
of preparation was performed in the single case With tantalum implants in structural defects,
report using this method, and the cage was stable implantation, structural support, and lim-
implanted and compressed into the bone ends of ited local stress shielding are feasible.
a humeral fracture associated with a gunshot
wound. They suggested the use of orthogonal
plating to impart greater stability and allow for 15.5.2 Bone Ingrowth Potential
early motion [30].
The porosity of current tantalum implants has
been designed to optimize bone ingrowth
15.5 Metal Tantalum for Defects potential [37]. A recent in vivo study sought to
evaluate the interaction between human osteo-
The use of metal alloys for structural substitution blasts and porous tantalum and convincingly
is an atypical technique and reserved for situa- demonstrated that porous tantalum is a good
tions where regeneration is unfeasible, unlikely, substrate for the attachment, growth, and differ-
or the patient declines other techniques. Any of entiated function of human osteoblasts.
these applications would certainly be considered The current tantalum implants used for defects
“off-label” techniques since none of the currently are designed for bone defect management around
available tantalum devices are intended for joint replacements. While not primarily designed
trauma applications. for trauma, many of the shapes have been
Fig. 15.7 31-year-old male motorcyclist who suffered autologous and allogeneic bone graft. d–f The cage,
complex intraarticular distal femur fracture with extensive demonstration of packing the cage with bone graft, and a
bone loss. a–c Patient underwent initial Masquelet clinical photograph demonstrate the technique. (The cage
technique after extensive debridement and bone grafting images courtesy of Brian J. Cross, DO, Broward Health
was assisted by the integration of a titanium mesh cage Medical Center, Plantation FL, USA)
supported by screw fixation through the plate and massive
adaptable to the shape of common diaphyseal This size defect will require multiple implants
and meta-diaphyseal defects (Fig. 15.9). used end to end but allows for full defect
reconstruction and immediate weight bearing
(Fig. 15.10).
15.5.3 Applications Metaphyseal. We have used tantalum most
frequently in the setting of metaphyseal bone loss
Tantalum can be used for reconstruction of dia- —for both complete and incomplete defects.
physeal defects of large size. Our experience has Metaphyseal reconstructions can be done with
been primarily in knee arthrodesis in conjunction either a plate of an intramedullary device for
with an intramedullary device for stabilization. stabilization. Defects can be modified to accept
Fig. 15.8 a Scanning electron micrograph of porous caused by crystal growth during the process of tantalum
tantalum showing the cellular structure formed by the deposition. c Photographs showing transcortical implants
tantalum struts. There is the occasional smaller opening or with small and large pore sizes (From Bobyn et al. [34]
portal that interconnects with the larger pores or cells. with permission of The British Editorial Society of Bone
b Higher power scanning electron micrograph of a single & Joint Surgery)
pore illustrating the surface microtexture on the struts
the flat ends of the implant, and load can be Outcomes. There are no published outcomes
applied. The most common indication for tanta- of using tantalum for defect reconstruction.
lum is critical size bone defect in patients who There are ongoing concerns about risk of infec-
have declined traditional approaches to bone tion with the use of tantalum in traumatic wounds
regeneration (bulk grafting or distraction osteo- and the required resection in the setting of a
genesis), patients who cannot comply with lim- fulminant infection. However, in theory, this
ited weight bearing postsurgically, and patients approach is not significantly different than the
with poor bone regeneration potential (elderly, currently utilized induced membrane technique,
systemically ill). Patients must be informed that which utilizes a PMMA spacer. If an infection
this is an off-label application. were to occur, there is likely formation of a
grafts in long bone loss. The fibula as a bony

