Nuclear Medicine
Nuclear Medicine
Nuclear Medicine
BIOSCIENCE V ad e me c u m
Nuclear
Medicine
William D. Leslie
I. David Greenberg
v a d e m e c u m
Nuclear Medicine
LANDES
BIOSCIENCE
GEORGETOWN, TEXAS
U.S.A.
VADEMECUM
Nuclear Medicine
LANDES BIOSCIENCE
Georgetown, Texas U.S.A.
ISBN: 1-57059-644-1
While the authors, editors, sponsor and publisher believe that drug selection and dosage and
the specifications and usage of equipment and devices, as set forth in this book, are in accord
with current recommendations and practice at the time of publication, they make no
warranty, expressed or implied, with respect to material described in this book. In view of the
ongoing research, equipment development, changes in governmental regulations and the
rapid accumulation of information relating to the biomedical sciences, the reader is urged to
carefully review and evaluate the information provided herein.
Contents
1. An Introduction to Nuclear Medicine ............................... 1
Brian Lentle and Anna Celler
Introduction ............................................................................................... 1
History ....................................................................................................... 1
Comparative Imaging and the Role of Nuclear Medicine ............................ 4
Radionuclide Production ............................................................................ 5
Radionuclide Decay .................................................................................... 8
Detection Systems .................................................................................... 10
Clinical Practice ........................................................................................ 13
A Perspective on the Future ...................................................................... 14
2. Radiation Effects and Safety ........................................... 16
Michael J. Chamberlain
Introduction ............................................................................................. 16
Radiation Dosimetry ................................................................................ 16
Radiation Effects and Carcinogenesis ........................................................ 19
Principles of Radiation Protection ............................................................. 24
Practical Aspects of Radiation Protection .................................................. 25
Frequently Asked Questions (FAQs) ......................................................... 28
3. Myocardial Perfusion Imaging ........................................ 31
Robert Corne and I. David Greenberg
Introduction ............................................................................................. 31
Physiologic and Technical Considerations ................................................. 31
Clinical Role in the Diagnosis of Coronary Artery Disease ........................ 41
Clinical Role in Prognosis and Risk Stratification ..................................... 50
Clinical Role in Defining Myocardial Viability ......................................... 54
Frequently Asked Questions (FAQs) ......................................................... 55
4. Equilibrium Radionuclide Angiocardiography ............... 60
I. David Greenberg and Robert Corne
Introduction ............................................................................................. 60
Technical Considerations .......................................................................... 60
Clinical Applications ................................................................................. 64
Frequently Asked Questions (FAQ)s ......................................................... 72
5. Thromboembolic Disease ............................................... 75
Daniel F. Worsley and Philip S. Wells
Introduction ............................................................................................. 75
Technical Considerations in Lung Scanning ............................................. 75
Diagnosis of Acute Pulmonary Embolism ................................................. 83
Evaluation of Pulmonary Hypertension .................................................... 88
Frequently Asked Questions (FAQs) ......................................................... 88
6. Bone Densitometry ......................................................... 93
William D. Leslie and Bruce E. Roe
Pathophysiology of Bone Loss and Osteoporotic Fractures ........................ 93
Technical Aspects of Bone Densitometry .................................................. 97
Clinical Role of Bone Densitometry ....................................................... 104
Clinical Management of Osteoporosis .................................................... 110
Frequently Asked Questions (FAQs) ....................................................... 117
7. Skeletal Disorders ......................................................... 121
Leonard Rosenthall and Peter MacDonald
Introduction ........................................................................................... 121
Skeletal Anatomy and Physiology ........................................................... 121
Technical Considerations ........................................................................ 121
Trauma ................................................................................................... 122
Osteomyelitis .......................................................................................... 129
Vascular Disorders .................................................................................. 131
Joint Prostheses ....................................................................................... 134
Radionuclide Synovectomy ..................................................................... 137
Frequently Asked Questions (FAQs) ....................................................... 138
8. Skeletal Oncology ......................................................... 141
Leonard Rosenthall and Ralph Wong
Introduction ........................................................................................... 141
Primary Benign Bone Tumors ................................................................. 141
Primary Malignant Bone Tumors ............................................................ 149
Diagnosis and Follow-Up of Skeletal Metastases ..................................... 152
Frequently Asked Questions (FAQs) ....................................................... 159
9. Kidney .......................................................................... 163
Michael Hoskinson and Keevin Bernstein
Introduction ........................................................................................... 163
Renal Physiology ..................................................................................... 163
Technical Considerations ........................................................................ 165
Clinical Role in Acute Renal Failure ........................................................ 172
Clinical Role in Hydronephrosis ............................................................. 174
Clinical Role in Renovascular Hypertension ........................................... 179
Clinical Role in the Renal Transplant Patient .......................................... 188
Frequently Asked Questions (FAQs) ....................................................... 195
10. Gastrointestinal ............................................................. 196
Peter Hollett and Ford Bursey
Introduction ........................................................................................... 196
Clinical Role in Esophageal Motility Disorders ....................................... 196
Clinical Role in Gastric Motility Disorders ............................................. 199
Clinical Role in the Localization of Gastrointestinal Bleeding ................. 203
Clinical Role of Urea Breath Testing ....................................................... 206
Frequently Asked Questions (FAQs) ....................................................... 208
11. Hepatobiliary Imaging .................................................. 211
Reinhard Kloiber and Gary R. May
Introduction ........................................................................................... 211
Radiopharmaceuticals ............................................................................. 211
Clinical Role in the Evaluation of the Biliary Tree ................................... 214
Clinical Role in the Characterization of Liver Masses .............................. 225
Frequently Asked Questions (FAQs) ....................................................... 231
12. Inflammatory Disorders ................................................ 233
William D. Leslie and Pierre Plourde
Pathophysiology of Inflammation ........................................................... 233
Technical Considerations ........................................................................ 234
Clinical Role: General Principles ............................................................. 244
Clinical Role in Fever of Unknown Origin (FUO) .................................. 249
Clinical Role in Vascular Graft Infection ................................................. 252
Frequently Asked Questions (FAQs) ...................................................... 255
13. Thyroid Disorders ......................................................... 260
Albert A. Driedger and Thomas J. McDonald
Thyroid Anatomy and Physiology ........................................................... 260
Technical Aspects of Thyroid Scintigraphy .............................................. 262
Thyrotoxicosis ........................................................................................ 265
Hypothyroidism ..................................................................................... 269
Thyroid Nodules .................................................................................... 270
Thyroid Cancer ...................................................................................... 270
Frequently Asked Questions (FAQs) ....................................................... 275
14. Radionuclide Therapy of Thyroid Disorders ................ 276
Albert A. Driedger and Thomas J. McDonald
Introduction ........................................................................................... 276
Benign Thyroid Disorders ....................................................................... 276
Follicular Cell-Derived Thyroid Cancers ................................................. 282
Frequently Asked Questions (FAQs) ....................................................... 294
15. Tumor Imaging ............................................................. 297
A.J.B. McEwan
Introduction ........................................................................................... 297
Mechanisms of Radiopharmaceutocal Uptake ......................................... 297
Radiopharmaceuticals Used in Cancer Management ............................... 299
Contributions of Nuclear Medicine to Cancer Imaging .......................... 305
Radioisotope Therapy ............................................................................. 333
Frequently Asked Questions ................................................................... 336
16. Neuropsychiatric Disorders ........................................... 340
Jean-Paul Soucy, Denis Lacroix and Catherine Kissel
Introduction ........................................................................................... 340
Regional Cerebral Perfusion .................................................................... 340
Energy Metabolism and Neurotransmission Studies ................................ 350
Cerebrospinal Fluid Assessment .............................................................. 353
Intracranial Mass Lesions ........................................................................ 358
Conclusions ............................................................................................ 360
Frequently Asked Questions (FAQs) ....................................................... 361
17. Pediatric Nuclear Medicine ........................................... 365
David Gilday
Introduction ........................................................................................... 365
Technical Considerations ........................................................................ 365
Clinical Role in the Assessment of Childhood
Musculoskeletal Disorders ....................................................................... 367
Clinical Role in Childhood Malignancies ................................................ 370
Clinical Role in Neonatal Jaundice ......................................................... 373
Clinical Role in Rectal Bleeding .............................................................. 376
Clinical Role in Genitourinary Disorders ................................................ 378
Frequently Asked Questions (FAQs) ...................................................... 382
Appendix ....................................................................... 384
Half-lives and prinicipal emissions from common radionuclides ............. 384
Effective dose from common radiologic and nuclear
medicine procedures ............................................................................... 385
Index ............................................................................. 386
Editors
William D. Leslie, MD, FRCPC, ABNM, MSc
Associate Professor of Medicine and Radiology
University of Manitoba
Winnipeg, Manitoba, Canada
Chapters 6 and 12
Contributors
Keevin Bernstein Albert A. Driedger
Associate Professor of Medicine Professor of Nuclear Medicine
University of Manitoba and Oncology
Winnipeg, Manitoba, Canada University of Western Ontario
Chapter 9 London, Ontario, Canada
Chapters 13 and 14
Ford Bursey David Gilday
Associate Professor of Medicine Professor of Radiology
Memorial University of Newfoundland University of Toronto
St. Johns, Newfoundland, Canada Chapter 17
Chapter 10
Peter Hollett
Anna Celler Professor of Radiology
Medical Imaging Research Group Memorial University of Newfoundland
Nuclear Medicine St. Johns, Newfoundland, Canada
University of British Columbia Chapter 10
Vancouver, British Columbia, Canada
Chapter 1 Michael Hoskinson
Nuclear Medicine
Michael J. Chamberlain University of Alberta
Professor of Radiology Edmonton, Alberta, Canada
University of Ottowa Chapter 9
Ottawa, Ontario, Canada
Chapter 2 Catherine Kissel
Associate Professor of Medicine
Universit de Montreal
Robert Corne Montreal, Quebec, Canada
Associate Professor of Medicine Chapter 16
and Radiology
University of Manitoba
Winnipeg, Manitoba, Canada
Chapters 3 and 4
Reinhard Kloiber Pierre Plourde
Clinical Professor of Radiology Associate Professor of Medical
University of Calgary Microbiology
Calgary, Alberta, Canada University of Manitoba
Chapter 11 Winnipeg, Manitoba, Canada
Chapter 12
Denis Lacroix
Assistant Professor of Psychiatry Bruce E. Roe
Universit de Montreal Associate Professor of Medicine
Montreal, Quebec, Canada University of Manitoba
Chapter 16 Winnipeg, Manitoba, Canada
Chapter 6
Brian Lentle
Emeritus Professor Leonard Rosenthall
Radiology Professor of Radiology
University of British Columbia McGill University
Vancouver, British Columbia, Canada Montreal, Quebec, Canada
Chapter 1 Chapter 7, 8
William D. Leslie
I. David Greenberg
Comprehensive References
1. Harbert JC, Eckelman WC, Neumann RD, eds. Nuclear Medicine Diagnosis and Therapy.
New York: Thieme Medical Publishers, Inc., 1996.
2. Wagner HN, Szabo Z, Buchanan JW, eds. Principles of Nuclear Medicine. Second Ed.
Philadelphia: W.B. Saunders Company, 1995.
3. Murray ICP, Ell PJ, Strauss HW, eds. Nuclear Medicine in Clinical Diagnosis and
Treatment. Second Ed. Edinburgh: Churchill Livingstone, 1994.
4. Sandler MP, Patton JA, Coleman RE, Gottschalk A, Wachers FJT, Hoffer P. Diagnostic
Nuclear Medicine. Third Ed. Baltimore: Williams & Wilkins, 1996.
5. Henkin RE, Boles MA, Dillehay GL, Halama JR, Karesh SM, Wagner RH et al, eds.
Nuclear Medicine. St. Louis: Mosby, 1996.
CHAPTER 1
CHAPTER 1
the stem of the plant with an instrument capable of detecting and measuring radio-
activity. This use of radioactive atoms, present in minute amounts but acting as a
marker of other, non-radioactive atoms came to be called the tracer principle. It
only required that Hevesys insight be translated to people instead of plants, and for
the tracer to be administered by injection instead of through a plants root system,
for the power of nuclear medicine to become clear.
4 Nuclear Medicine
related to the function of an organ. The term nuclear medicine encompasses sev-
eral different imaging techniques ranging from positron emission tomography (PET)
and single photon emission computed tomography (SPECT) through whole body 1
planar images or scans.
There are several ways in which the analysis of nuclear medicine data can be
done ranging from a simple and qualitative visual inspection of planar images up to
a full numerical and quantitative analysis of the three- or even four-dimensional
(including temporal) data sets. Creation of these quantitative images is quite complex
and usually requires application of several corrections to the data (attenuation, scatter,
normalization, for example) as well as iterative reconstruction methods. Also the use
of quantitative analytical methods usually involves kinetic modeling and sophisticated
computer-based operations. The information obtained from such analyses can be
directly related, however, to physiological processes and may provide a very useful
and comprehensive picture of a disease. At present this type of data analysis is available
mostly in centers with strong research programs because only the most advanced
nuclear medicine systems using modern image reconstruction techniques are able to
realize fully quantitative data. Therefore, diagnostic applications of this approach
remain in development.
The relative sensitivities of PET and MRI for the detection of metabolic changes
in vivo are such that PET can detect concentrations of metabolites several orders of
magnitude smaller than those detectable with MRI. Thus, while functional and
anatomic imaging are converging, the methods each have strengths and weaknesses
which suggest that both will have a role to play in the future. Indeed there is growing
interest not only in fusing anatomical and functional images but also in obtaining
such images with hybrid technology combining, for example, PET and CT.
Radionuclide production
Radioactive atoms (radionuclides) are fundamental to the tracer principle. Thus
their production is an important step in the clinical practice of nuclear medicine.
The radionuclides used in imaging usually emit gamma (a) rays. Occasionally a
particle is emitted as well. X-rays and a-rays are both part of the electromagnetic
spectrum. Visible light, radiowaves and microwaves are also part of this spectrum.
However, X-rays and a-rays are of short wavelength and thus are energetic and can
penetrate tissues. The penetrating power of X-rays and a-rays is a function of their
energy usually measured in electron voltsthe gamma rays from technetium-99m
are of 140 thousand electron volts (140 keV). While a-rays are in general more
energetic than X-rays the real distinction between them stems from their origins: a-
rays are produced in nuclear decay processes whereas X-rays derive from orbital
electron perturbations produced, for example, by electrons accelerated in an X-ray
tube.
Radionuclides may be created by one or other of the following methods.
Cyclotron Irradiation
Radionuclides may be made by irradiation of stable atoms with cyclotron-
accelerated particlesusually protons. This method of production is especially
important for those radioactive atoms used in PET since these usually have very
short half-lives and must be manufactured where they are to be used. The reaction
6 Nuclear Medicine
Figure 3. Beta-minus decay with the emission of a a-ray (iodine-131 decays in this
way).
can be represented as follows, where the oxygen atom is stable, p is the accelerated
proton, n is the neutron produced and fluorine-18 is the resulting tracer (half-life
about 2 hours):
18
O + p = 18F + n
In this reaction oxygen has eight protons and fluorine nine in the respective
nuclei. The superscripts used throughout this volume refer to the atomic mass of the
atom, essentially the sum of the protons and neutrons in the atomic nucleus since
electrons and other particles have negligible mass.