anatomic unit is quite versatile as it is similar to the
radius and ulna in shape and size, can be used
intramedullary in the humerus and can even be
medialized to substitute for the tibia. The vascular
supply is from the peroneal artery and veins, which
provide a dual endosteal and periosteal supply
from both the nutrient artery and the
musculo-periosteal vessels [38]. It can be utilized
as a purely osseous, or with the overlying skin and
muscle depending on the amount of type of bone
and soft tissue loss associated with the injury. One
of the disadvantages of utilization of the vascu-
larized fibula is the small caliber of the bone,
which can be compensated for with the double
barrel technique that allows for the long fibula
donor (which can be a maximum of 26 cm in
length). Modifications such as this allow for broad
application of the graft with the only limitation
being the technical nature of the surgical harvest
and implantation requiring a skilled microvascular
surgeon [38, 39]. Outcomes of use of the vascu-
larized fibula in the upper extremity demonstrate
excellent incorporation at 3 months. In the lower
extremity the fibula can be applied to the foot and
ankle in standard fashion if there is an adequate
location for anastomosis outside the zone of injury,
Fig. 15.9 Multiple different tantalum implants that can
but in the tibia different techniques might be used
be fashioned for critical bone loss substitution
(Fig. 15.11). Medialization of the fibula to substi-
tute for segmental bone loss can be performed
primarily or with Ilizarov techniques, but must at
vascularized scar response around the implant all times account for the condition of the soft tis-
that could ultimately be grafted. sues [40].
Rib. Defects of the clavicle are rare but difficult to
heal lesions that can be associated with long-
15.6 Tissue Transfer standing nonunions. They occur in patients who
often have had multiple failed procedures and there
is no single answer to the reconstruction of these
Except for distraction osteogenesis, the other defects. Traditionally tricortical iliac crest with
techniques described above involve the use of compression has been the standard of care with
bulk bone grafting that provides an avascular variable success rates. The advantage of this
technique is the relative simplicity and ability to
healing zone that requires creeping substitution reconstructs small defects to equalize the affected
with cells migrating from the intact bone through clavicle to the length of the contralateral side.
the matrix. The risk of nonunion, fracture of the Larger defects may benefit from both a vascular-
transplanted bone, and overall poor microarchi- ized bone graft and compression. A free pedicled
transfer is less than ideal in this region of the body.
tecture of the healed environment places the Free transfer of a vascularized rib pedicle has been
patient at high risk. This can be obviated by utilized for mandibular, maxillary and extremity
distraction osteogenesis or vascularized defects (tibial, calcaneal and humeral) [41]. This
tissue/bone transfer. The maintenance of perios- graft has also been studied in a rotational manner
for the clavicle where a serratus anterior flap is
teal and endosteal blood supply allows for heal- taken with the seventh and eighth rib and tunneled
ing and remodeling through both the vascular under the pectoral musculature then embedded into
pedicle and the local supporting vasculature with the debrided clavicle with compression fixation.
osteoblast induction. This is done in a double barreled fashion that
allows for adequate strength [42]. In a few case
Fibula. The vascularized fibula (pedicled or free) studies this has demonstrated long term success
is the most well-studied of all vascularized bone and although technically challenging this
Fig. 15.10 This patient had purulent infection after replacement of beads with a solid spacer, he had removal
ORIF distal tibia fracture with articular extension. a, b of his antibiotic spacer, placement of an intramedullary
He underwent radical resection of infection and had nail through a tantalum spacer, and autogenous cancellous
antibiotic beads placed after moderate bone resection was grafting. d At 5 months, he is free of infection with full
required. c After 6 weeks of intravenous antibiotics and incorporation of tantalum implant
technique can provide both mechanical and bio- reports, but all identify the value of a vascularized
logic advantages for the patient. graft particularly in the post-radiation, recalcitrant
Medial Femoral Condyle. The medial femoral nonunions, and necrotic bone loss patients.
condyle has been extensively studied for small
defects as it is taken as a cortico-periosteal graft
that is supplied by the descending genicular artery.
Traditionally this was described as a thin, 15.7 Summary
non-biomechanically strong graft that is easily
isolated and transferred into a site of defect with Nonunion care requires significant thought and
cancellous bone harvested most commonly from
precision with respect to achieving a sterile zone
the iliac crest. The harvest site is highly reliable
and it provides an answer for small areas of of injury along with an adequate understanding
necrotic or missing bone (Fig. 15.12). The average of the causative factor in a failure of bony
size of flap is 5 cm in length; however; recent regeneration. The occurrence of bone loss is not
studies have demonstrated the potential harvest site
at all uncommon with respect to nonunion, and
to be as much as 13 cm [43, 44].
when critical cortical defects occur beyond 4 cm,
Other sources of vascularized bone transport special techniques must be employed to achieve
will likely be identified in the future. There have complete reconstruction and return to function
been limited reports of others including a vascu- for patients.
larized pelvic flap for calcaneal substitution [45]. For massive defects, especially in the setting
These reports are limited case series or single case of current or prior infection, distraction
Fig. 15.11 55-year-old female injured in a fall from a transfer to achieve both soft tissue coverage and a
burning building, with a complex open fracture disloca- talonavicular arthrodesis. f–g Radiographs at 1.5 years
tion of the hindfoot. Complete traumatic loss of the talar demonstrate bony healing with a nonantalgic gait and
head at the talonavicular joint. a–e Primary subtalar complete return to function
arthrodesis performed along with vascularized free fibula
osteogenesis remains as the technique of choice to critical size defects or cannot comply with
for regeneration. weight bearing limitations. Vascularized bone
For large diaphyseal or metaphyseal defects, transfer is an alternative at centers that have
especially with no evidence of infection and microvascular expertise and in settings where the
stable fixation constructs, an induced membrane local blood supply will likely not support vigorous
technique with cancellous grafting can be safely osteogenesis.
utilized. Metal cages can be used during recon- BMP with allograft cancellous chips remains
struction to contain the cancellous graft, provide an alternative for small to medium size defects,
stability benefit, and potentially improve graft especially in the setting of a diaphyseal defect
efficiency. treated with an intramedullary rod. However, the
More unique approaches can be utilized in quality of bone regenerate created by this approach
more challenging cases. Metal substitution is an and potential local inflammatory consequences
alternative when patients decline other approaches limit the use of BMPs as a first-line approach.
Fig. 15.12 The medial femoral condyle-free vascular- useful in the distal aspects of the extremities. (Image
ized graft is ideal for small areas of necrotic bone. The courtesy of John S. Reach, Jr., MSc, MD, Yale University
harvest site is easily identifiable (a) and yields a graft that School of Medicine, New Haven CT, USA)
can be interposed into areas of bony loss (b), particularly
11. De Bastiani G, Aldegheri R, Renzi BL. Dynamic

References axial fixation. A rational alternative for the external
fixation of fractures. Int Orthop. 1986;10(2):95–9.
1. Stevenson S, Emery SE, Goldberg VM. Factors 12. Marsh JL, Nepola JV, Meffert R. Dynamic external
affecting bone graft incorporation. Clin Orthop Relat fixation for stabilization of nonunions. Clin Orthop
Res. 1996;324:66–74. Relat Res. 1992;278:200–6.
2. Masquelet AC, Begue T. The concept of induced 13. Gulabi D, Erdem M, Cecen GS, Avci CC, Saglam N,
membrane for reconstruction of long bone defects. Saglam F. Ilizarov fixator combined with an
Orthop Clin North Am. 2010;41(1):27–37. intramedullary nail for tibial nonunions with bone
3. Taylor BC, French BG, Fowler TT, Russell J, loss: is it effective? Clin Orthop Relat Res. 2014;472
Poka A. Induced membrane technique for recon- (12):3892–901.
struction to manage bone loss. J Am Acad Orthop 14. Mahboubian S, Seah M, Fragomen AT,
Surg. 2012;20(3):142–50. Rozbruch SR. Femoral lengthening with lengthening
4. Harshwal RK, Sankhala SS, Jalan D. Management of over a nail has fewer complications than intramedul-
nonunion of lower-extremity long bones using lary skeletal kinetic distraction. Clin Orthop Relat
mono-lateral external fixator–report of 37 cases. Res. 2012;470(4):1221–31.
Injury. 2014;45(3):560–7. 15. Lee DH, Ryu KJ, Song HR, Han SH. Complications
5. Keating JF, Simpson AH, Robinson CM. The of the intramedullary skeletal kinetic distractor
management of fractures with bone loss. J Bone (ISKD) in distraction osteogenesis. Clin Orthop
Joint Surg Br. 2005;87(2):142–50. Relat Res. 2014;472(12):3852–9.
6. Aronson J. Limb-lengthening, skeletal reconstruc- 16. Kirane YM, Fragomen AT, Rozbruch SR. Precision
tion, and bone transport with the Ilizarov method. of the PRECICE internal bone lengthening nail. Clin
J Bone Joint Surg Am. 1997;79(8):1243–58. Orthop Relat Res. 2014;472(12):3869–78.
7. Codivilla A. On the means of lengthening, in the 17. Schiedel FM, Vogt B, Tretow HL, Schuhknecht B,
lower limbs, the muscles and tissues which are Gosheger G, Horter MJ, et al. How precise is the
shortened through deformity. 1904. Clin Orthop PRECICE compared to the ISKD in intramedullary
Relat Res. 1994;301:4–9. limb lengthening? Reliability and safety in 26
8. Putti V. The operative lengthening of the femur. procedures. Acta Orthop. 2014;85(3):293–8.
1921. Clin Orthop Relat Res. 1990;250:4–7. 18. Oh CW, Apivatthakakul T, Oh JK, Kim JW,
9. Green SA. Ilizarov method. Clin Orthop Relat Res. Lee HJ, Kyung HS, et al. Bone transport with an
1992;280:2–6. external fixator and a locking plate for segmental
10. Catagni MA. Treatment of tibial nonunions. In: tibial defects. Bone Joint J. 2013;95-B(12):
Bianchi Maiocchi A, editor. Treatment of fractures, 1667–1672.
nonunions, and bone loss of the tibia with the 19. Viateau V, Bensidhoum M, Guillemin G, Petite H,
Ilizarov method. 4th ed. Milan: Medi Surgical Video; Hannouche D, Anagnostou F, et al. Use of the
2007. p. 1–8. induced membrane technique for bone tissue
engineering purposes: animal studies. Orthop Clin segmental defect: description of a new technique and
North Am. 2010;41(1):49–56. report of two cases. J Orthop Trauma. 2000;14
20. Pelissier P, Masquelet AC, Bareille R, Pelissier SM, (1):54–9.
Amedee J. Induced membranes secrete growth 33. Levine BR, Sporer S, Poggie RA, Della Valle CJ,
factors including vascular and osteoinductive factors Jacobs JJ. Experimental and clinical performance of
and could stimulate bone regeneration. J Orthop Res. porous tantalum in orthopedic surgery. Biomaterials.
2004;22(1):73–9. 2006;27(27):4671–81.
21. Viateau V, Guillemin G, Bousson V, Oudina K, 34. Bobyn JD, Stackpool GJ, Hacking SA, Tanzer M,
Hannouche D, Sedel L, et al. Long-bone critical-size Krygier JJ. Characteristics of bone ingrowth and
defects treated with tissue-engineered grafts: a study interface mechanics of a new porous tantalum
on sheep. J Orthop Res. 2007;25(6):741–9. biomaterial. J Bone Joint Surg Br. 1999;81(5):907–
22. Viateau V, Guillemin G, Calando Y, Logeart D, 14.
Oudina K, Sedel L, et al. Induction of a barrier 35. Bobyn JD, Toh KK, Hacking SA, Tanzer M, Kry-
membrane to facilitate reconstruction of massive gier JJ. Tissue response to porous tantalum acetab-
segmental diaphyseal bone defects: an ovine model. ular cups: a canine model. J Arthroplasty. 1999;14
Vet Surg. 2006;35(5):445–52. (3):347–54.
23. Masquelet AC, Fitoussi F, Begue T, 36. Heiner AD, Brown TD, Poggie RA. Structural
Muller GP. Reconstruction of the long bones by the efficcay of a novel porous tantalum implant for
induced membrane and spongy autograft. Ann Chir osteonecrosis grafting. Trans Orthop Res Soc.
Plast Esthet. 2000;45(3):346–53. 2001;26:480.
24. Stafford PR, Norris BL. Reamer-irrigator-aspirator 37. Bobyn JD, Pilliar RM, Cameron HU, Weatherly GC.
bone graft and bi Masquelet technique for segmental The optimum pore size for the fixation of
bone defect nonunions: a review of 25 cases. Injury. porous-surfaced metal implants by the ingrowth of
2010;41(Suppl 2):S72–7. bone. Clin Orthop Relat Res. 1980;150:263–70.
25. McCall TA, Brokaw DS, Jelen BA, Scheid DK, 38. Soucacos PN, Kokkalis ZT, Piagkou M, Johnson EO.
Scharfenberger AV, Maar DC, et al. Treatment of large Vascularized bone grafts for the management of
segmental bone defects with reamer-irrigator-aspirator skeletal defects in orthopaedic trauma and recon-
bone graft: technique and case series. Orthop Clin structive surgery. Injury. 2013;44:S70–5.
North Am. 2010;41(1):63–73. 39. Soucacos PN, Korompilias AV, Vekris MD, Zou-
26. Aho OM, Lehenkari P, Ristiniemi J, Lehtonen S, bos A, Beris AE. The free vascularized fibular graft
Risteli J, Leskela HV. The mechanism of action of for brisdging large skeletal defects of the upper
induced membranes in bone repair. J Bone Joint Surg extremity. Microsurgery. 2001;31:190–7.
Am. 2013;95(7):597–604. 40. Rahimnia A, Fitoussi F, Pennecot G, Mazda K.
27. Eck KR, Bridwell KH, Ungacta FF, Lapp MA, Treatment of segmental bone loss of the tibia by
Lenke LG, Riew KD. Analysis of titanium mesh tibialisation of the fibula: a review of the literature.
cages in adults with minimum two-year follow-up. Trauma. 2011;16(4):154–9.
Spine (Phila Pa 1976). 2000;25(18):2407–15. 41. Bhathena HM, Savant DN, Kavarana NM,
28. Hertlein H, Mittlmeier T, Piltz S, Schurmann M, Parikh DM, Sanghvi VD. Reconstruction with
Kauschke T, Lob G. Spinal stabilization for patients different free flaps in oro-facial cancer patients. Acta
with metastatic lesions of the spine using a titanium Chir Plast. 1996;38(2):43–9.
spacer. Eur Spine J. 1992;1(2):131–6. 42. Brenner P, Zwipp H, Rammelt S. Vascularized
29. Ostermann PA, Haase N, Rubberdt A, Wich M, double barrel ribs combined with free serratus
Ekkernkamp A. Management of a long segmental anterior muscle transfer for homologous restoration
defect at the proximal meta-diaphyseal junction of of the hindfoot after calcanectomy. J Trauma.
the tibia using a cylindrical titanium mesh cage. 2000;49(2):331–5.
J Orthop Trauma. 2002;16(8):597–601. 43. Pelzer M, Reichenberger M, Germann G.
30. Attias N, Lehman RE, Bodell LS, Lindsey RW. Osteo-periosteal-cutaneous flaps of the medial
Surgical management of a long segmental defect of femoral condyle: a valuable modification for selected
the humerus using a cylindrical titanium mesh cage clinical scenarios. J Rec Microsurgery. 2010;26
and plates: a case report. J Orthop Trauma. 2005;19 (5):291–4.
(3):211–6. 44. Iorio ML, Masden DL, Higgins JP. The limits of the
31. Attias N, Lindsey RW. Case reports: management of medial femoral condyle corticoperiosteal flaps.
large segmental tibial defects using a cylindrical J Hand Surg Am. 2011;36(10):1592–6.
mesh cage. Clin Orthop Relat Res. 2006;450:259–66. 45. Kurvin LA, Volkering C, Kebler SB. Calcaneus
32. Cobos JA, Lindsey RW, Gugala Z. The cylindrical replacement after total calcanextomy via vascularized
titanium mesh cage for treatment of a long bone pelvis bone. Foot Ankle Surg. 2008;14:221–4.
Index
A Bone loss, 2, 27, 323 See also Nonunions