Reactor Irradiation
Atomic nuclei may be made radioactive by the flux of neutrons in a nuclear
reactor or from the fission of heavier atoms. Molybdenum, for example, may be
made by either of the following reactions:
98
Mo (n, a) A 99Mo (low specific-activity)
235U (n, fission) A 99Mo (high specific activity)
In these reactions n stands for a neutron and a for a gamma ray. Specific activity
is a measure of the fraction of radioactive atoms present of the total in the sample
important to consider in the preparation of some radiopharmaceuticals.
Reactor produced atoms are often rich in neutrons and thus decay by `- emissions
(e.g., iodine-131) (Fig. 3). `- particles (energetic electrons) give large local radiation
doses which may be destructive. This may make them important in therapy as in the
use of radioactive iodine to treat Graves disease and thyroid cancer.
Generator Production
A common strategy in nuclear medicine practice is to take delivery, at a hospital
or clinic, of a generator containing a long-lived precursor of a short-lived daughter
isotope. The precursor may be made in either a reactor or cyclotron. Molybdenum-
99/technetium-99m generators (Fig. 4) are very widely used in nuclear medicine,
the molybdenum most often being reactor produced. Technetium-99m has many
useful features (short half-life, no particulate emissions to cause large radiation
exposures to patients and a a-ray energy ideal for gamma-camera detection) and is
An Introduction to Nuclear Medicine 7
more widely used than any other radionuclide at present. Yet with a half-life of
about six hours it would not be readily available on the scale on which it is used were
it not for its availability from a generator. The generator contains a long-lived
molybdenum-99 parent absorbed onto a column and, as this radionuclide decays,
technetium (which being different chemically is not so absorbed) is eluted (milked)
from the column in the generator on a daily or twice-daily basis. The molybdenum
decays as follows:
99 99 m <
MoA Tc + ` + i
and the technetium used in preparing the radiotracer then decays as follows (Fig. 5):
99mTc A 99Tc + a
Radionuclide Decay
Some nuclei are unstable and decay at a rate described by the half-life (the time
taken for 50% of the nuclei of a given radioactive sample to decay). The life expectancy
of an individual atom is impossible to predict but in the large numbers in which
they are produced the whole-population characteristic half-life can be described by
bulk averaging. Radioactive instability is related to the excess of energy contained in
a given excited nucleus and often results from an imbalance of the numbers of
protons and neutrons in the nucleus. Radioactive atoms decay by a number of
processes each with different implications for nuclear medicine practice. Decay
processes may be classified according to whether in the atoms in question the
imbalance leads to a neutron-rich nucleus (usually reactor produced) or proton-rich
nucleus (usually accelerator produced). The commoner forms of decay are described.
Half-lives and principal emissions from common radionuclides are summarized in
Appendix I.
Electron (or Beta-) Particle Emission (Fig. 3)
Neutron-rich atoms decay by the transformation of a neutron into a proton and
electron. The proton remains in the nucleus but the electron is emitted, and for
historical reasons, it is in this context called a `-minus particle:
<
nA p+e +i
where the n is a neutron, p a proton, e - the negative electron (beta particle) and
v an antineutrino (an undetectable and nearly mass-less particle). In this reaction
An Introduction to Nuclear Medicine 9
the daughter atom has the same mass number as the parent but an increase in one of
the atomic number because of the increase in number of protons in the nucleus.
Isomeric Transition (Gamma Decay) (Fig. 5) 1
This method involves an internal rearrangement of the nuclear structure with
minimal change in atomic weight. However, an alteration in the energy state of the
nucleus results in the emission of a-rays. Such a-rays are very energetic forms of
electromagnetic radiation, like light, and lose little of their energy in the body. Thus
little radiation damage results to the tissues while the gamma rays are well-suited to
imaging the distribution in the body of the physiological molecule-of-interest to
which they are attached. For such reasons technetium-99m which decays by this
mechanism is widely used in nuclear medicine, the decay schematic being as follows:
99m
Tc A 99Tc + a
Electron Capture
If a nucleus is not sufficiently energetic to decay by positron emission (see below)
it may capture an orbital electron. A proton is then transformed into a neutron and
a neutrino emitted. Since a vacancy is created in the inner electron shell, this is filled
from outer rings and a succession of so-called characteristic X-rays result (characteristic
because of their specific and recognizable energies). Examples used in nuclear medicine
are indium-111 and iodine-123. The decay schematic is as follows:
p + + e- A n + i
Internal Conversion
This type of decay occurs in parallel to a-decay. It is the result of an energetic
radioactive nucleus transferring its energy to an orbital electron which is ejected,
rather than a a-ray. The result is again characteristic X-rays (as the orbital vacancy is
filled) and the electron (called an Auger electron) of discrete energy. This process is
particularly important in calculations of radiation doses resulting from radioactive
decay.
Positron (Beta+) Emission
Positrons (or positive electrons) are an example of anti-matter so beloved of science
fiction writers. However, positrons have a very important role in nuclear medicine.
They result from the decay of proton-rich nuclei. It so happens that the only externally
detectable isotopes of carbon, hydrogen, oxygen and nitrogen (which make up a
major part of bodily tissues) are carbon-11, oxygen-13 and nitrogen-15, all proton
rich. All three decay by positron-emission albeit with very short half-lives. Add in
fluorine-18, which can substitute for hydroxyl groups and which is also a positron-
emitting radionuclide, and it is apparent that detection of positron emissions might
be a very powerful tool for studying disease. Indeed positrons might also be used to
study the mechanisms of disease and the behavior and localization of important
molecules in the body.
Positrons do not decay themselves but are extremely short-lived. When they lose
their kinetic energy after collision(s) with electrons they finally meet a negative electron
and the particles mutually annihilate. The energy associated with their rest mass
appears as two high-energy electromagnetic photons (each of an energy of 0.511
10 Nuclear Medicine
Figure 6. Positron decay and the principle of positron emission tomography. The
positron, after a short path and scattering off negative electrons, interacts with such
an electron. As both annihilate, their rest mass results in two photons detected as
coincidental events in the detector ring.
all combine detectors with electronic amplification and analysis of the signal. The
detectors in gamma cameras (Fig. 7) are scintillators made of sodium iodide crystals,
whereas in positron emission detection bismuth germinate is often used but other
materials are being explored for both applications including semi-conductor detection
systems. A discussion of the relative advantages and disadvantages of these materials
is beyond the scope of this chapter.
The electronic processing firstly involves pulse height analysis. This determines
if the parcel of energy associated with each detected a-ray of electromagnetic radiation
corresponds with the energy anticipated knowing the radionuclide injected. If it
does not, and it might, for example, come from cosmic radiation or the naturally
occurring radioactive tracer potassium-40 present in each of us, then that signal is
rejected as having the potential to degrade the image. Secondly the electronic
processing involves giving the signal an address which describes the coordinates of
the interaction in the crystal. The collimator in the imaging system makes the image
coherent much as the lens in a camera focuses light. The vertical perforations in the
collimator between the patient being studied and the crystal make the radiation
12 Nuclear Medicine
reaching the crystal reflect the distribution of the nuclide in the patients body. In
this way an image of that distribution can be built up.
1 It is important to note that since the data acquired are inherently digital (the
positional address is computed) then nuclear medicine readily lends itself to
quantitative methods. Several specific detection methods exist.
Static Imaging
The gamma camera is used to image a single organ within its field-of-view, such
as heart or lung, after injection of an appropriate tracer. An example is a map of lung
perfusion obtained after injection of 99mTc-protein aggregates which trap in lung
capillaries.
Dynamic Imaging
This technique is often combined with static imaging. The arrival and uptake of
tracer in an organ as well as its washout may be imaged or analyzed from repeated
images taken over a span of time.
Gated Imaging
To arrest body motion during the time it takes to make an image, and thus
reduce image blurring, the image may be gated by linking image acquisition to
particular times in the cardiac or respiratory cycles. Images from the same segment
of many such cycles are thus combined into what is in effect a cin loop of the organ
in motion.
Whole-Body Imaging
Among other imaging techniques involving exposure to ionizing radiation, such
as computed radiography, the radiation exposure to the patient increases with each
additional image made. This is not the case in nuclear medicine. In nuclear imaging
the radiation exposure is determined by the injection of the tracer and additional
images take time but do not otherwise expose the patient to risk. Radionuclide
imaging is a particularly powerful method to search the whole body for disease, the
distribution of which is unknown. Examples are bone scans done to detect metastases,
tumor scans (for example with 18F-fluoro-deoxyglucose [FDG]) and scans with labeled
white cells to detect occult infections.
Region-of-Interest (ROI) Analysis
Because, as we have seen, the data acquired in nuclear medicine images are
inherently digital, it is easy to obtain quantitative information about organ function.
From the gamma camera image the organ, or a part of it, is defined by a computer-
generated outlinethe region-of-interestand the activity within this area
measured, either as total uptake or rate-of-uptake in an activity-time plot. At present
much research is being done to ensure that these measurements are accurately
corrected for scattered radiations and attenuation of activity because of the depth of
the organ in the body.
An Introduction to Nuclear Medicine 13
99m. No stable form of this element is found on earth and it has no role in human
metabolism. However, it is cheaply available from a generator, has advantageous
1 physical properties as noted above, and has proved to be the work-horse for nuclear
medicine for the past three decades. This fact owes a great deal to the innovations by
radiopharmacists in finding ways to label biologically interesting molecules with
technetium-99m.
A Perspective on the Future
Diagnostic imaging, by any method, does not exist in isolation but must respond
to the changing context in which medicine is practiced. Several trends in healthcare
stand out as the twentieth century gives way to the twenty-first:
The high cost of care, and societal imperatives to contain such costs, and re-
align spending for other social purposes;
The focus on the patient (sometimes called the client in this context) as a
partner in health promotion rather than as the passive recipient of care;
The publics increasing interest in and use of complementary medicine -
either traditional methods such as acupuncture, or new techniques such as Bach
flower therapy. By implication, this interest seems to be a measure of public skepticism
about allopathic medicine and certainly represents a use of resources that might
otherwise be used in conventional care;
A further change in the traditional relationships between physicians and patients
dictated by the accessibility of information, good and bad, about health from sources
such as the world-wide-web;
Increasing emphasis on the public health and the social determinants of illness
as distinct from the medical view of illness;
A requirement that any medical interventions be evidence-based. Again modern
information technology will impact on this as decision support tools are developed
and on-line records become a tool for audit and outcomes analyses; and
Against this pattern of social change in the context in which medicine is
practiced, medicine itself is also poised on the threshold of revolution. The insights
afforded by molecular biology and the unfolding of the human genome project are
about to change forever the human view of disease and the ability to treat it.
Nuclear medicine methods are, by the standards of technology-intensive
medicine, relatively low-cost, safe and minimally invasive as well as often able to be
done for people as out-patients. This makes them likely to be important to the
future of care.
Nuclear medicine clearly will not and should not be expected to respond to
every change in the social context in which medicine is practiced. Nevertheless it is
capable of being made user-friendly. At the same time images are intrinsically an
effective way to communicate with patients and might be used more often for this
purpose.
In the longer view the movement to an evidentiary basis for practice will serve
nuclear medicine well given its rich tradition of intellectual inquiry. A considerable
literature has already emerged, for example, to show that positron emission
tomography with 18F-fluoro-deoxyglucose is, while expensive, both a cost-effective
technique in cancer diagnosis and staging and one which positively influences patient
outcomes.
An Introduction to Nuclear Medicine 15
Additional Reading
The following articles supply more information on the roots of nuclear medi-
cine: 1
1. Brucer M. A Chronology of Nuclear Medicine. St. Louis: Heritage Publications
Inc., 1990:piii.
2. Brucer M. Nuclear Medicine Begins with a Boa Constrictor. New York: Society of
Nuclear Medicine, 1979:v-xxvi.
3. Cohen M. Ernest Rutherford at McGill University. In: Aldrich JE, Lentle BC, eds.
A New Kind of Ray: The Radiological Sciences in Canada. 1895 - 1995. Montreal:
The Canadian Association of Radiologists, 1995.
4. Levi H. George Hevesy and his concept of radioactive indicatorsIn retrospect.
Eur J Nucl Med 1976; (1):3-7.
5. Rntgen WC. On a new kind of rays (English translation). Nature 1896;
53:274-276.
6. Mould RF. Discovery of radioactivity and radium. In: A Century of X-rays and
Radioactivity in Medicine. London: The Institute of Physics Publishing, 1993:10.
CHAPTER 2
Figure 1. Linear energy transfer (LET). Individual ionization events (small circles)
occur along paths followed by the radiation (lines). The large circles contain equal
numbers of ionizations for low LET and high LET radiation. In the case of high LET
radiation, all energy deposition occurs along a single short ionization track With
low LET radiation it takes a larger number of radiation emissions as each one has a
relatively small chance of tissue interaction.
Dose equivalent was developed in an effort to incorporate biology into the phys-
ics of radiation exposure. Not all forms of radiation (alpha, beta, neutron, gamma,
X-ray) produce the same biological effect for equal absorbed doses. For example,
neutron irradiation has five times the biologic effect of X-rays while alpha radiation
has twenty times the biologic effect of X-rays. This is denoted by the Quality Factor
(1 for X-ray, gamma and beta irradiation; 5 for neutron and 20 for alpha irradiation).
The dose equivalent is the absorbed dose multiplied by the radiations Quality Factor.
For instance, 0.01 Gy (1 rad) of beta radiation produces a dose equivalent of 0.01 Sv
(1 rem) while 0.01 Gy (1 rad) of alpha radiation produces a dose equivalent of 0.2
Sv (20 rem).
Effective Dose
In order to be able to compare different radiation doses with differing forms of
radiation affecting different parts of the body it is necessary to express each radiation
dose as if it had been distributed evenly throughout the body. This uses information
on the radionuclides biodistribution, tissue weighting factors which adjust for the
varying susceptibility of different tissues and organs to radiation damage, and quality
factors which reflect the RBE of different forms of ionizing radiation. A radiation
dose expressed in this way is termed an effective dose equivalent (or simply an effective
dose). These principles are illustrated in the effective doses shown in Table 3 which
compare a thyroid radioiodine uptake (assuming a 24-hour uptake of 25%) with a
bone scan (assuming normal renal function). Despite considerable differences in the
radiation emissions, activities and biodistributions (critical organs) it is possible to
see that the bone scan should result in about twice the biologic effect of the uptake
Radiation Effects and Safety 19
Absorbed dose 1 gray (Gy) = 1 1 rad = 100 erg per Gy = 0.01 x rad
Joule per kg gram rad = 100 x Gy
measurement. Effective dose from common radiologic and nuclear medicine proce-
dures are summarized in Appendix 2.
Background Radiation
Living on Earth has always meant continuous exposure to natural background
radiation including cosmic radiation, radiation from naturally-occurring minerals
in natural surroundings and construction materials as well as internal radiation from
radionuclides within our own bodies. Finally there is the inevitable but highly variable
inhalation of the ubiquitous radioactive gas radon, a decay product of the uranium-
238 which is a component of the Earths crust.
In Eastern North America typical natural background radiation is in the order of
1.0 mSv per year with exposure to radon contributing a further 1-2 mSv. Natural
background radiation varies by a factor of five to eight times from place to place in
the world and the variation is even greater when differences in inhaled radon are
taken into account. Numerous large surveys have compared cancer incidence and
deaths in high and low background regions in an attempt to measure the presumed
carcinogenic effect of low level radiation. No association between cancer incidence
and level of background radiation has been found. It is ironic that for radiation
safety purposes we may struggle to regulate and reduce radiation exposures which
are less than those which may be incurred by moving from one city to another
because of the geographic variation in background radiation levels!
20 Nuclear Medicine
when treating painful bony metastases with an agent such as radioactive strontium-
89 bone marrow depression may occur as a side effect.
With stochastic effects the probability of the event increases with radiation dose
with the assumption that there is no threshold (i.e., no dose too small not to have an
effect). In practice there may well be a point below which the effect is imperceptible
or so small as to be negligible. Examples of stochastic effects are the increased 2
probability of developing leukemia or a solid cancer following radiation exposure.
Linear No Threshold Hypothesis
This hypothesis states that the stochastic dose/response effects (e.g., cancer in-
duction) observed at high doses and dose rates can be extrapolated in a straight line
passing through the origin (zero dose and zero effect). This implies that there is no
dose so low that it does not have an adverse effect even though no effect is directly
observable. This is a perfectly valid scientific and safety device providing it is appre-
ciated that it is not proven or provable, and that other hypotheses may fit the ob-
servable phenomena such as a linear/quadratic relationship between dose and effect
for low doses and even that there is indeed a threshold (Fig. 2).
Sources of Radiation
Along with natural background radiation there comes exposure to man-made
radiation. Some of this is unavoidable such as the fall-out from nuclear weapons
testing (now at very low levels), the generation of electricity by nuclear power, and
the radiation from luminous dials and signs. By far the greatest contribution to the
man-made radiation in the developed world comes from medical uses in diagnostic
radiology and nuclear medicine. When distributed over the adult North American
population this is approximately equivalent to two-thirds of natural background.
Radiation-Induced Cancer
The clinical presentation of leukemia induced by ionizing radiation is likely to
be delayed by several years from the time of exposure. In the case of solid tumours
the latent period may be several decades. It is believed that excess cancers are still
arising in Japanese atomic bomb survivors. The risk of inducing additional neoplasms
is generally quoted as the average for an adult population. The additional lifetime
risk of developing a fatal cancer is generally taken as 1 in 4,000 for an effective dose
of 10 mSv. Lifetime risk from radiation exposure is relatively less in elderly individuals
(due to limited life expectancy from other illness) and relatively greater in healthy
children.
Atomic bomb survivor data show an excess of leukemias and solid cancers over
the expected spontaneous occurrence but the excess of solid cancers only exceeds
95% confidence limits for victims with estimated doses in the range 50-250 mSv
and higher with no excess demonstrable for exposures below 50 mSv.
Effects of Ante-Natal (in utero) Radiation Exposure
The effects of radiation exposure of the conceptus depends upon the time of
exposure relative to conception (Table 4). Prior to the beginning of organogenesis
(three weeks after conception) damage to the small number of relatively
undifferentiated cells is most likely to result in failure of implantation or undetectable
death of the conceptus rather than a damaged liveborn child. When major
22 Nuclear Medicine
A B
Practice Point
Exclusion of pregnancy is particularly important when therapeutic
radiopharmaceuticals are to be administered. Serum beta-HCG becomes detectable
about two weeks after ovulation (around the time of implantation) and should be
tested whenever pregnancy cannot be reliably excluded otherwise. Particular tact
may be required when dealing with a thyrotoxic teenage candidate for radioiodine
therapy accompanied by a parent.
Genetic Disorders
Theoretically genetic disorders may occur in first and subsequent generation
children born to parents with radiation exposure prior to conception due to radiation-
induced mutation in gamete precursors. This effect has never been demonstrated to
occur in human populations (including the atomic bomb survivors) but its occurrence
24 Nuclear Medicine
is predicted from plant and animal work. By extrapolation, it has been estimated
that a doubling of the spontaneous mutation rate (non-radiation induced plus
background radiation induced) requires on the order of 50 to 250 mSv.
Principles of Radiation protection
2
Increased Distance from the Source
Increasing the distance from a point source of radiation causes the exposure rate
to drop off rapidly according to the Inverse Square Law. For example moving from
10 cm to 30 cm from a point source will decrease exposure rate by a factor of nine.
You can illustrate this effect by warming your hands at an open fire and then taking
a step backwards.
Practice Point
Increasing distance from the source brings major benefits in reducing radiation
exposure in nuclear medicine but the major source of radiation exposure, the patient,
is not a point source. Simple inverse square law calculations do not apply when the
distance from the patient is less than 3 m. This principle can be easily demonstrated
by taking a survey meter and measuring exposure rates at increasing distances from
a patient recently injected with a 99mTc agent for a bone or myocardial perfusion scan.
Minimize Time of Exposure
Carefully plan procedures to require the least time handling or in the vicinity of
a radiation source.
Shielding
Gamma and X-ray radiation is attenuated by the material through which it passes
to an extent determined by the energy of the radiation, the density of the material
and its thickness. The thickness of a shielding material required to reduce radiation
exposure from a given radionuclide by half is called its half value layer (HVL). The
HVL of lead for 99mTc is 0.3 mm whereas for the more energetic emission of 131I a
thickness of 3.0 mm is needed (Table 5). Lead is frequently used for shielding because
its high density minimizes the thickness and hence the volume occupied by the
shielding needed to provide a required attenuation. Where space is available, concrete,
compacted earth or water may provide effective and economical shielding.
Practice Point
The wearing of lead or lead equivalent (usually 0.25 mm lead equivalent) aprons
to shield the trunk is popular among technologists and mandated in some
jurisdictions. These are effective shields at typical diagnostic X-ray energies but the
theoretical attenuation achieved for 99mTc is < 50%, while for 131I it is negligible.
The risk of wearing the heavy device may not be negligible in terms of back strain.
Elements of distance, time and shielding are often used together to achieve the
desired reduction in radiation exposure. Occasionally over emphasis on one principle
can have a net negative effect. For example particularly heavy and clumsy shielding
of a radioactive source may increase the time needed to handle it and increase the
possibility of a spill.
Radiation Effects and Safety 25
99m 131
Tc I
Photon energy keV 140 364
HVL in tissue (cm) 4.6 6.3 2
HVL in lead (mm) 0.3 3.0
odds and dismisses the risk. Conversely the layperson concentrates on the dread
event, ignores the odds and is concerned by the risk. (If you doubt the above consider
the person who dreams the afternoon away planning how he will spend a lottery
prize while ignoring the odds against winning.)
Equivalent risks may be helpful in putting radiation risks in context but be warned
2 that your attempt to be helpful may be counterproductive. Expressing the risk of a
nuclear medicine procedure in terms of an equivalent number of chest radiographs
may be thought helpful in that the latter is more familiar to most patients. Anyone
told that a bone scan is equivalent to 60 chest X-rays or that a rest/stress myocardial
perfusion scan equates to 250 chest X-rays is unlikely to know that annual natural
background radiation is equivalent to 30 chest X-rays or focus on the safety of chest
X-rays rather than the apparent danger of the nuclear medicine procedure. Describing
the risk of nuclear medicine procedures in terms of cigarettes smoked, slices of cream
pie eaten or minutes of mountain climbing is more likely to make the physician
look foolish than to educate the patient. It may be more acceptable to relate risk to
that of a fatal accident when driving a car since that is such a familiar, recurrent and
difficult to avoid activity.
The risk of a fatal accident during a working lifetime in a safeoccupation, such
as a bank manager or sales assistant, is approximately 1 in 10,000. This is equivalent
to a 4 mSv radiation dose. For many people this will help to put a bone scan in
context. The Health Physics Society recommends against quantitative estimation of
health risks below an individual dose of 50 mSv in one year or a lifetime dose of 100
mSv in addition to background radiation.
Dose limits
Legislation changes over time and differs between countries and the reader should
consult local authorities for the current dose limits. Current radiation effective dose
limits in Canada, Britain and Europe are 1 mSv (0.1 rem) annually for the general
public and 20 mSv (2 rem) annually for someone who is occupationally exposed.
The corresponding limits for the United States are 1 mSv (0.1 rem) and 50 mSv (5
rem) respectively. At these limits it can be seen that nuclear medicine technologists
with a typical exposure of 1.5-2.0 mSv annually exceed the limit for the general
public but fall far short of that for the occupationally exposed.
Dose limits for pregnant women are less uniform. In Canada the dose limit for
the occupationally-exposed pregnant worker is 4 mSv (0.4 rem) effective dose to the
worker from the summed internal and external sources of radiation exposure for the
remainder of the pregnancy after the pregnancy has been declared. In the United
States the dose to an embryo/fetus during the entire pregnancy due to occupational
exposure of a declared pregnant woman must not exceed 5 mSv (0.5 rem). There is
no dose rate restriction on this maternal exposure. In Europe and the United Kingdom
the dose limits for the pregnant worker are 2.0 mSv to the surface of the abdomen
and 1.0 mSv to the fetus for the remainder of the pregnancy after the pregnancy has
been declared. The rate of fetal exposure must not exceed 0.06 mSv in any two week
period in the remainder of the pregnancy. In practical terms this means that pregnant
women who wish to carry out the normal work of a nuclear medicine technologist
may generally do so.
Radiation Effects and Safety 27
pregnancy?
Occasionally, despite all precautions, it will be discovered that a procedure has
been performed on a woman who is subsequently found to have been pregnant at
the time. The question of performing a therapeutic abortion may then be raised. In
such cases as much factual information as possible must be supplied as quickly as
possible to what will inevitably be an anxious situation. This will include calculation
2
of the radiation dose likely to have been received by the products of conception
(taking a worst case scenario), consideration of the stage of the pregnancy and the
threshold considerations for deterministic effects as discussed above. These must be
seen in the context of the rates of spontaneous occurrence of the effects it is feared
that radiation may induce and all the other risks of pregnancy. Consideration should
also be given to the opportunities for detection of fetal abnormalities by ultrasound
and amniocentesis. Rarely if ever will the additional risk to an unborn child resulting
from in utero irradiation alone justify interruption of the pregnancy. The nuclear
medicine physician can be a helpful resource to the woman, her family and her
obstetrician/gynecologist by dispassionately presenting them with the facts and the risks.
Why is explaining radiation risk so difficult?
Nuclear medicine professionals may sometimes think that radiation risks are
somehow unfairly magnified in comparison with all the other threats to continuing
health and happiness. They may wonder why they seem to be so bad at explaining
this. The following factors may have something to do with it.
1. The Hiroshima and Nagasaki legacy. The atomic bombing of these Japanese
cities and their largely civilian populations was such a dramatic and horrifying event
that the general perception of the scale of their long term radiation consequences is
greatly magnified. Consequently the risks of radiation generally are exaggerated..
Visitors to the Radiation Effects Research Foundation official web page may be
surprised by the figures for excess cancers and leukemias among atomic bomb
survivors with average exposure of 200 mSv The 428 excess cancer deaths attributable
to radiation in the survivor population studied represented 8.8% of the total cancer
deaths observed in 1950-1990.
2. The difficulty in explaining the small incremental risk of death from cancer
seen against a several thousand fold greater pre-existing risk.
3. Obsessional concentration on the feared (anticipated) event rather than the
remote odds of it happening (lottery syndrome).
4. The ALARA principle and linear no threshold hypothesis preclude a safe
radiation dose making it difficult to reassure an anxious patient
5. The difficulty in explaining stochastic risk.
6. Dependence on thermoluminescent dosimetry which provides only delayed
reassurance rather than using a direct reading device whenever concern over radiation
exposure is expressed.
7. Irresponsible colleagues who either treat radiation in a cavalier fashion
describing the radiation dose received from a rest/stress myocardial perfusion scan as
30 Nuclear Medicine
Treadmill Exercise
Graded exercise testing (GXT) in conjunction with myocardial perfusion imag-
ing is most commonly performed on a treadmill. GXT is preferred over pharmaco-
Figure 1. Cornoary artery plaque and plaque rupture. (A) Plaque is made up of a
lipid-core containing free lipid and foam cells covered by a fibrous cap denuded of
normal endothelium. (B) Unstable plaque may rupture exposing circulating blood
to underlying thrombogenic material with the consequent formation of a thrombus.
Figure 2. The effect of stenosis on mycocardial flow. At rest (curve A), flow is
maintained until a critical stenosis of approximately 90% is reached. With stress
(curve B), flow increases by a factor of 4. As the degree of stenosis increases, flow
is maintained until approximately a 50% stenosis is reached. In this example, at
rest there is no difference in flow between myocardium supplied by a normal artery
and that supplied by an artery with a 70% stenosis. However, at stress flow will be
reduced to myocardium supplied by the stenotic artery.
Pharmacologic Agents
Pharmacologic vasodilatation with adenosine or dipyridamole is indicated for
patients who are unable to exercise or are unable to increase their heart rate (rate-
limiting medications) and patients with a left bundle branch block (LBBB) or paced
ventricular rhythm.
Due to delayed septal activation in patients with LBBB, there is a corresponding
delay in relaxation of the septum resulting in less time for diastolic flow. At rest, this
34 Nuclear Medicine
effect is negligible but as the heart rate increases (and diastole shortens) flow to the
septum may be compromised in the absence of coronary stenosis. Vasodilator stress
is preferred in this circumstance because it is not associated with a significant in-
crease in heart rate.
Vasodilator stress with adenosine or dipyridamole is contraindicated in patients
with significant reactive airway disease. Methylxanthines such as caffeine and
theophylline competitively block adenosine receptors and must be avoided for 24
3 hours prior to the examination.
Exercise and dobutamine increase myocardial oxygen demand and in the presence
of hemodynamically significant coronary stenosis result in myocardial ischemia.
Vasodilator stress with dipyridamole or adenosine usually creates flow heterogeneity
by causing a greater increase in coronary blood flow in normal coronary arteries
compared with coronary arteries with significant stenosis. Although myocardial
ischemia occurs much less commonly with these agents, there is the potential for the
development of ischemia due to a decrease in distal perfusion pressure and/or the
development of a coronary steal. Side effects occurring with vasodilator stress can be
rapidly reversed by the intravenous administration of aminophylline (100-200 mg
over 2-5 min).
Adenosine
Adenosine activates specific receptors in vascular smooth muscle resulting in
smooth muscle relaxation and vasodilatation. This results in approximately a four-
fold increase in coronary flow. Adenosine is administered intravenously at 140 g/
kg per minute for three minutes followed by radiotracer injection into a different
vein and continuation of the adenosine infusion for an additional three minutes.
Dipyridamole
Dipyridamole blocks the cellular re-uptake of endogenously produced adenosine.
It is administered as an intravenous infusion of 0.56 mg/kg over 4 minutes. The
radiotracer is injected after an additional 4 minutes (i.e., 8 minutes after the start of
the dipyridamole infusion) when there is maximal increase in coronary flow. The
use of low-level supplemental exercise (e.g., Bruce stage I) following the infusion is
used in some centers. It serves to lessen symptoms as well as to improve image
quality by reducing infradiaphragmatic splanchnic activity.
Dobutamine
Dobutamine is a beta adrenergic agonist with both positive chronotropic and
inotropic effects resulting in coronary arteriolar dilatation secondary to an increase
in myocardial oxygen demand. Dobutamine is given as an intravenous infusion in
incremental doses starting at 5 g/kg/min and gradually increasing, at three-minute
intervals, to 40 g/kg/min. If there is a submaximal increase in heart rate, atropine
is often administered during maximal dobutamine infusion. Dobutamine is less
potent than either adenosine or dipyridamole for maximizing coronary blood flow
(two to threefold increase in coronary flow) and may be associated with a lower
sensitivity in the detection of CAD. It is primarily used for patients with reactive
airway disease in whom adenosine and dipyridamole are contraindicated.