Acetabular nonunion .See Pelvic nonunions Bone Marrow Aspiration (BMA), 31, 46, 50
Adjunct therapy, 35 Bone Morphogenetic Proteins (BMPs), 1, 31, 32, 106,
Adjuvant therapy, 233 117, 329
Aging, 2, 45, 62 Bone regeneration, 323
Alcoholism, 66, 96, 289 Bone Specific Alkaline Phosphatase (BSAP), 20
Angiopoietin-dependent pathway, 56 Bone turnover markers, 20
Ankle injury treatment, 7
Antegrade nails, 229, 243, 248
Anterior Superior Iliac Spine (ASIS), 183 C
Arthroplasty, 108 Cage technique, 330
functional restoration, 222 Capacitively Coupled Electric Field (CCEF), 38
proximal femur fracture nonunions Carpal bones
femoral neck, 215 hamate fractures, 172
for failed osteotomy, 216 scaphocapitate fracture syndrome, 171
proximal humerus nonunions scaphoid fracture, 157
options, 111 anatomy and biomechanics, 158
reverse total shoulder, 110 clinical evaluation, 159
unconstrained, 108 treatment, 161
Atrophic nonunion, 17, 100, 103, 128, 174, 187, 230, scaphoid with osteonecrosis, 165
295, 311, 315 postoperative management, 169
Augmentation scaphoid without osteonecrosis, 163
proximal humerus nonunions SNAC, 169
biologic augmentation, 106 Cast stabilization, 2, 4
structural graft augmentation, 104 Cathepsin K (CK), 20
supracondylar humeral nonunions, 117 Clavicle fracture nonunion
Augmentative plate fixation, 231 classification, 77
Autogenous bone graft, 30, 31, 104, 250, 279 complication, 75, 90
distal femur fracture nonunion, 249 distal, 75, 76
ORIF, clavicle fracture nonunion epidemiology, 77
innovative method, 85 midshaft, 75, 77
literature review, 85 rates of, 78
multicenter and randomized controlled trials, 86 risk factors, 76, 78
surgical method, 81 treatment
Autogenous cancellous bone graft, 147, 151 nonoperative, 76, 80
Autograft bone grafting, 108 operative interventionSee also Operative
Autologous bone graft, 117, 219, 278, 297, 303 intervention
Avascular nonunions, 20, 21 symptomatic, 80
Clavicle fractures, 75
Claviculectomy, 80
B Combined Magnetic Fields (CMF), 38
Bicondylar tibial plateau fracture, 282, 284 Compartment syndrome, 4
Biologic augmentation, 106 Computed Tomography (CT) scan, 14, 16
Bisphosphonates, 9, 67, 68, 208, 222 distal femur fracture nonunion, 256
Blade plate, 243 distal radioulnar joint instability, 154

A. Agarwal (ed.), Nonunions, DOI 10.1007/978-1-4939-7178-7
340 Index
distal radius fracture nonunion, 146 salvage procedure, 249

femoral shaft nonunions, 229 TKA, 249
proximal tibial fracture nonunion, 276 Distal humeral nonunions. See Supracondylar humeral
scaphoid fracture nonunion, 159 nonunions
supracondylar humeral nonunions, 115 Distal RadioUlnar Joint (DRUJ), 125
tibial shaft fracture nonunions, 291 Distal radioulnar joint instability, 151
Contact healing, 46 anatomy and biomechanics, 151
Corticocancellous graft, 164 clinical evaluation, 153
C-Reactive Protein (CRP), 17, 115, 187, 247, 251, 291 CT scan, 154
MRI, 154
post operative management, 157
D radiographic evaluation, 154
Delayed union, 1 treatment
Demineralized Bone Matrix (DBM), 117 acute DRUJ, 154
Diabetes mellitus, 6, 65, 128, 246 chronic DRUJ, 155
Diaphyseal bone healing, 57 Distal radius fracture nonunion, 143
Direct fracture healing, 46 anatomic alignment, 144
contact healing, 46 classification systems, 144
gap healing, 46 clinical evaluation
lamellar bone regeneration, 46 CT scan, 146
rigid stability, 46 laboratory analysis, 145
Disabilities of the Arm, Shoulder and Hand (DASH), 104 patient history, 145
Distal femur fracture, 243 physical examination, 145
Distal femur fracture nonunions radiographic examination, 146
case studies, 251 incidence, 143
evaluation and diagnosis management, 143
CT scan, 247 post operative care, 151
laboratory analysis, 247 radiographic parameters, 144
MRI, 247 risk factors, 143
nuclear imaging, 247 treatment, 144
patient medical history, 246 DRUJ salvage procedures, 148
physical examination, 246 nonoperative, 146
radiographs, 247 surgical, 147
incidence, 243 Distal tibia and ankle nonunions
ramification, 244 amputation, 321
risk factors, 244 analysis, 309
comminuted injuries, 245 arthodesis, 321
infection, 246 atrophic nonunions, 315
locked plating, 244 classification, 311
obesity, 246 clinical evaluation, 313
open fractures, 245 delayed/true nonunion, 310
osteoporosis, 246 external fixation, 317
smoking, 246 hypertrophic nonunions, 314
soft tissue damage, 244 infected nonunions, 317
treatment intramedullary nailing, 320
antegrade nails, 248 plating techniques, 317
atrophic nonunions, 249 Distraction osteogenesis, 303, 323
author’s preferred methods, 250 Dorsal distraction bridge plating, 149
double plate fixation with autologous bone graft, Dual-Energy X-ray Absorptiometry (DEXA) scan, 251
248 Dual plating technique, 147
external fixation, 248 Dynamic Compression Plating (DCP), 85
fixed angle implant, 248 Dynamic Condylar Screw (DCS), 243
hypertrophic nonunions, 249 Dynamization, 33, 230, 238
infected nonunions, 249
intramedullary nailing, 248
literature review, 247 E
ORIF, 248 Electrical stimulation, 38
plate fixation, 248 Electromyography (EMG), 98
prosthetic replacement, 249 Endochondral ossification, 46
Index 341
Endocrinopathies muscle, 126

distal femur nonunions, 247 proximal radioulnary joint, 128
tibial shaft fracture nonunions, 290 anticipated outcomes/recurrence, 132
Erythrocyte Sedimentation Rate (ESR), 17, 187, 247, 251, case study, 132
291 patient variables, 132
Exchange nailing, 231 risk factors, 128
External fixation, 4, 5, 16, 232, 250, 275, 301, 302, 317, surgical treatment
324, 326 approach and exposure, 130
ExtraCorporeal Shock Wave Therapy (ECSW), 38, 305 bone preparation, 131
essentials of exposure, 131
operating room setup, 130
F preoperative planning, 129
Femoral neck fracture nonunion, 12 technical points, 132
proximal femur fracture Fracture healing, 69
anatomic reduction, 210 acute inflammation, 65
angle of fracture line, 209 assessment
arthroplasty, 215 clinical criteria, 61
arthroplasty for failed osteotomy, 216 fracture union, 61
fixation, 212 radiologic scores, 61
incidence, 209 biomechanics
neglected fracture, 216 biologic healing response, 58
surgical options, 210 fracture gap and hydrostatic pressure, 58
valgus osteotomy, 212 interfragmentary strain, 58
varus collapse, 209 role in skeletal repair, 58
vascularized graft, 214 timing, 59
RUSH score, 13 comorbidities
Femoral shaft nonunions, 227, 233 aging, 62
case study, 233, 236 endocrine disorders, 65
classification, 229 habits, 66
diagnosis medications, 67
CT scan, 229 metabolic bone disease, 62
radiographs, 229 direct
epidemiology, 228 contact healing, 46
etiology, 227, 228 gap healing, 46
evaluation operative fixation, 46
laboratory analysis, 229 rigid stability, 46
patient history, 229 indirect, 46
physical examination, 229 bone regeneration, 47
treatment, 230 bridge plating, 49, 55
adjuvant therapy, 233 external fixation, 49, 53
augmentative plate fixation, 231 inflammatory phase, 49
dynamization, 230 micromotion, 47
exchange nailing, 231 remodeling phase, 57
external fixation, 232 reparative phase, 49
IM nails, 227 metaphyseal, 57
intramedullary nails, 230 physiology, 45
Masquelet technique, 232 process, 1
plate fixation, 231 Fracture stabilization, 5, 292, 294
Fibular strut allograft, 104 Functional bracing, 155, 299
Fine wire circular fixation, 324, 325
Fixation methods, 46
Fixed angle devices, 243, 245 H
Flexor Carpi Radialis (FCR) tendon, 163 Hand and wrist nonunions, 143 See also Carpal bones;
Fluoroscopy, 16, 149, 279 Distal radioulnar joint instability; Distal radius
Forearm nonunion fracture nonunion; Metacarpal and phalangeal
anatomy, 125 nonunions
bone, 125 Hemiarthroplasty, 103, 106, 108–110, 218, 220
DRUJ, 127 High Intensity Focused Ultrasound (HIFUS), 305
interosseous membrane, 127 Horse hoof/foot nonunion, 21
342 Index
Humanitarian Device Exemption (HDE), 106 M