Myocardial Perfusion Imaging 35
Figure 3. Flow-chart outlining the criteria for the selection of treadmill, dipyridamole/
adenosine or dobutamine stress.
201 99m
Tl-chloride Tc-perfusion agents
Physical properties
decay mode electron capture isomeric transition
physical half life 72 hours 6 hours
emission 63-80 keV 140 keV
3
Effective dose 23 mSv (2.3 rems) 13 mSv (1.3 rems) for
for 100 MBq 750 MBq x 2 (stress &
rest)
Blood flow
Extraction efficiency 85% 65%
Tracks high flow rates good moderate
201
Tl is the radioactive form of thallium used for scanning. It has a half-life of 72
hours and decays by electron capture. Characteristic x-rays (range of 63-80 keV)
produced during electron capture in addition to low abundance 135 and 167 keV
gamma rays are imaged. 201Tl has relatively poor imaging characteristics. Its low
photon energy makes it subject to attenuation by overlying soft tissues while its long
half life limits the activity that can be administered rendering the images relatively
count poor.
99m Tc-sestamibi
This tracer is the most commonly used 99mTc-perfusion agent. It is a lipophilic
monovalent cation that is passively taken up by myocytes along an electochemical
gradient (and therefore requires cell membrane integrity). The extraction efficiency
is approximately 65%. In the physiologic flow range, uptake is proportional to flow.
However the deposition of the tracer does not increase linearly with flow but rather
tends to level off at higher flow rates (Fig. 4). The distribution of 99mTc-sestamibi
remains relatively fixed for several hours. This allows imaging to be delayed up to
several hours after injection thereby facilitating the evaluation of patients presenting
with acute chest pain (who can be injected during pain and imaged several hours
later once stabilized).
Separate injections are required at stress and rest. These may be done on the
same day (e.g., rest injection of 300 MBq and several hours later a stress injection of
1000 MBq) or as a two day protocol (e.g., 750 MBq injections at rest and stress on
separate days). A third option is to administer 201Tl for the rest study and 99mTc-
sestamibi for the stress study.
38 Nuclear Medicine
Figure 4. Schematic representation of myocardial uptake versus flow for 201Tl and
99mTc-sestamibi. Uptake increases linearly with flow at low-flow rates for both tracers.
At higher flow rates, there is a roll-off which is less marked for 201Tl (curve A)
than for 99mTc-sestamibi. For two areas of the heart with flows of x and y, there is
a greater difference in uptake with 201Tl (a) than with 99mTc-sestamibi (b).
Hepatobiliary excretion of tracer may result in liver and/or gut activity obscur-
ing the inferior wall of the left ventricle. To minimize adjacent infra-diaphragmatic
activity imaging is delayed for at least 30 minutes after a stress injection and 60
minutes after a rest injection.
99m Tc-tetrofosmin
99m
Tc-tetrofosmin is another commonly used 99mTc-perfusion agent with
properties similar to 99mTc-sestamibi. While its extraction efficiency is somewhat
lower than that of 99mTc-sestamibi there is less hepatobiliary extraction potentially
decreasing interference with inferior wall assessment.
PET Tracers
While not available in many centers, PET radiopharmaceuticals can be used to
assess myocardial flow and viability. 18F-fluorodeoxyglucose (FDG) is a glucose
analogue. In the fasting state the myocardium normally uses fatty acids as its metabolic
substrate. In ischemic myocardium, glucose is used preferentially. Increased uptake
of FDG (in comparison to flow which is assessed with a separate radiopharmaceutical)
in an ischemic segment indicates sustained metabolic activity and implies myocardial
viability.
PET tracers to assess myocardial blood flow include 13N-ammonia and rubidium-
82 (82Rb). 13N-ammonia has a high extraction efficiency of 90%. The energy of the
positron and consequently its path range is low yielding improved resolution. Because
of its short half-life of ten minutes, an on-site cyclotron is needed to produce 13N.
82Rb has properties similar to potassium and thallium. Since it is produced from a
strontium-82 generator, an on-site cyclotron is not required. 82Rb has a short half
life of 1.3 minutes with an extraction efficiency lower than that of 13N-ammonia. Its
positron energy is greater than that of 13N-ammonia and hence resolution is poorer.
Myocardial Perfusion Imaging 39
Figure 5. Normal planar 201Tl scan with images in the anterior, LAO and LAT
projections. Panel A shows the nomenclature of the myocardial segments. There is
uniform uptake of tracer in all segments of the left ventricle on the immediate post-
stress images (panel B) and on the delayed images 4 hours post injection (panel C).
Typically, the counts in the heart decrease by approximately 50% over 4 hours. For
the purpose of comparison, however, the intensity of uptake within the heart is
normalized to the same brightness as the stress images and therefore the decrease
in counts is not apparent. (Ant-anterior; AL-anterolateral; Ap-apex; Inf-inferior; IA-
inferoapical; IS-inferoseptal; Post-posterior; PL-posterolateral; Sep-septum; RV-right
ventricle)
Figure 6. Schematic showing the three orthogonal planes used for SPECT myocardial
imaging and segmental nomenclature. The short-axis (SA) slices can be considered
as cucumber slices through the left ventricle. The horizontal long axis (HLA)
slices are similar in orientation to the echocardiographic four chamber view. The
vertical long axis (VLA) slices can be thought of as viewing the heart from a lateral
projection. (ant-anterior; ap-apex; inf-inferior; lat-lateral; sep-septal)
data spanning the R-R interval. This allows for assessment of wall motion and using
commercially available software the determination of the left venticular ejection
fraction (Fig. 8). Gated SPECT has become the standard technique for MPI.
Interpretation and Quantification
The interpretative approach to myocardial perfusion imaging is illustrated in
Figure 9. A defect at stress which improves or normalizes at rest is termed a reversible
defect and indicates ischemia. An infarct will result in diminished or absent uptake
of tracer at stress and at rest. This pattern is termed a fixed defect.
The distribution of radiotracer in the heart is assessed qualitatively by examining
the tomographic slices in orthogonal planes. It is also possible to generate a polar
plot that condenses the three-dimensional information into a two-dimensional image
(Fig. 10). Tracer distribution in the myocardium for a group of patients with a low
likelihood of CAD can be used to generate a normal database. In a polar plot, segments
of the myocardium below the normal range can be highlighted (Fig. 11).
A semi-quantitative method of representing the extent and severity of the perfusion
abnormalities at stress has become popular. The heart is divided into 17 or 20 segments
(Fig. 12). Each segment is graded on a five-point scale from normal (0) to absent
uptake (4). The sum of these values is the summed stress score (SSS) and is a useful
index of cardiac risk. In a similar manner a summed rest score (SRS) can be generated
as well as a summed difference score (= SSS-SRS).
Myocardial Perfusion Imaging 41
Figure 7. Normal MPI study in a 54 year old woman referred because of a positive
GXT. There is mildly diminished activity anteriorily both at stress and rest (arrow)
attributable to breast attenuation. No other fixed or reversible abnormalities are
identified. The top row are short axis (SA) images of the left ventricle at stress and
the second row are corresponding SA slices at rest going from apex (on the left)
towards the base of the heart (on the right). The next two rows are vertical long
axis images (VLA) at stress and rest going from septum to lateral wall. The bottom
two rows are horizontal long axis (HLA) at stress and rest going from inferior to
superior slices. For subsequent MPI studies the same format has been used.
Two ancillary findings, increased lung thallium activity and transient ischemic
dilatation are useful in defining patients at high risk for subsequent events. Increased
lung activity is related to increased LV filling pressure and accumulation of 201Tl in
the pulmonary interstitial space is a marker for patients with multivessel coronary
disease and/or LV dysfunction. It is assessed quantitatively in the anterior view per-
formed immediately after the injection of 201Tl during stress, by comparing activity
in lung to that in the hottest area of the myocardium. The lung/heart ratio is less
useful with 99mTc-sestamibi and tetrofosmin, partly because of the 30-60 minute
delay before imaging begins. Dilatation of the LV cavity at stress compared to rest is
variously termed transient ischemic dilatation or stress-induced dilatation. It is caused
by subendocardial ischemia at stress (causing an apparent increase in LV size) or by
true dilatation of the LV cavity post-stress.
42 Nuclear Medicine
Figure 8. Gated SPECT. (A) 3 short axis slices are shown at ED on the left and ES on
the right. Note that the ventricle is smaller and brighter at ES. The increase in
brightness is related to systolic thickening of the ventricular wall. On the right
hand side of the panel, single VLA and HLA mid-ventricular slice are shown at ED
and ES. (B) A program identifies the endocardial margins of the left ventricle for all
frames and generates a volume curve from which the ejection fracture (EF) is derived.
Myocardial Perfusion Imaging 43
Figure 9. (A) Fixed and reversible defects. A defect (light area) present at stress and
rest is termed a fixed defect and typically represents an area of infarction. A defect
present at stress but not at rest is termed a reversible defect and implies ischemia.
(B) Schematic showing induction of a perfusion defect at stress. One segment of
the ventricle is supplied by a normal artery and a second segment by a stenoic
artery. At rest, there is equal flow to both myocardial segments and equal uptake of
tracer. At stress, flow to the myocardium supplied by the patent artery increases
markedly. Flow to the myocardium supplied by the stenoic artery increases
marginally. Less tracer will be taken up in the myocardium supplied by the stenotic
vessel than in normally supplied myocardium, i.e., a perfusion defect is induced at stress.
44 Nuclear Medicine
Figure 10. Schematic showng derivation of the polar plot. The heart is divided into
one pixel slices (A). A single slice is shown on the right side of panel A. The septum
and lateral wall show normal uptake while the inferior wall shows a mild defect
and the anterior wall a severe defect. Information from this slice (B) and all other
slices is inserted into a polar plot (C). In this example, the septum and lateral wall
are normal. The entire inferior wall shows mildly diminished activity while the
entire anterior wall and apex show markedly dimished uptake.
Myocardial Perfusion Imaging 45
Figure 11. Polar plot of study shown in Figure 17. (A) Stress polar plot showing
diminished activity (darker area) in the inferior wall. (B) Rest polar plot showing
improved uptake in the inferior wall. (C) Using the stress polar plot, areas considered
to be abnormal (based on normal population database) are highlighted in black.
(D) Using this blackout image, areas considered to improve significantly between
stress and rest (i.e., the reversible inferior wall defect) are highlighted in white.
Figure 12. Twenty segment model used for calculation of the stress severity score
(SSS). The left ventricular apex is divided into two segments (anteroapical and
inferoapical). The remainder of the ventricle is divided into three cylindrical slices
representing the distal, mid and basal thirds of the ventricle. Each of these thirds is
divided into six segments (Ant-anterior; AL-anterolateral; PL-posterolateral; Inf-
inferior; IS-inferoseptal; AS-anteroseptal) for a total of 20 segments. To calculate
the SSS, each segment is assigned a score from 0 (normal) to 4 (absent) uptake. The
sum of the scores for all 20 segments is the SSS.
disease in the population under study (for a simplified approach to Bayes theorem
see FAQs). Diagnostic testing is most effective in patients with an intermediate
prevalence of disease estimated from the patients age, gender and chest pain history,
together with consideration of coronary risk factors (Fig. 13).
Sensitivity
The overall sensitivity of SPECT imaging using either 201Tl or 99mTc-sestamibi is
85-90%. SPECT detects stenosis in individual vessels with a sensitivity for the left
anterior descending artery of 80%, right coronary artery 80% and circumflex 70%.
In those patients in whom SPECT cannot be performed (unable to lift left arm,
unable to lie still, body weight considerations), quantitative planar studies yield a
similar sensitivity (on a per patient basis) although the identification of individual
vessel stenoses is reduced.
Sensitivity varies depending upon:
Myocardial Perfusion Imaging 47
Table 3. Indications for MPI for detection of coronary artery disease (CAD)
Intermediate Pretest Probability of CAD
Abnormal Resting Electrocardiogram
Non-specific ST-T Abnormalities
Left Ventricular Hypertrophy
Conduction Disturbance
Ventricular Pacing
Pre-excitation (WPW syndrome). 3
Non-diagnostic Treadmill Stress Test
Inability to Reach * 85% Maximal Predicted Heart Rate
Digoxin
Figure 13. 85 year old male with atypical chest pain. The MPI study was performed
for diagnostic purposes. There is extensive severe reduction of activity in the septum
(long arrow), anterior wall (arrowhead) and apex (short arrow) at stress with
normalization at rest. A subsequent angiogram showed an occluded LAD with filling
of the LAD via collaterals from the RCA. Note the dilation of the LV cavity at stress
(transient ischemic dilation) which is evident in all three planes.
Figure 14. 70 year old male with stable angina. The MPI study was performed for
risk assessment. There is reduced activity in the anterior wall (narrow arrowhead),
apex (short arrow) and septum (long arrow) at rest with improvement at rest. There
is reduced activity inferiorly at stress with improvement of the distal inferior wall at
rest (broad arrowhead). There is evidence of both LAD and RCA disease.
Figure 15. 71 year old male with post MI angina. An angiogram showed moderately
severe stenosis of the proximal RCA and severe stenosis of the circumflex artery.
The MPI study was performed to identify the culprit artery prior to
revascularization. At stress, severe reduction of activity is seen in the lateral wall
on the short axis and HLA images (arrows) with improvement at rest. Based on the
finding of a reversible lateral wall abnormality, the patient underwent a PTCA of
the circumflex artery with excellent results.
motion the projection images are examined. This is typically done by reviewing the
images in cine mode looking for evidence of vertical or horizontal movement of the
heart. Another cause of an apparent myocardial defect is attenuation from overlying
soft tissue structures. In women breast tissue frequently overlaps the anterior
myocardial wall on many of the SPECT projections. In these projections fewer
anterior wall counts will be detected and consequently, upon reconstruction of the
data, the anterior wall will show apparently diminished uptake (Fig. 7). Attenuation
of the inferior wall is frequently seen in men with a protuberant abdomen and an
elevated left hemidiaphragm.
Clinical Role in Prognosis and Risk Stratification
MPI provides important prognostic information in a wide spectrum of patients
with CAD. The ability to identify patients at high and low risk has been validated
for both planar and tomographic techniques using 201Tl or 99mTc-sestamibi with
exercise or pharmacologic stress.
Myocardial Perfusion Imaging 51
(dipyridamole or adenosine) stratifies patients into low, intermediate and high risk.
The extent and severity of the perfusion defect and the degree of reversibility are
predictors for in-hospital and late cardiac events whether or not patients receive
thrombolytic therapy.
Early risk stratification is important, since most early (< 1 year) post-MI cardiac
events (e.g., unstable angina, recurrent infarction, cardiac death) occur within 4-6
52 Nuclear Medicine
Figure 16. Flow chart outlining a cost-effective strategy for the evaluation of coronary
artery disease (MPI-myocardial perfusion imaging; GXT-graded exercise stress test).
weeks of the acute event. The ability to perform vasodilator stress imaging safely two
to four days following acute infarction with greater separation of low and high risk
patients compared with submaximal exercise imaging makes this an attractive strategy.
Patients identified as low risk can be considered for early discharge whereas those at
high risk can be referred for early angiography and possible revascularization.