Humeral shaft fracture, 37 Macrophage Colony-Stimulating Factor (M-CSF), 49, 57
Huntington procedure, 304 Magnetic Resonance Imaging (MRI), 16
Hyperparathyroidism, 65 osteomyelitis, 292
Hypertrophic nonunion, 21, 22, 33, 99, 128, 174, 229, proximal humerus nonunions, 100
249, 293, 300, 311, 314 scaphoid fracture nonunion, 160
Hypervascular nonunions, 20 tibial shaft fracture nonunions, 291
Masquelet technique, 232, 298, 323, 327
bone grafting timing, 330
I features, 328
Indirect fracture healing induced m, 327
bone regeneration, 47 induced membrane, 327, 328
bridge plating, 49 outcomes, 329
external fixation, 49 two-stage technique, 329
inflammatory phase, 49 Matrix Metalloproteinases (MMPs), 56
micromotion, 47 Matti-Russe procedure, 164
remodeling phase Medical comorbidity, fracture healing
activation and regulation, 57 aging, 62
woven bone resorption, 57 endocrine disorders
reparative phase diabetes mellitus, 65
angiogenesis, 56 hyperparathyroidism, 65
characteristics, 49 habits
hard callus formation, 56 alcohol consumption, 66
MSC recruitment, 50 smoking, 66
revascularization, 56 medications
soft cartilaginous callus formation, 55 bisphosphonates, 67
Indium scan, 19 NSAIDs, 67
Infected tibial nonunions, 299 PTH, 69
adequate soft tissue coverage, 295 metabolic bone disease. See Osteoporosis, 62
antibiotics, 295 Megaprosthesis, 250
bone defect restoration, 296 Mesenchymal Stem Cells (MSCs), 1, 9, 50, 51
osteomyelitis, 294, 295 Metabolic disorders, 19
surgical debridement, 295 Metacarpal and phalangeal nonunions
treatment atrophic nonunion, 174
external fixation, 295 hypertrophic nonunions, 174
fracture stabilization, 295 radiographic evidence, 173
Inflammatory cytokines, 56 risk factor, 173
1,2 Intercompartmental Supraretinacular Artery (ICSRA), treatment, 174
166 Metal tantalum device, 331
Interosseous Membrane (IOM), 125 applications, 332
Intertrochanteric fracture nonunion, 217 bone ingrowth potential, 331
Intramedullary (IM) nails, 227, 230 material, 331
Intramedullary (IM) nails distraction Micromotion, 56
device plus external fixation, 326 Musculoskeletal injury, 45
telescopic intramedullary, 326
with plate osteosynthesis, 327
Intramedullary Skeletal Kinetic Distractor (ISKD), 326 N
Neck Resorption Ratio (NRR), 14
Neurofibromatosis, 301
L Nicotine, 10, 66, 289
Less Invasive Stabilization System (LISS), 243, 275 Nonsteroidal Anti-inflammatory Drugs (NSAIDs), 9, 67,
Locking Compression Plates (LCP), 245 290
Locking Condylar Plate (LCP), 245 Nonunion Risk Determination (NURD) Score, 15
Locking plate technology, 108 Nonunions
Long bone nonunions, 2 and biological markers, 3
Lower Extremity Assessment Project (LEAP), 66 classification
Low Intensity Pulsed Ultrasound (LIPUS), 35, 37, 305 and scoring systems, 22
Index 343
avascular, 20, 21 IM nails .See Intramedullary (IM) nails distraction,

based on radiographic evaluation, 22 326
hypervascular, 20 Masquelet technique .See Masquelet technique, 327
motion amount at fracture site, 22 metal tantalum for defects See also Metal tantalum
pseudarthrosis, 20, 21 device, 331
costly burden, 2 tissue transfer, 334
criteria and timepoints for, 2 Nuclear Imaging, 100, 247, 291
definition, 1, 20, 61
diagnosis, 10
evaluation of O
extremity examination, 3 Oligotrophic nonunions, 21, 30, 230, 293, 315
injury mechanism, 3 Open Reduction and Internal Fixation (ORIF), 4, 80, 275
patient medical history, 3 clavicle fracture nonunion
physical examination, 3 with autogenous bone graft .See Autogenous bone
preoperative planning, 3 graft
previous surgery critics, 4 without bone grafting, 86
prior complications, 4 distal femur fracture nonunions, 243
laboratory evaluation Operative intervention
serologic markers, 20 clavicle fracture nonunion
WBC, CRP and ESR, 19 IM pinning, 88
management principles, 22 intramedullary fixation, 80
adjunct therapy, 35 ORIF .See Open reduction and internal fixation
limb mechanical environment, 35 (ORIF)
local biological environment, 27 ORIF, 88
local mechanical environment, 33 reconstructive procedures, 81
systemic biological environment, 27 resection, 89
quality of life, 2 vascularized bone graft reconstruction, 88
radiographic evaluation tibial shaft fracture nonunions
bone, orthogonal views of, 11 adjunct treatments, 303
criteria, 11 bone graft substitutes, 303
CT scan, 16 bone grafting, 302
fluorodexoglucose PET scan, 19 compression plating, 300
fluoroscopy, 16 electrical technqiues, 305
MRI scan, 16 external fixation, 301
nuclear medicine tests, 16, 19 fibular osteotomy, 302
oblique radiograhs, 11 growth factors, 303
PET/CT scan, 19 intramedullary nailing, 300
plain radiographs, 11 ultrasound .See Ultrasound, 305
RUST, 12 Osteoblasts, 46, 57
ultrasonography, 16 Osteocalcin (OC), 20, 57
risk factors Osteoclasts, 46, 56, 57
biological and mechanical factors, 4 Osteoporosis, 8, 62
inflammation, 9 Osteoprotegerin (OPG), 20, 57
medications, 9 Osteosynthesis
open fractures, 4, 5 of distal femur fracture, 244
patient factors, 6 plate, 227, 230, 231, 287, 300, 331
patient medical factors, 6 proximal humerus nonunions
smoking, 9, 10 ORIF with blade plates, 104
surgeon factors, 5 ORIF with standard plates, 103
scoring system
fracture line visibility, 12
NURD score, 15 P
presence/absence of callus, 12 Parathyroid Hormone (PTH), 8, 69, 293
RUSH score, 12, 14 Peg graft, 104
special techniques for traumatic bone loss, 335 Pelvic nonunions
cage technique, 330 case study, 196
distraction osteogenesis, 323 clinical assessment and pain, 183
fine wire circular fixation, 324 deformity, 185
344 Index
ilioinguinal approach, 194 rate of, 95

Injuries, 185 patient evaluation
malunions and displaced nonunions, 195 clinical examination, 98
painful nonunions, 192 laboratory analysis, 100
patient expectations, 185 radiographic imaging, 99
Pfannenstiel incision, 192 risk factors
radiographic assessment, 187 distracting forces, 98
rami factures, 192 glenohumeral joint arthrosis, 97
results, 195 loss of bone mineral density, 96
sacral nonunions, 195 metabolic bone disease, 96
treatment, 187 metaphyseal comminution, 96
Perren’s strain theory, 58 nutritional deficiency, 96
Pfannenstiel incision, 194 premature rehabilitation, 97
Plate fixation, 231, 295, 315 smoking, 96
Platelet Derived Growth Factor (PDGF), 49 soft tissue interposition, 97
Platelet-Rich Plasma (PRP), 31 translated surgical neck fractures, 96
Plate osteosynthesis, 227, 230, 231, 287, 300, 331 surgical timing, 101
Plating techniques, 103 treatment, 111
Poly(methyl methacrylate) (PMMA) block, 328 augmentation .See Augmentation, 104
PRECICE intramedullary nail, 326 avascular necrosis, 103
Preoperative antibiotics, 19 interlocking intramedullary nails, 107
Prostaglandins, 9, 67 internal fixation devices, 108
Proximal femur fracture nonunions, 222 non-operative, 102
diagnosis, 207 osteosynthesis .See Osteosynthesis, 103
femoral neck surgical, 103
anatomic reduction, 210 unreamed intramedullary rods, 106
angle of fracture line, 209 Proximal Radioulnar Joint (PRUJ), 125
arthroplasty, 215 Proximal tibial fracture nonunion, 281
arthroplasty for failed osteotomy, 216 case study, 283
fixation failure, 209 diagnosis, 276
incidence, 209 CT scan, 276
neglected fracture, 216 MRI, 276
revision fixation, 212 nuclear medicine examinations, 276
surgical options, 210 surgeon factors, 276
valgus osteotomy, 212 epidemiology, 275
varus collapse, 209 outcomes, 280
vascularized graft, 214 treatment, 277
intertrochanteric fracture nonuninon, 217 functional brace, 276
physical examination, 207 locked plating, 283, 284
risk factors Poller screws, 276
biologic factors, 208 Pseudarthrosis, 21, 66
medications, 208 Pulsed Electromagnetic Fields (PEMF), 38
surgeon factors, 209
subtrochanteric nonunion, 220
timing and adequacy of, 207 R
Proximal femur fractures Radiographic Union Score for Hip (RUSH), 12, 61
device used in, 207 Radiographic Union Score for Tibia (RUST), 12, 61
occurrence, 207 Radionucleotide bone scans, 150
Proximal humerus fractures, 111 Reamer-Irrigator/Aspirator (RIA) system, 30, 250, 298
incidence, 95 Receptor Activator of Nuclear factor kB Ligand
prevalence, 95 (RANKL), 20, 57
Proximal humerus nonunions, 111 Retrograde nailing, 262
arthroplasty, 108 Reverse total shoulder arthroplasty, 110
options, 111
reverse total shoulder, 110
unconstrained, 108 S
classification, 98 Scaphocapitate fracture syndrome, 171
epidemiology Scaphoid Nonunion Advanced Collapse (SNAC), salvage
incidence, 95, 96 procedures for, 169
Index 345
Serologic markers, 61 instability, 288