Unstable Angina
Patients with an acute non-ST segment elevation coronary syndrome (unstable
angina or non-Q wave myocardial infarction) who have recurrent angina, ST segment
depression or elevated troponin should be considered for early coronary angiography.
Patients with unstable angina without ST segment depression or biochemical markers
of myocardial injury who respond to initial medical treatment can be effectively risk
stratified with perfusion imaging. In the latter group, there is no reduction in hard
cardiac events with early coronary angiography and revascularization compared with
Myocardial Perfusion Imaging 53
Figure 17. 50 year old male with previous inferior MI. A GXT demonstrated mild
ST segment depression at a high workload without angina. The MPI study was
done for risk stratification. A large area of severely decreased uptake is seen in the
inferior wall at stress (arrows). This area normalizes at rest. Subsequently,
angiography demonstrated an occlusion of the RCA which was dilated and stented.
initial strategy of MPI in both intermediate and high risk patients reserving angiog-
raphy for those with extensive inducible ischemia.
The negative predictive value of a normal scan is approximately 99% while ap-
proximately 30% of intermediate risk vascular surgical patients with scintigraphic
evidence of inducible ischemia will have a perioperative cardiac ischemic complication.
Perioperative beta-blockers reduce cardiac death and non-fatal infarction in high-
risk patients. Alternatively patients considered to have a high probability of multivessel
3 CAD and left ventricular dysfunction by non-invasive assessment can selectively be
referred for coronary angiography and possible myocardial revascularization (CABG
or PTCA) with a view to improving their long-term survival.
Clinical Role in Defining Myocardial viability
Impaired left ventricular function is an important determinant not only of
symptoms but also of survival. Regional LV dysfunction may result from (a) a
transmural myocardial infarction (b) a non-transmural (subendocardial) myocardial
infarction (c) a transient period of myocardial ischemia (stunning) causing a wall
motion abnormality for a variable length of time after the transient ischemic insult
or (d) a prolonged period of resting myocardial hypoperfusion (hibernation). It
has been suggested that hibernating myocardium may actually represent the effects
of repetitive stunning i.e., patients with hibernating myocardium have reduced
coronary flow reserve without reduced resting coronary blood flow.
If the myocardium in the abnormally contracting region is viable i.e., has sustained
metabolic activity, left ventricular dysfunction in patients with CAD may be
potentially reversible with restoration of coronary blood flow for those patients with
stunned or hibernating myocardium. On the other hand regional left ventricular
dysfunction due to myocardial infarction (scar) is irreversible after revascularization.
Recovery of regional left ventricular dysfunction may occur (a) after thrombolytic
therapy or PTCA for acute myocardial infarction (b) after exercise induced ischemia
(c) in patients who have experienced unstable angina and (d) in the setting of chronic
silent ischemia (when overt or compensated cardiac failure may be the dominant
clinical finding) after CABG or PTCA.
There are a number of non-invasive techniques available for the assessment of
myocardial viability (Table 4). Metabolic imaging with 18F-FDG is often considered
the gold standard for assessing myocardial viability. 18F-FDG uptake in a region of
decreased perfusion (mismatch pattern) supports a determination of myocardial
viability. Presently, few centers have the capacity to perform PET studies.
201Tl and 99mTc-sestamibi myocardial perfusion imaging assess perfusion and cell
for viability has a positive predictive accuracy of approximately 70% and a negative
predictive accuracy of 90% for improved regional left ventricular function after
revascularization. The predictive accuracy of myocardial perfusion imaging for
functional recovery can be improved by considering the level of regional tracer activity
as a continuum and documenting whether there is inducible ischemia.
Myocardial Perfusion Imaging 55
Figure 18. 53 year old male with a previous history of an inferior MI being considered
for revascularization. Both a ventriculogram and an equilibrium radionuclide
angiocardiogram showed widespread wall motion abnormalities and global decrease
in systolic function (ejection fraction = 28%). The study was performed as a
dipyridamole 201Tl stress study with additional images at four hours and following
reinjection. Only the stress and reinjection images are shown. There is severe (<50%
of peak) fixed reduction of activity in the inferior wall (short arrow), a reversible
septal abnormality (long arrow) and a partially reversible anterior wall abnormality
(arrowhead). In summary, the examination demonstrates extensive ischemia and,
with the exception of the inferior wall, viability of all segments.
probability (Fig. 19). As can be seen the greatest changes in probability occur when
applying this test to patients who have an intermediate pre-test probability of disease.
Sequential Bayesian analysis refers to the situation where the post-test probabil-
ity after a diagnostic test becomes the pre-test probability for the next test. Suppose
that on the basis of risk factors a patient is assigned a pre-test probability of 20%. He
then undergoes a GXT after which the (post-test) probability is raised to 60%. This
60% probability now becomes the pre-test probability for any subsequent examina-
tion such as a myocardial perfusion study.
Why is it necessary to have polar maps based on a normal
database? Shouldnt perfusion to all portions of the heart be
Myocardial Perfusion Imaging 57
CAD
yes no
3
MPI +ve 54 4 58
-ve 6 36 42
60 40
equal?
With a perfect imaging system the answer would be yes. However, there are
certain factors that prevent this. Firstly, activity in the heart will be attenuated to a
greater or lesser extent by soft tissues. For instance in women, photons arising from
the anterior myocardial wall will in many SPECT projections have to traverse the
breast. A certain number of photons will be absorbed or scattered in the breast and
will not be detected by the gamma camera. Because fewer anterior wall counts will
be detected, the reconstructed images will contain fewer counts in the anterior wall
(i.e., there will be an apparent anterior wall defect). In men, the inferior wall often
demonstrates decreased activity because of attenuation from infra-diaphragmatic
structures. Secondly scattered photons emanating from outside the heart will be
detected as originating from within the heart causing inhomogeneity of uptake.
Lastly, resolution in deeper structures is poorer than resolution of more superficial
structures contributing to non-uniformity. There are various approaches to correct
for this non-uniformity.
Transmission imaging with an external radioactive source can be used to correct
for the effects of variable attenuation (Fig. 20). A gadolinium-153 sealed source(s) is
mounted on the gamma camera. The camera is rotated about the patient and a
transmission map (which can be thought of as a low resolution CT scan) is
reconstructed from the projection data. With the transmission data in hand the
emission (SPECT) data can be corrected for variable attenuation. This technique is
not as yet being used on a widespread clinical basis.
Additional Reading
1. Beller GA, Zaret BL. Contributions of nuclear cardiology to diagnosis and prognosis
of patients with coronary artery disease. Circulation 2000; 101:1465-1478.
Excellent review of myocardial perfusion imaging.
2. Bergmann SR. Cardiac positron emission tomography. Semin Nucl Med 1998;
28(4):320-340.
A review of cardiac PET applications.
3. Berman DS, Germano G, Shaw LJ. The role of nuclear cardiology in clinical decision
making. Semin Nucl Med 1999; 29(4):280-297.
An excellent article focusing on risk stratification.
4. Botvinick EH, ed. Unit 2: Pharmacologic stress and associated topics. In : Nuclear
58 Nuclear Medicine
Figure 19. The relationship of pre- and post-test probability is shown for a test with
90% sensitivity and specificity. For example, if the pretest probability of disease is
0.6 and the test is positive, the post-test probability of disease will be 93% (top
curve). If the test is negative, the likelihood of disease is 14% (bottom curve).
Figure 20. (A) Vertical long axis image shows decreased activity in the anterior wall
in this 74 year old female. (B) With attenuation correction, the anterior wall
normalizes, suggesting that the reduction is due to breast attenuation.
Equilibrium Radionuclide
Angiocardiography
I. David Greenberg and Robert Corne
Introduction
A quantitative evaluation of ventricular function gives an objective assessment of
ventricular performance in asymptomatic patients and provides prognostic
information in patients with ischemic and valvular heart disease. The technique of
radionuclide angiography provides an accurate and reproducible method of non-
invasively assessing global left ventricular systolic function in patients receiving
cardiotoxic chemotherapy. The extent and severity of regional left ventricular wall
motion abnormalities including ventricular aneurysm can be defined. Right
ventricular function can also be evaluated.
Technical Considerations
The nuclear medicine technique most commonly used to assess ventricular
function is termed equilibrium radionuclide angiocardiography (ERNA). The first
ERNA program was developed by Medical Data Systems which coined the term
MUGA (multiple gated acquisition). This proprietary term is still frequently used as
is the generic description gated blood pool imaging. The American Society of
Nuclear Cardiology has suggested that ERNA be adopted as standard terminology.
In this technique a tracer is administered that remains in the intravascular
compartment. After allowing several minutes for equilibration (hence the term
equilibrium), the heartthe largest blood containing structure in the bodycan
be imaged. While planar images could identify the cardiac chambers, they would
not provide information on cardiac function. Gating the studies allows for the
generation of a series of images representing the heart through the entire cardiac
cycle. This is achieved by attaching electrodes to the patient and using the R wave to
synchronize image acquisition with the cardiac cycle (Fig. 1). Effectively one acquires
information over several hundred heartbeats to produce a composite cardiac cycle
consisting of 16 to 24 images. These images can then be viewed in an endless loop
format allowing wall motion to be assessed. The left ventricular ejection fraction
will also be derived from these images.
Typically image sets are obtained in three projections: anterior, left anterior ob-
lique and steep left anterior oblique (or left lateral). In this way the various walls of
the left ventricle can be seen tangentially and regional wall motion evaluated (Fig.
2). The LAO view is not taken at a specific angle but rather the technologist will
adjust the projection to maximize separation between the left and right ventricles.
The camera is angled caudally to separate left ventricle and left atrium. This modi-
Figure 2. Normal ERNA. Typically, three projections are obtained: anterior (A),
best-septal (B) and steep LAO or lateral (C). The left ventricle is seen in all three
projections (arrow). When describing wall motion, the same segmental
nomenclature is used as that for planar 201Tl or 99mTc-sestamibi studies (AL-
anterolateral; Ap-Apex; IS-inferoseptal; PL-posterolateral; IA-inferoapical; Sep-
Septum; Ant-Anterior; Inf-Inferior; Post-Posterior). The spleen (arrowhead) is often seen.
fied LAO projection, termed the best septal view, separates the left ventricle from
other structures and is used for determination of the ejection fraction. In our center,
the best-septal view is acquired first. The anterior view is obtained by rotating the
camera head 40 counter-clockwise from the best septal view while the steep LAO
Equilibrium Radionuclide Angiocardiography 63
projection is obtained by rotating the camera 30 clockwise from the best septal
view.
Intravascular Agents
A tracer is needed that remains in the blood pool for sufficient time to allow for
acquisition of the gated blood pool images. There are two approaches to achieve
this: labeling of the patients own red blood cells (RBCs) with technetium-99m or
administering a technetium-99m labeled macromolecule such as 99mTc-HSA (human
serum albumin). The latter technique is not commonly used since the residence
time of 99mTc-HSA in the blood pool is less than that for 99mTc-RBCs. While we are
not aware of any case in which hepatitis or HIV has been transmitted through the
use of 99mTc-HSA, the use of autologous RBC labeling obviates this concern.
There are three methods available for labeling red blood cells:
(a) In Vivo: In this technique 10-20 g/kg of stannous (Sn2+) ions is injected.
After 20-30 minutes, 99mTc-pertechnetate is administered. This sequence produces
hemoglobin labeling with technetium-99m (Fig. 3). Labeling efficiency (ie. RBC/
Whole blood activity x 100%) with this technique ranges from 80-90%.
In Vitro: In this technique, several ml of blood are withdrawn from the patient
labeled in a test tube (using Sn2+ and 99mTc-pertechnetate) and then reinjected.
Labeling is reliable with typical labeling efficiencies in excess of 95%.
Modified in vivo (or in vivitro): In this technique, Sn2+ is injected in the same
manner as for the in vivo technique. After 30 minutes, several ml of blood are
withdrawn into a syringe containing 99mTc-pertechnetate. Several minutes are allowed
for RBC labeling within the syringe after which the labeled RBCs are re-injected.
Labeling efficiency is typically in the 90% range.
The in vitro technique provides the most reliable and highest quality labeling.
However, the commercial kit used for in vitro labeling (Ultratag) is not inexpensive.
Moreover RBC labeling is performed on the bench removed from the patient. If
multiple patients are being imaged, it is crucial that procedures are in place to pre-
64 Nuclear Medicine
vent the inadvertent mixing of blood samples. In our laboratory, we use the in vitro
technique in patients in whom we have experienced (or anticipate) poor RBC label-
ing.
Ejection Fraction
The ejection fraction (EF) as the name implies is the percentage of blood in the
left (or right) ventricle that is ejected during contraction of the heart. In
echocardiography (and contrast ventriculography) the technologist traces the outline
of the left ventricle at end-diastole and end-systole. On the assumption that the left
ventricle has an ellipsoidal shape, end-diastolic and end-systolic volumes are calculated
4 from which the ejection fraction is determined. If all ventricular walls contract equally
this geometric assumption is accurate. It is less reliable if regional wall motion
abnormalities are present. In ERNA, the left ventricular ejection fraction is determined
without assumptions as to the shape of the heart; rather it is based on the principle
that counts in the left ventricle are proportional to volume (Fig. 4). Using the best-
septal view, outlines of the left ventricle at ED and ES are manually drawn or generated
by an edge detection program (Fig. 5). Counts in the left ventricle can assumed to
be proportional to volume. To correct for background activity in front and behind
the heart and scatter from adjacent structures, a background region of interest is
drawn adjacent to the heart and the EF calculated. Note that no geometric
assumptions are made as to the shape of the heart. Normally the left ventricular EF
is *50% while the right ventricular EF is >40-45%.
Parametric Images
In addition to analyzing the images in cine-mode for assessment of regional wall
motion, it is possible to generate valuable information from parametric images. (A
parametric image can be thought of as an image that condenses information from a
series of images). Phase and amplitude images are two commonly generated parametric
images. The phase image provides information as to the timing of contraction in
various portions of the heart and the amplitude image on the degree of contraction
(Fig. 6).
Clinical Applications
Various modalities are available to assess left ventricular function including ERNA,
echocardiography, gated myocardial perfusion imaging (see Chapter 3) and contrast
venticulography. The relative strengths and weaknesses of these procedures are listed
in Table 1. The technique employed to evaluate ventricular function will depend on
local expertise, test availability and the clinical scenario. For instance, in a patient
with valvular disease, echocardiography can assess valve structure and function as
well as ventricular function. On the other hand ERNA has proven to be an accurate
and reproducible method of serially following the EF of patients being treated patients
with cardiotoxic drugs such as doxorubicin.
Congestive Heart Failure
Congestive heart failure may be due to left ventricular systolic and/or diastolic
dysfunction (Fig. 7). Systolic dysfunction is due to impaired contractility or afterload
mismatch (increased wall tension due to an increase in intraventricular pressure or
Equilibrium Radionuclide Angiocardiography 65
Figure 4. Count-based ejection fraction. The ejection fraction (EF) is the percentage
of the end-diastolic volume (EDV) that is ejected during systolic contraction. In this
example, EDV is 100 ml, end-systolic volume (ESV) is 40 ml and the EF is 60% (A).
Assuming that the counts in the LV are proportional to volume the end-diastolic
counts (EDC) and end-systolic counts (SC) can be used to calculate the EF (B).
66 Nuclear Medicine
Figure 5. Left ventricular time-activity curve. Using the best septal view, regions of
interest (ROIs) encompassing the left ventricle are either drawn manually or
generated using an edge-detection program. Shown are computer-derived ROIs
around the LV at end-diastole (A) and end-systole (B). ROIs are generated for all
frames of the image set. After correcting for the contribution of background, a left
ventricular time-activity curve is produced (C). Examining the curve, one can see
that in the first frame (end-diastole) there are 20,000 counts in the left ventricle.
The seventh frame has the fewest counts in the left ventricle (8,000) and is the end-
systolic frame. The ejection frame is therefore 60%.
Figure 6. Phase and amplitude images. The series of ERNA images can be analyzed
in such a way as to generate images depicting the degree of pixel change between
end-diastole (ED) and end-systole (ES) as well as the timing of pixel change. The
top row depicts a normal patient with the amplitude image showing a horseshoe
area of activity (long arrow) in the left ventricle (since the peripherally located LV
pixels show the greatest change between the ED and ES). Similarly, a crescentic
area is seen for the RV (short arrow). The phase image shows a similar gray color
for both the left and right ventricles indicating that they contract simultaneously. In
the lower row there is a low amplitude area (i.e., little change between ED and ES)
inferoapically which on the phase image is seen to be out of pahse with the
remaining ventricular mass implying that it is dyskinetic (arrowhead).
grafting in patients with multivessel CAD and mild to moderate left ventricular
dysfunction (EF 35-49%).
Table 2. Factors associated with failure of the ejection fraction to increase with
exercise
Cause Comment
Gender Women increase their stroke volume with exercise mainly
by an increase in end-diastolic volume (Frank-Starling
mechanism) rather than by an increase in contractility
Hypertension Patients with hypertension may have a fall in EF as a
result of an exaggerated blood pressure response to
exercise
4 Anti-anginal drugs The negative inotropic action of beta-blockers and
calcium channel blockers may blunt the normal increase
in the ejection fraction with exercise
Age Elderly patients may have an abnormal exercise response
in the absence of coronary artery disease
Non-coronary disease Myocardial contractility may be impaired in patients with
a cardiomyopathy or valvular heart disease
Left bundle branch Patients with a LBBB may have an abnormal response to
block exercise in the absence of coronary artery disease
Doxorubicin Cardiotoxicity
Doxorubicin and other anthracyclines such as daunorubicin, epirubicin and
mitoxantrone are commonly employed chemotherapeutic drugs which may be
associated with significant cardiotoxicity. ERNA provides an accurate, reproducible
and non-invasive method of serially monitoring left ventricular function. It has been
recommended that ERNA be obtained prior to chemotherapy and if left ventricular
function is normal (EF *50%) subsequent studies be obtained at a total doxorubicin
dose of 250-300 mg/m2, at 400-450 mg/m2, and before each subsequent dose.
Cardiotoxicity, defined by a decrease in the EF of *10% and an EF <50% is predictive
of subsequent clinical cardiac failure if treatment is continued. If there is left
ventricular dysfunction (EF <50%) on the pre-chemotherapy ERNA, additional
studies should be obtained prior to each dose, with treatment discontinued if the EF
falls by 10% or the EF decreases to 30% (Fig. 8).
Valvular Heart Disease
Echocardiography is preferred for the non-invasive assessment of patients with
valvular heart disease since it provides information on valve structure in addition to
assessing ventricular function. ERNA provides an alternative method of assessing
ventricular function in patients with technically suboptimal echocardiographic studies
or in patients with borderline left ventricular dysfunction for whom an independent
assessment of left ventricular function may be helpful. Surgery is recommended
when patients with valvular regurgitation develop symptoms or resting left ventricular
dysfunction (EF <50% in aortic regurgitation and <60% in mitral regurgitation).
Although the exercise ejection fraction response in asymptomatic patients with severe
aortic regurgitation is an independent predictor of death, left ventricular dysfunction
or the development of symptoms, there are no data to support aortic valve surgery
based only on a fall in ejection fraction with exercise.
Equilibrium Radionuclide Angiocardiography 71
Figure 9. First pass study of right ventricle. Immediately after tracer injection into
an antecubital vein, a gated study was acquired in the brief period that activity
traversed the right heart but before arrival of tracer in the left heart. The study is
typically acquired in an anterior or RAO projection to optimally separate right
atrium (thin arrow) and right ventricle (arrowhead). With this technique, the right
ventricular ejection fraction can be determined in the same manner as the left
ventricular ejection fraction in an ERNA study.
Right Ventricle
Evaluation of right ventricular function is important in patients who are
hemodynamically unstable following inferior infarction, patients with pulmonary
hypertension and in patients suspected of having arrhythmogenic right ventricular
dysplasia (fibro-fatty infiltration resulting in a dilated hypokinetic right ventricle).
Right ventricular size and wall motion can be assessed on an ERNA study. Overlap
of other cardiac structures (especially the right atrium) on the right ventricle limits
the accuracy of a right ventricular EF determination. For this reason, a first pass
study (see FAQs) is commonly used to assess the right ventricular EF (Fig. 9).
72 Nuclear Medicine
Figure 10. List mode acquisition. The rhythm shown in (A) is sinus (beats a, b, c
and e) with a single ventricular premature beat (d). The volume curve (B)
demonstrates the effect of the premature beat on the ejection frame (EF). Because
of a longer diastole and greater ventricular filling following a ventricular premature
beat (d), the EF of the next sinus beat (e) is increased. The premature beat itself has
a decreased EF while the interrupted sinus beat (c) has a normal EF. With a list
mode acquisition, an R-R histogram is generated (C) and only those beats with a
normal R-R interval (between the dotted lines) are used to produce the series of
ERNA images. As can be appreciated, this is not a perfect solution since sinus beat
e (with an elevated EF) will be included while the interrupted sinus beat c (with
normal EF) will be excluded from the acquisition.
Figure 11. Normal SPECT ERNA in left panel and abnormal in right panel. The end-
diastolic contours are shown in a cage display while the end-systolic contours are
represented as solids. The number of voxels at end-diastole and end-systole are
determined for both the right and left ventricles and from these the right ventricular
ejection frame (RVEF) and left ventricular ejection frame (LVEF) are calculated.
Note that this differs from the count-based approach used for planar ERNA studies.
(Cases provided by ADAC Laboratories)
analyzed, one is not limited to tracers that remain in the intravascular blood pool
(beyond the first pass through the heart). A variety of techniques can be used. These
will not be discussed in detail. However, some factors to be considered are:
In a first pass study, separation of the right ventricle and left ventricle is achieved
temporally (Fig. 9). While activity is in the right ventricle, right ventricular function
can be assessed; after the activity leaves the right ventricle and reaches the left ventricle,
left ventricular function is assessed.
An anterior or RAO projection is usually chosen to allow separation of right
atrium from right ventricle and left atrium from left ventricle,
The study can be gated in the same way as an ERNA. Non-gated techniques
are also available.
After injection, the bolus of activity spreads. This can be tolerated for assessment
of right ventricular function but is a more severe problem with the assessment of left
ventricular function. To maintain a tight bolus, a jugular injection is often required.
Does irregularity of the cardiac rhythm cause difficulties in
interpreting an ERNA?
The ERNA technique generates a composite cardiac cycle based on several hun-
dred heart beats. The data is then used to generate a time activity curve from which
the ejection fraction is determined. Conceptually, one can consider several hundred
individual time activity curves superimposed to form the composite time-activity
74 Nuclear Medicine
curve. Referring to Figure 10, it can be appreciated that if the data from ventricular
premature beats is included, the evaluation of ventricular function during sinus
rhythm will be distorted. Cardiac programs allow us to deal with this problem in
two ways:
Many programs calculate the average R to R interval and reject data from beats
with an R-R interval which are beyond a predetermined acceptance limit (e.g., beats
with an R-R interval less than 80% or greater than 120% of the mean R-R interval).
Another approach is to collect the data serially in list mode. This data includes
scintigraphic information, timing pulses and R wave pulse information. An R-R
histogram is generated and the operator selects the appropriate R-R interval for
4 reformatting the representative cardiac cycle (Fig. 10).
Gated SPECT studies are done for myocardial perfusion
imaging. Can gated SPECT ERNA studies be performed as well?
Yes. Gated SPECT myocardial perfusion studies are now routine. The same tech-
nique can be applied to ERNA (Fig. 11). Indeed, any camera with the capability of
acquiring gated SPECT myocardial perfusion images is also capable of acquiring
gated SPECT ERNA. Potential advantages of this technique include: 1) absolute
volumes can be calculated in addition to the EF, 2) no corrections is needed for
background activity, 3) assessment of wall motion is not limited to those segments
seen tangentially on the 3 routine views, 4) three-dimensional phase images can be
generated. Once the software to generate volumes, ejection fractions and 3-D phase/
amplitude images is readily available, this technique will probably supplant planar
ERNA studies.
Additional Reading
1. DePuey EG, Garcia EV. Updated imaging guidelines for nuclear cardiology
procedures, Part 1. J Nucl Cardiol 2001; 8(1):G1-G58.
An expert consensus on how to perform nuclear cardiology studies. Pages G17-29 deal
with first pass and ERNA studies.
2. Jain D. Cardiotoxicity of doxorubicin and other anthracycline derivatives. J Nucl
Cardiol 2000; 7(1):53-62.
Nicely written article dealing with mechanisms, diagnosis and prevention of
cardiotoxicity.
3. Lee K, Pryor DB, Pieper KS et al. Prognostic value of radionuclide angiography in
medically treated patients with coronary artery disease: A comparison with clinical
and catheterization variables. Circulation 1990; 82:1705-1717.
A valuable report on the extensive experience with exercise ERNA at Duke University
in risk stratifying patients with chronic coronary artery disease.
4. Ritchie JL et al. ACC/AHA guidelines for clinical use of cardiac radionuclide
imaging. J Am Coll Cardiol 1995; 25:521-547.
Guidelines on the use of nuclear cardiology procedures published by the American College
of Cardiology and the American Heart Association
5. Ryan TJ et al. ACC/AHA guidelines for the management of patients with acute
myocardial infarction. Circulation 1999; 100:1016-1030.
Guidelines published by the American College of Cardiology and the American Heart
Association on the investigation and treatment of patients with an acute MI.
CHAPTER 1
CHAPTER 5
Thromboembolic Disease
Daniel F. Worsley and Philip S. Wells
Introduction
Pulmonary embolism (PE) is a relatively common and potentially fatal disorder
for which treatment is highly effective and improves patient survival. The diagnosis
of acute PE requires an interdisciplinary team approach and may be difficult because
of nonspecific clinical, laboratory and radiographic findings. The incidence of venous
thromboembolism is approximately 1 in 1,000 per year. Approximately 10% of
patients with PE die within one hour of the event. For those patients who survive
beyond the first hour following PE, treatment with heparin or thrombolytic agents
are both effective therapies. The mortality in patients with PE who are untreated has
been reported to be as high as 30%. In contrast, the correct diagnosis and appropriate
therapy significantly lowers mortality to between 2.5% and 8%. Although
anticoagulant therapy is effective in treating PE and reducing mortality, it is not
without some risk. The prevalence of major hemorrhagic complications has been
reported to be as high as 10-15% among patients receiving anticoagulant therapy.
Therefore, the accurate and prompt diagnosis of PE is not only essential to prevent
excessive mortality but also to avoid complications related to unnecessary
anticoagulant therapy.
Technical Considerations in Lung Scanning
Perfusion Lung Scanning
The agent of choice for perfusion imaging is 99mTc-labeled macroaggregated al-
bumin (99mTc-MAA). Technetium-99m MAA particles range in size from 10-150
m with over 90% of injected particles measuring between 10-90 m. The injection
of labeled particles should be performed with the patient in the supine position to
limit the effect of gravity on regional pulmonary arterial blood flow. Following the
intravenous administration of 99mTc-MAA, particles are mixed within the heart and
then lodge within precapillary arterioles in the lungs. The usual administered activity
is between 74-148 MBq (2-4 mCi) bound to 0.5-2.0 mg of human serum albumin.
The distribution of particles within the lungs is proportional to regional pulmonary
blood flow at the time of injection. Approximately 200,000-500,000 particles are
injected during a routine clinical perfusion lung scan. The blockage of pulmonary
precapillary arterioles by 99mTc-MAA is transient, and the biological half-life within
the lung ranges between 2-6 hours. In pediatric patients, patients with right to left
shunts, pulmonary hypertension or those who have undergone pneumonectomy or
single lung transplantation, the number of particles injected should be reduced. A
minimum of 60,000 particles is required to obtain an even distribution of activity
within the pulmonary arterial circulation and avoid potential false positive interpre-
tations.
When performing perfusion scintigraphy, at least six views of the lungs should
be obtained. These include anterior, posterior, right and left anterior oblique and
posterior oblique views (Fig. 1). Additionally, right and left lateral views may be
helpful in selected cases. Animal studies have demonstrated that perfusion imaging
will detect greater than 95% of emboli which completely occlude pulmonary arterial
vessels greater than 2 mm in diameter. In spite of this sensitivity, the perfusion scan
may underestimate some perfusion abnormalities. For example, a perfusion defect
limited to the medial basal segment of the right lower lobe is completely surrounded
by normal lung; consequently a perfusion defect in this segment will not be detected
on planar perfusion imaging.
Ventilation Lung Scanning
Perfusion scintigraphy is sensitive but not specific for diagnosing pulmonary
diseases. Virtually all parenchymal lung diseases (including tumors, infections, chronic
obstructive pulmonary disease [COPD] or asthma) can cause decreased pulmonary
arterial blood flow within the affected lung zone. Ventilation imaging was combined
with perfusion scintigraphy to improve the diagnostic specificity for PE based upon
the principle that PE characteristically cause abnormal perfusion with preserved
ventilation (mismatched defects) (Fig. 2). In contrast, parenchymal lung disease
most often causes both ventilation and perfusion abnormalities in the same lung
region (matched defects) (Fig. 3). Conditions in which the ventilation abnormality
appears larger than the perfusion abnormality (reverse mismatch) include airway
obstruction, mucous plug, airspace disease, atelectasis or pneumonia (Fig. 4). Patients
Thromboembolic Disease 77
Figure 2. 99mTc-DPTA aerosol ventilation (A) and 99mTc-MAA perfusion (B) images
demonstrate multiple segmental and subsegmental perfusion defects (arrows) in
regions which are ventilated normally (V/Q mismatch). The findings indicate a
high probability of acute pulmonary embolism. (Same image order as Fig. 1.)
with metabolic alkalosis, limited pulmonary vascular reserve or patients treated with
inhaled bronchodilators may also have failure of hypoxic pulmonary vasoconstriction
resulting in reverse mismatch.