Smoking, 1, 2, 9, 66, 96, 128, 132, 145, 208, 246, 247, medical comorbidities, 289
260, 289 medications, 290
Social issues and nonunions, 3 open fracture, 289
Soft fracture callus, 38 poor blood supply, 288
Staphylococcus aureus, 290 smoking, 289
Stiff nonunions, 314 treatment, 292
Structural grafts, 304 nonoperative, 299
Subtrochanteric femoral region, 1 nonunion type
Subtrochanteric nonunion, 220 atrophic, 293
Supracondylar humeral nonunions hypertrophic, 292
complications, 118 infected nonunions, 294
laboratory analysis, 115 location, 296
physical examination, 115 oligotrophic, 293
postoperative rehabilitation, 117 prior failed treatment, 299
radiographic imaging, 115 segmental defects, 297
risk factors, 115, 118 operative intervention .See Operative intervention,
surgical approaches, 118 299
augmentation, 117 Tibial shaft fractures, 15, 230, 287
posterior incision, 117 Tissue transfer
total elbow arthroplasty, 117 medial femoral condyle, 335
triceps splitting, 117 rib, 334
ulnar nerve transposition, 117 risk factor, 334
symptoms, 115 vascularized fibula, 334
treatment, 116 Total elbow arthroplasty, 117
Surgical neck fracture, 107 Total Hip Arthroplasty (THA), 197, 219
Total Knee Arthroplasty (TKA), 250
Transforming Growth Factor-beta (TGF-(b)), 20
T Triangular Fibrocartilage Complex (TFCC), 144
Tartrate-Resistant Acid Phosphatase (TRAcP), 20 Tumor Necrosis Factor-alpha (TNF-a), 49
Taylor Spatial Frame (TSF), 35
Technetium-99m-pyrophosphate scan, 100
Teriparatide, 69 U
Three-part proximal humerus fracture, 105 Ulnar bone graft, 168
Tibial plateau fracture nonunion Ulnocarpal joint stress testing, 153
case study, 282 Ultrasonography (US), 16
epidemiology, 275 Ultrasound, 35
outcomes, 281 fracture healing, 16
treatment, 277 three-dimensional, 16
external fixation, 275 tibial shaft fracture nonunions
LISS, 275 ECSW, 305
minimally invasive plating, 275 LIPUS, 305
Tibial shaft fracture nonunions, 2
diagnosis
CT scan, 291 V
MRI, 291 Valgus Osteotomy, 212
nuclear imaging, 291 Vascular Endothelial Growth Factor (VEGF), 20, 51
radiographs, 291 -dependent pathway, 56
evaluation Vascularized bone graft reconstruction, 88
laboratory analysis, 291 Vascularized Graft, 214
patient history, 290 Vitamin D deficiency, 1–3, 8, 27, 208, 247
physical examination, 291
incidence, 287
ramifications, 288 W
risk factors, 288 White Blood Cell (WBC) count, 17, 187, 251, 291
endocrionopathies, 290 Wrist arthrodesis, 150
infection, 290
346 Index
X Y
X-ray, initial injury, 234 Young and active individuals, 78
X-ray, post-operative, 237 primary fixation, 76
X-rays, plain, 187 trauma in, 75
Young and active patients, recent study, 80

Nonunions

Uploaded by

Copyright:

Available Formats

Nonunions

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Nonunions

Uploaded by

Copyright:

Available Formats

Animesh Agarwal

ISBN 978-1-4939-7176-3 ISBN 978-1-4939-7178-7 (eBook)

Library of Congress Control Number: 2017940364

© Springer Science+Business Media LLC 2018

Printed on acid-free paper

This Springer imprint is published by Springer Nature

It is an honour to be invited to write the foreword to this important contri-

Manchester, VT, USA James D. Heckman, MD

An investment in knowledge pays the best interest.

Nonunions have been a challenging aspect for orthopedic trauma surgeons.

Boerne, TX, USA Animesh Agarwal, MD

15 Special Techniques for Nonunions Associated

Animesh Agarwal MD Department of Orthopedics, Division of Orthopedic

Michael P. Leslie DO Department of Orthopedics and Rehabilitation, Yale

There are however many factors to take into

© Springer Science+Business Media LLC 2018 1

Fig. 1.1 Patient with right

Fig. 1.2 Patient with a long

Fig. 1.3 Patient with a right

Fig. 1.4 (a–c. Computed

Fig. 1.10 Patient with

1.6.1 Biological Environment: options for bony defects are discussed in

interdependent processes; disruption in key steps

© Springer Science+Business Media LLC 2018 45

2.2.1.2 Indirect Fracture Healing anatomical reduction. Rather, micromotion, to an

Fig. 2.3 Cutting cones. Low

b Fig. 2.8 Secondary healing with bridge plating. The

Table 2.1 Cytokines and their roles in fracture healing

Cartilaginous callus formation is driven by bone formation require adequate oxygen

2.2.2 Biomechanics of Fracture histomorphometric and ﬁnite element analyses

2.2.2.1 Interfragmentary Strain 2.2.2.2 Fracture Gap

2.2.3.2 Radiologic Scores

Table 2.2 Calculation of Score per cortex Callus Fracture line

Fig. 2.12 Radiographic

2.3.2.2 Alcohol Consumption

2.3.3 Medications 115, 116]. Aspirin, at doses equivalent to

difference in time to union, or in radiographic or The short-term results of bisphosphonate use

callus has formed. In an article published in the

© Springer Science+Business Media LLC 2018 75

Orthopedic Trauma Unit at the Royal Inﬁrmary of

3.9.2 An Innovative Method In a series comparing the use of dynamic com-

3.12 Vascularized Bone Graft

3.14 Conclusion 11. Robinson CM, Cairns DA. Primary nonoperative

4.1 Introduction 4.2 Epidemiology

© Springer Science+Business Media LLC 2018 95

4.7.4.1 ORIF with Standard Plates

4.7.5.2 Biologic Augmentation and 10 with hemiarthroplasty based on an intra-

4.7.7 Interlocking Intramedullary surgical stabilization using an interlocking

infection as a cause of their nonunion. Screening

© Springer Science+Business Media LLC 2018 115

Fig. 5.1 Anteroposterior

associated with unsuccessful humeral condylar frac-

tion appropriately. The forearm axis of rotation

© Springer Science+Business Media LLC 2018 125

Subsequently, Anderson et al. [31] reported on Preoperative planning includes, if needed:

6.4.4 Essentials of Exposure The surgeon should avoid extensive subperiostal

Technique for bone preparation: 6.5 Patient Variables

direct blow to his ulnar forearm, resulting in a Case 3 (Fig. 6.3)

Case 2 (Fig. 6.2) Case 4 (Fig. 6.4)

cement spacer and insert a corticocancellous procedure. At that point, an intra-operative