78 Nuclear Medicine
Figure 4. 99mTc-DTPA aerosol and 99mTc-MAA images in an intensive care unit patient
with a low PaO2 demonstrates generalized decreased ventilation within the entire
left lung. The ventilation defects are more prominent then the perfusion defects
(reverse mismatch) indicating partial airway obstruction within the left main
bronchus. The patient subsequently underwent bronchoscopy which confimed
mucous plugging of the left main bronchus.
are between 0.5-3 microns in size and are produced by adding the appropriate ra-
diopharmaceutical to a commercially available nebulizer. For a routine ventilation
study, 1.11 GBq (30 mCi ) of 99mTc-DTPA in 3 ml of saline is placed within the
nebulizer. Oxygen is then forced through the nebulizer at high pressure to produce
aerosolized droplets which are inhaled by the patient through a mask or mouth
piece. The patient generally breathes from the nebulizer for 3-5 minutes or until 37
MBq (1 mCi) of activity is deposited within the lungs. The distribution of activity
within the lungs is proportional to regional ventilation. Multiple image projections
can be obtained which correspond with those obtained during subsequent perfusion
imaging. Ventilation studies with 99mTc-labeled radioaerosols require minimal patient
cooperation, and portable studies or studies in patients on respirators can be
performed relatively easily. In general, ventilation is performed prior to perfusion
imaging but ventilation imaging with radiolabeled aerosols can also be effectively
performed following perfusion imaging. Thus, in patients with a normal or near
normal perfusion study, ventilation imaging can be omitted. Disadvantages of 99mTc-
labeled radioaerosols relate primarily to central deposition of activity in patients
with COPD or airway obstruction and the amount of activity which is wasted within
the nebulizer.
Because of the problem with central deposition of 99mTc-labeled radioaerosol in
patients with COPD, newer agents have been developed. These include 99mTc-
Technegas and 99mTc-Pertechnegas. Both of these agents are formed by burning
99mTc-pertechnetate in a carbon crucible at very high temperatures which produces
Figure 7. Single coronal SPECT image through the mid thorax demonstrates increased
accumulation of 99mTc-DMP-444 (99mTc-GP IIb/IIIa antagonist) in a patient with
documented PE within the right lower lobe, left main and left lower pulmonary
arteries (arrows).
of the patients with a normal V/Q scan had PE. The overall diagnostic performance
of the V/Q scan was similar in men and women, in patients of varying ages, and
among patient with or without pre-existing cardiac or pulmonary disease.
Patients with suspected pulmonary embolism can be safely managed based on
pretest probability and results of V/Q scanning. One prospective study in 1239
patients with suspected pulmonary embolism examined a clinical model used to
categorize pretest probability of pulmonary embolism as low, moderate, or high
(Wells, 1998). V/Q scanning and bilateral deep venous ultrasonography were
performed. Only 3 of the 665 patients (0.5%) with low or moderate pretest
probability and a non-high-probability scan had PE or DVT during the 90-day
follow-up period.
5 Interpretation Pitfalls
False negative lung scan interpretations (low probability despite angiographic
evidence of PE) do occur and patients who have a recent history of immobilization,
surgery, trauma to the lower extremities and central venous instrumentation are a
particularly high risk group. In patients with low or very low probability V/Q scan
interpretations and no history of immobilization, recent surgery, trauma to the lower
extremities or central venous instrumentation, the prevalence of PE is only 4.5%. In
contrast, when patients with a low or very low probability V/Q lung scan
interpretations have one of the above mentioned risk factors, the prevalence of PE is
12%; with more than one risk factors, the prevalence rises to 21% (Table 1). Patients
with false negative lung scan interpretations tend to have non-occlusive and/or
subsegmental thrombi with low pulmonary clot burden. In recent years, concern
has been raised that a low probability lung scan interpretation may be misleading
and result in unnecessary mortality as a sequela of PE. Nevertheless, the prognostic
value of a low probability scan is excellent, particularly in patients with a low clinical
pre-test likelihood of disease or negative ultrasound. In a recent series of 536
consecutive patients with this finding, there was no evidence that PE was a causative
or contributing factor among patients who died within six months of imaging
(Rajendran, 1999).
The most common cause of V/Q mismatch in patients who do not have acute
PE are related to chronic or unresolved PE. Other causes of V/Q mismatch in the
absence of PE (false positive interpretation) include extrinsic compression of the
pulmonary vasculature (mass lesions, adenopathy, mediastinal fibrosis), vessel wall
abnormalities (pulmonary artery tumors, vasculitis), intraluminal obstruction (tumor
emboli, foreign body emboli) and congenital vascular abnormalities (pulmonary
artery agenesis or hypoplasia). In patients with unilateral V/Q mismatch
(hypoperfusion or absent perfusion) within an entire lung or multiple contiguous
segments, and normal perfusion in the contralateral lung, extrinsic compression of
the pulmonary vasculature, congenital abnormalities or proximal PE all need to be
considered in the differential diagnosis. These patient will often require further
imaging with CT or angiography.
Interpretation Criteria
Several diagnostic criteria have been developed for the interpretation of V/Q
lung scans. The original PIOPED criteria had the highest likelihood ratio for
predicting the presence of PE on pulmonary angiography. However, these also had
Thromboembolic Disease 85
Table 1. Value of combining selected risk factors and the lung scan
interpretation
Prevalence of PE
V/Q scan 0 Risk factors* 1 Risk factor* *2 Risk factors*
interpretation (%) (%) (%)
High 82% 84% 97%
Intermediate 25% 37% 45%
Low/Very Low 4% 12% 21%
Normal 0% 0% 0%
*risk factors include immobilization for >3 days prior to presentation, recent
surgery, trauma to lower extremeties or central venous instrumentation within 3 5
months of presentation
Normal
No perfusion defects
The stripe sign is defined as a rim of perfused lung tissue between the perfu-
sion defect and the adjacent pleural surface. The presence of the sign excluded the
diagnosis of PE within the affected zone in 93% of cases. Therefore, perfusion de-
fects which demonstrate a stripe sign are unlikely to be due to PE and in the absence
of perfusion defects elsewhere should be interpreted as representing a low probabil-
ity for PE. Patients with partially resolving perfusion defects may have a similar
appearance to the stripe sign, therefore this sign should be interpreted with caution
in patients with more chronic symptoms.
Finally, the nuclear medicine physicians subjective estimate of the likelihood of
PE (without using specific interpretation criteria) correlated well with the fraction
of patients with angiographic evidence of PE. Thus, experienced readers (such as the
PIOPED investigators) can provide an accurate estimate of the probability of PE
based on radiographic and scintigraphic findings.
Thromboembolic Disease 87
Diagnostic Approach
Prospective and outcome-based studies permit the following conclusions in the
diagnostic evaluation of patients with suspected PE (summarized in Fig. 8):
A normal V/Q scan interpretation excludes the diagnosis of clinically
significant PE.
Patients with very low or low probability V/Q scan interpretation and low
clinical likelihood of PE do not require angiography or anticoagulation.
Patients with very low or low probability V/Q scan interpretation,
intermediate or high clinical likelihood of PE, and negative serial
noninvasive venous studies of the lower extremities do not require
anticoagulation or angiography. If noninvasive venous studies of the lower
extremities are positive, patients should be treated.
Clinically stable patients with an intermediate probability V/Q scan
interpretation require noninvasive venous studies of the legs and, if
negative, require pulmonary angiography or CT angiography for a definite
diagnosis.
88 Nuclear Medicine
new symptomatic episodes. Any resolution that occurs on the V/Q scan occurs within
3 months. A follow-up V/Q scan, therefore, is particularly useful at 3 months after
the acute PE. If the patient is unable to return in 3 months, a V/Q scan at discharge
may be useful if there is a prolonged hospitalization.
If the clinical circumstances indicate the necessity of an
intervention to reduce clot burden in patients with suspected
acute PE, and if thrombolytic therapy is the intervention of
choice, is pulmonary angiography always necessary to confirm
the diagnosis?
There is disagreement on this point. Some believe that the risk of serious and
perhaps fatal bleeding with thrombolytic therapy is so great that a pulmonary
angiogram must be obtained in all patients. Others believe that if there is a strong
5
clinical suspicion of acute PE supported by a high probability V/Q scan (or in an
extreme emergency, a noninvasive leg test showing DVT or an echocardiogram
showing right ventricular dysfunction) then thrombolytic therapy may be
administered on the basis of these noninvasive tests without a pulmonary angiogram.
In the absence of a pulmonary angiogram, a high probability V/Q scan in the
appropriate clinical setting offers the most useful information.
Does a ventilation scan need to be performed in all patients, or
are there patients in whom the combination of a plain chest
radiograph and perfusion scan are sufficient?
If V/Q scans are obtained with 99mTc-aerosols, a chest radiograph and a perfusion
scan can be obtained initially and evaluated by the nuclear medicine physician. If a
ventilation scan is deemed necessary, it can be performed by postperfusion techniques.
If the perfusion scan is normal, no ventilation scan is needed. If the perfusion scan
shows characteristic vascular defects in regions where the chest radiograph is normal,
this would indicate a high probability for PE and a ventilation scan is unnecessary.
In patients with prior cardiopulmonary disease, in whom matched perfusion and
ventilation defects are likely to occur, both a ventilation and perfusion scan are often
indicated.
What is the role of contrast-enhanced helical (spiral) CT?
This diagnostic modality is still under investigation and no firm general conclu-
sions can be made without more extensive experience. It appears to be a useful
addition to the panel of tests available for the diagnosis of PE, particularly in central
arteries. In institutions where experience and skill with this modality are available, it
can have value in diagnosing central PE in circumstances wherein established
diagnostic tests are not immediately available. A normal contrast-enhanced CT scan
does not exclude PE, particularly in subsegmental arteries. In patients with renal
insufficiency, consideration should be given to the consequences of the load of
92 Nuclear Medicine
Bone Densitometry
William D. Leslie and Bruce E. Roe
Pathophysiology of Bone Loss and Osteoporotic Fractures
Pathophysiology of Bone Loss
Bone is a dynamic tissue consisting of cellular, organic and inorganic components
with a complex internal structure (Fig. 1). It is much more than passive scaffolding
for the rest of the body and undergoes considerable metabolic activity and remodeling.
The mature skeleton consists of a mixture of cortical bone (85%) and trabecular
bone (15%). The relative amounts of these vary widely between different anatomic
sites (Fig. 2).
Bone tissue reacts to stress and injury through a well-orchestrated sequence for
removing old bone and building new tissue. Bone remodeling is carried out by the
basic multicellular unit (BMU), which consists of both osteoclasts and osteoblasts.
The BMU typically takes 3-6 months to complete a cycle (Fig. 3). Bone remodeling
affects 3-5% of cortical bone per year, but involves up to 25% of trabecular bone
due in part to its greater relative surface area. The osteoclast, a multinucleated cell of
monocyte origin, resorbs bone through the release of acid and enzymes such as
cathepsin K from its ruffled border. Osteoblasts, derived from mesenchymal cells,
enter the resorption pit and lay down organic matrix (osteoid). The osteoblast then
dies or enters a dormant stage and is known as an osteocyte (if trapped within calcified
tissue) or a lining cell (if found on the surface of calcified tissue). The osteoid is
subsequently mineralized over a period lasting several months. There is close coupling
of osteoclast and osteoblast activities, although the intercellular signaling involved is
incompletely understood. It is clear, however, that processes which stimulate (or
suppress) one cell type result in stimulation (or suppression) of the other. For example,
after menopause, osteoblast activity increases in an attempt to compensate for
increased osteoclastic resorption. On the other side of the equation, antiresorptive
treatments targeted at suppressing osteoclast activity are only able to achieve a slight
gain in bone mass as there is a parallel reduction in osteoblast activity (Fig. 4). The
bone remodeling cycle is regulated by a myriad of factors, including growth factors
and interleukins. A new molecule, osteoprotegerin ligand (OPGL), has recently been
discovered and shown to be the primary regulator of osteoclast activity. Estrogen
deficiency, glucocorticoid excess and vitamin D deficiency are common conditions
where bone metabolism is altered.
Bone cell activity can be evaluated through the measurement of biochemical
markers. Osteoblasts produce type I collagen (the primary collagen of bone tissue),
non-collagenous proteins (such as osteocalcin) and enzymes (such as alkaline
phosphatase). Many of these are measured in the serum as indices of bone formation.
Figure 2. Percent of trabecular bone in skeletal sites commonly assessed with bone
densitometry.
Bone Densitometry 95
Figure 3. Bone remodelling cycle. Proceeding from left to right the sequence is (a)
osteoclast resorption, (b) osteoblast proliferation, (c) osteoid matrix deposition and
(d) mineralization.
6
Nutritional
Low calcium intake
Vitamin D deficiency
Malabsorption
Drugs
Excess glucocorticosteroids
Anticonvulsants
Heparin
6 Lifestyle
Immobility
Smoking
Alcohol
Endocrine/Metabolic
Hypogonadism
Early menopause or prolonged hypoestrogenism
Glucocorticosteroid excess
Hyperparathyroidism
Hyperthyroidism
Diabetes mellitus
Renal insufficiency
Other
Myeloma
Malignancy
(especially calcium intake). The specific genes that influence peak bone mass have
yet to be elucidated, but twin studies suggest that 80% of the variation in peak bone
mass is determined by genetic factors. Ethnicity is also important. Blacks have higher
average bone mass than Caucasians who in turn have higher average bone mass than Asians.
After early adulthood, both men and women start a slow, inexorable decline in
bone mass that continues until death. Bone turnover accelerates in women at the
time of menopause, especially in trabecular bone, and usually results in the loss of 5-
15% of bone mass over the first 5 years after menopause. After early menopause,
age-related bone loss continues at a rate of 0.5-1% per year. The pathogenesis of
age-related bone loss is unclear, but may be related in part to changes in calcium
absorption and vitamin D availability. Other factors which can increase bone loss
are detailed in Table 1.
Pathophysiology of Osteoporotic Fractures
Osteoporosis is defined by the World Health Organization as a systemic skeletal
disease characterized by low bone mass and microarchitectural deterioration of bone
tissue, with a consequent increase in bone fragility and susceptibility to fracture.
Bone Densitometry 97
Bone mass is the primary determinant of bone strength, as studies of excised bone
have demonstrated that 80-90% of bone strength is determined by the amount of
bone. Measures of bone mass from different skeletal sites correlate moderately well,
but site-to-site differences are not uncommon and may reflect genetic factors,
hormonal influences and the individuals level and pattern of activity.
Microarchitectural deterioration also has an important effect on bone strength.
Typical changes in trabecular bone include reduction in trabecular thickness and
number, and perforation of trabeculae by deep resorption pits (Fig. 1). This results
in loss of trabecular connectivity which is currently believed to be irreversible. These
microarchitectural changes may not be reflected by reductions in bone density but
still contribute to fracture susceptibility.
Although this discussion has focused on bone strength as a determinant of fracture
risk, it is important not to overlook the many clinical factors that independently
contribute to fracture risk assessment (see Section on Clinical Risk Factors to Predict
Bone Density). 6
Technical Aspects of Bone densitometry
General Principles
All bone measurement techniques rely upon the ability of bone to block
transmission of energy. In general, the more dense the bone the less energy is
transmitted. The physical forms of energy used in clinical bone densitometry are x-
rays (generated from an x-ray tube), gamma rays (released from decaying
radionuclides) and sound (emitted from an ultrasonic transducer). Ideally, a bone
density method would have high accuracy and precision (or reproducibility), and
would be rapid, inexpensive, painless and safe (with little or no ionizing radiation).
At the present time, no method completely satisfies all of these criteria.
Bone Density Technologies
Dual Energy X-ray Absorptiometry (DEXA)
DEXA grew out of dual photon absorptiometry (DPA) which used gadolinium-
153 as the photon source. This method suffered from the need for radionuclide
source changes, poor image resolution and reproducibility, and has been largely
replaced by DEXA. DEXA uses an x-ray tube to generate two different x-ray energies.
Bone blocks (or attenuates) x-rays to a greater degree than soft tissue, and lower x-
ray energies are attenuated more than higher energies. An x-ray detector records the
amount of attenuation for the two energies and can calculate both the amount of
soft tissue and the amount of bone calcium in the path of the beam. The x-ray tube
and detector scan over the area of interest, building up an image of bone mineral
content (BMC) expressed in grams of calcium. The densitometers software identifies
the projected bone area using an edge-detection algorithm. Dividing BMC (grams
of calcium) by the bone area (cm2) yields a real bone mineral density (BMD as
grams/cm2). DEXA has the advantage of being rapid (particularly with newer scanners
that use higher output x-ray tubes and a fan-beam configuration) and are able to
scan the structures of greatest clinical interest such as spine, hip (Fig. 5), forearm
and even total body. The radiation dose from a lumbar spine scan is much less than
10 Sv (1 mRem), a value similar to one days normal background radiation and of
98 Nuclear Medicine
Figure 5. Normal DEXA scans of the spine (top) and hip (bottom). Results are plotted
in the right panels in relation to the normal age- and gender-matched reference curves.
negligible risk. Radiation doses from scans of the hip, forearm and total body are
even lower.
devices avoid these limitations and impart an exceedingly small radiation dose of
less than 0.1 Sv (0.01 mRem).
Quantitative ultrasound
Ultrasound has recently emerged as another tool for characterizing bone strength
and has the advantages of being radiation-free and using relatively inexpensive,
portable devices. Ultrasound penetrates bone poorly, and higher frequencies are
attenuated more than lower frequencies. Two measures are typically derived from
quantitative ultrasound (QUS). One is the speed of sound (SOS), a measure of the
speed with which sound travels from one transducer to the other through the bone
(m/sec). The other is broadbeam ultrasound attenuation (BUA) which is the slope
of the relationship between attenuation and frequency (dB/mHz) as shown in Figure
6. Due to ultrasounds difficulty in penetrating deep structures, most devices measure
the more accessible bones such as the calcaneus, phalanges and tibia. While x-ray-
based techniques are calibrated against calcium content, there is considerable
100 Nuclear Medicine
show only moderate correlation with bone density measured at proximal skeletal
sites, though this is similar to the ability of peripheral x-ray-based measurements to
predict central bone density. The accuracy of QUS can be indirectly validated through
its ability to identify patients at risk of fracture. To date, this has only been reported
for a small number of the available instruments, but available data show that fracture
prediction with calcaneal QUS is comparable to that obtained with x-ray-based
techniques.
Precision
Precision (also referred to as reproducibility) is the ability of a system to obtain
the same results in repeated measurements of the same individual. A technique must
have good precision if serial measurements are to be used in following an individual.
Greater precision makes it possible to detect smaller changes in a subject. Current
methodologies typically demonstrate precision errors that are larger than annual
changes in bone density. Thus, in an individual patient, it may be difficult to
determine whether a small change in the bone mass measurement reflects precision
error or true change (Table 2).
DEXA reproducibility is influenced by instrument-, operator- and subject-
dependent factors. These last two tend to be much more important than the
instrument itself, and patient positioning is the single most important determinant.
Reproducibility is optimized through a systemic process that includes careful quality
control of the instrument, scanning technique and analysis (Table 3). Hip
measurements are less reproducible than those of the spine, in large part due to the
difficulty in obtaining consistent positioning. Reproducibility is further compromised
when examining smaller regions of interest. Femoral neck precision is much worse
102 Nuclear Medicine
than the total hip region, and Wards triangle (an area within the femoral neck that
contains little trabecular bone) is so variable that it is of little clinical value.
Precision can be stated as either standard deviation (SD) or percent coefficient of
variation (%CV, defined as 100 x SD/mean). Although precision error is commonly
stated as a %CV, several reports indicate that error is independent of bone mass and
will therefore be underestimated in the lower (osteoporotic) range. This suggests
that it may be preferable to express precision error as the absolute SD. The smallest
change that must be present before one can conclude (with 95% confidence) that
the change is not related to measurement error is 2.8 x SD (or 2.8 x %CV).
The best sources of short-term precision data come from large multicentre studies
that have used replicated patient measurements. Vendors frequently cite in vivo
Bone Densitometry 103
Figure 9. Spine artifact from a barium swallow performed one week earlier. Barium
can be seen throughout the colon (upper) and had cleared out after a month (lower).
Barium caused bone density to be underestimated (L2-L4 1.246 g/cm2 with T-score
+0.39) versus the repeat scan (1.541 g/cm2 with T-score +2.84).
Figure 10. Pagets disease of the right hip. The bone expansion and sclerosis give a
bone density measurement of the femoral neck that is well above normal (1.314 g/
cm2 with T-score +2.78) while the left hip is actually in the osteoporotic range
(0.634 g/cm2 with T-score -2.85).
differences are usually small (1-2%) but on occasion can be clinically significant
(exceeding instrument reproducibility). Therefore, measurements from different
machines are very difficult to compare, and whenever possible follow-up examinations
should be performed on the same machine.
Clinical Role of Bone densitometry
T-Score and Z-Score
Absolute measurements of bone density are of little value since they are determined
by the site of measurement, calibration used by the equipment manufacturer, and
even the particular instrument. Since bone density follows a bell-shaped (Gaussian)
distribution, measurements are conventionally described in terms of the number of
Bone Densitometry 105
standard deviations (SD) that a value deviates from the population mean. Age-re-
lated changes in bone density, described earlier, must be taken into account. The Z-
score refers to the number of SD above or below the mean for an age-matched
population. The T-score refers to the number of SD above or below the mean for a
young adult population. It is not enough for a bone density instrument to provide
accurate and reproducible measurements. To interpret a patient value there must be
valid normal data from a large reference population that is patient-appropriate (age,
gender and ethnicity).
In interpreting an individual patients test results, the following question is
confronted: Should you compare the patient with someone of the same age or with
a young adult? The former masks the increasing prevalence of osteoporosis with
advancing age, while judging an 80-year-old against the same standard used in a 30-
year-old seems unreasonable. In reality, both approaches have merit and are
complementary (Fig. 12). An age-adjusted measurement indicates whether the
individual is average for their age and, if not, how markedly they deviate from the
expected value. On the other hand, bone strength depends upon bone mass and not
the age of the subject, therefore predictions in terms of fracture risk are best based
upon comparison with an absolute standard (young adult).
Diagnosis of Osteoporosis from Bone Density Measurements
The World Health Organization (WHO) formulated diagnostic ranges for
osteoporosis based upon T-score. These ranges were originally intended to be used
epidemiologically, but have subsequently been applied to individuals. The data
reviewed for these recommendations was almost exclusively derived from post-
106 Nuclear Medicine
Figure 11. Quality assurance plots from three hypothetical systems with mean 1.000,
standard deviation 0.005 (coefficient of variation 0.5%). The top plot indicates a
stable machine, the middle plot indicates an abrupt shift in baseline (regression
lines plotted to the data prior to and after the shift) and the bottom plot indicates a
continuous drift in baseline (regression line plotted to all time points). The latter is
particularly insidious and would be difficult to identify by visual inspection of the
data alone.
6
Figure 12. An 80-year-old woman with bone mineral density (BMD) of the hip that
is average for her age (648 mg/cm2) will have a Z-score of -0.5 but a T-score of -2.5.
A 30-year-old woman with exactly the same bone density measurement would
have a T-score and a Z-score that are both -2.5.
(or T-score). The increase in fracture likelihood with decreasing bone density is ex-
ponential (not simply additive) (Fig. 14), and is given by the formula:
Fracture Risk = Average Risk x RR-Z-Score
An individuals risk of fracture is dependent upon (a) average risk for age, (b) the
Z-score of the bone density measurement and (c) relative risk per SD change in
bone density. The influence of age on fracture likelihood cannot be overstated as it
occurs independent of any age-related decline in bone density as seen in Figure 15.
This age-related fracture risk is in part secondary to an increased likelihood of falls
in the elderly as well as microarchitectural changes in bone (such as increased bone
turnover) that are not directly measurable with bone density.
Different technologies and skeletal sites show markedly different age-related
changes in the T-score (Fig. 16). Lumbar spine QCT shows the earliest and most
rapid change with the average woman crossing the osteoporotic threshold (T-Score
-2.5) by age 61. In contrast, peripheral measurements are much less age-responsive
and ultrasound of the calcaneus will not reach the same threshold until after age
100. T-scores cannot be used interchangeably between different sites and techniques.
Clinical Risk Factors to Predict Bone Density
There have been many previous attempts to use clinical risk factors to predict
BMD and these have been largely unsatisfactory. The largest experience comes from
the Study of Osteoporotic Fractures (SOF) which used a multivariate model based
upon over 100 clinical risk factors to predict BMD in 7,963 ambulatory Caucasian
women 65 years age or older. The best such model was only capable of predicting
21% of BMD variance, a level that is insufficient for clinical purposes.
Certain medical disorders can be important secondary causes of rapid bone loss
(see Section on Clinical Management of Osteoperosis). Among healthy peri- and
post-menopausal women is a subset that also loses bone at a rapid rate, appropriately
108 Nuclear Medicine
termed fast losers. Unfortunately, this group has few distinguishing features. There
is some evidence that underweight women and those with higher initial bone density
are at risk for more rapid bone loss. Such statements are of limited help in an individual
patient due to the large degree of overlap.
Clinical Risk Factors to Predict Fractures
It may at first appear paradoxical that clinical risk factors can still be useful in
6 predicting fracture risk given the poor correlation with bone density. Many of the
factors important in the pathophysiology of fracture are not measurable with bone
densitometry (Fig. 17). Several large studies have identified clinical markers for hip
fracture that operate independent of bone density measurement (Table 5). Data
from the Study of Osteoporotic Fractures (SOF) indicate that when there are few
clinical risk factors, hip fracture rates are very low (1.1-2.6 hip fractures per 1000
women-years) (Fig. 18). In this study, women in the lowest bone density tertile with
less than three clinical risk factors had a substantially lower fracture rate than those
in the highest bone density tertile with at least five clinical risk factors (2.6 hip
fractures per 1000 women-years vs 9.4 hip fractures per 1000 women-years).
The single most powerful predictor of future osteoporotic fracture is the presence
of previous such fractures. A single vertebral fracture places that individual at higher
risk of fracture than low bone density alone (Fig. 19). The combination of low bone
density and previous fracture increases risk 25-fold and the presence of two vertebral
fractures increases that risk 75-fold. Loss of height, a marker of asymptomatic vertebral
deformities, is also a predictor of hip fractures (RR 1.7 if height loss >3 cm).
Falling is central to the pathophysiology of hip fracture and it is not surprising
that studies have found that a history of fall in the preceding 12 months was a
powerful (RR 1.6-2.4) predictor of hip fracture. An increased risk of falls is also seen
with use of long-acting benzodiazepines and tricyclic antidepressant medications.
Smoking has also emerged as an independent predictor of hip fracture (RR 2.0),
likely due to direct effects on bone metabolism and indirect effects leading to low
body weight and reduced overall health.
Evaluation of these clinical risk factors is valuable in determining the need for therapy.
Following Osteoporosis with Bone densitometry
Follow-up bone mass measurements in patients not receiving active treatment
can help in the identification of individuals with rapid bone loss (fast losers).
Repeat testing may also be useful in confirming a positive treatment response,
although some evidence suggests that much of the anti-fracture effect of current
antiresorptive therapies is mediated through mechanisms other than increasing bone mass.
Bone Densitometry 109
6
Figure 14. A 70-year-old woman has bone density measurements of the hip with T-
score -4.2 and Z-score -3. Relative risk of hip fracture (compared with an average
70-year-old woman) increases with relative risk (RR) of 2.6 for each standard
deviation (SD) below average. Therefore, relative risk of hip fracture is increased
18-fold (2.63), placing the woman at markedly increased risk.
treatment induces a shift to the right of the bone loss distribution, decreasing the
number of cases in which rapid loss occurs.
The optimal time interval for follow-up measurements is a function of machine
precision and the expected rate of bone loss. For example, if a subject loses bone
mass at a rate of 1% per year then it would take 3 years for this to exceed (with 95%
confidence) the precision limits of a machine with optimal performance (CV 1%)
and 6 years for a typical machine (CV 2%). To aid in ordering a repeat bone mass
measurement, the relevant parameters are presented in Table 2 by relating test
precision (%CV), maximum expected rate of bone loss (percent per year) and the
minimum time interval between tests required to detect a change of this magnitude
(with 95% statistical confidence). Thus, to exclude bone loss in the total hip (CV
2.5%) that is three times an average value of one percent, a follow-up interval of 2.5
years is appropriate. Finally, it should be remembered that minimum detectable
bone loss is a statistical concept and needs to be distinguished from a clinically
significant change in skeletal mass. The latter will be influenced by a history of
symptomatic fractures, the severity of osteoporosis and use of active treatment.
Clinical Management of Osteoporosis
General Assessment
Although osteoporosis is most commonly seen in postmenopausal women and
the elderly, a number of secondary causes of bone loss should be considered in the
evaluation of an individual with low bone mass (Table 1). These factors may exist in
isolation, or accelerate bone loss in postmenopausal osteoporosis. The clinical
Bone Densitometry 111
Figure 15. Estimated incidence of fracture as a function of age and bone mass.
(From Hui SL et al. J Clin Invest 1988; 81:1804. Reproduced with permission of the
American Society for Clinical Investigation, Inc.)
assessment should be directed towards elucidating these potential causes and any
fracture history. Assessment of the risk factors for falls given in Table 5 is an important
adjunct to measuring bone density. Laboratory testing can be limited to the
measurement of serum calcium, alkaline phosphatase, creatinine and a complete
blood count. In individuals with postmenopausal or age-related osteoporosis, all of
these indices should be within the normal range. These investigations may be further
expanded to include serum TSH, parathyroid hormone (PTH), serum 25-
hydroxyvitamin D, protein electrophoresis, and 24 hour urinary calcium
determination, as guided by clinical judgment. Although not routinely required,
lateral x-rays of the thoracolumbar spine help to determine the number and type of
112 Nuclear Medicine
Figure 16. Age-related decline in mean Caucasian female T-scores for different
BMD technologies based on manufacturer reference ranges. Total hip values (
9) are from the NHANES data implemented by all DEXA manufacturers. Also
shown are L1-L4 PA spine (), L2-L4 lateral spine (), one-third forearm
() from the Hologic QDR-4500 DEXA; calcanus ultrasound () from the
Hologic Sahara, spinal QCT ( ) from the Image Analysis system. (From Faulkner
K et al. J Clin Densitometry 1999; 2:343-350. Reproduced with permission from
Humana Press Inc.)
pre-existing vertebral fractures, and their presence indicates high risk for further
fractures. Biochemical markers of bone metabolism may provide an indirect method
to evaluate the rate of bone turnover. Osteocalcin, bone-specific alkaline phosphatase
and procollagen I peptide can assess the level of bone formation while urinary
excretion of collagen crosslinks (such as deoxypyridinoline and N-telopeptide) reflect
the level of bone resorption. These markers have not been useful for diagnosing
osteoporosis or for predicting bone loss, but some studies suggest that they
independently predict fracture risk.
Non-Pharmacologic Therapy
General measures to reduce the risk of fractures should include assessment of
hazards in the home environment, sedative use, muscle weakness, postural
hypotension and uncorrected visual deficits. Exercise should also be encouraged in
an attempt to preserve bone mass and to maintain or improve muscular conditioning.
Calcium supplementation is often necessary in postmenopausal and elderly
women to reach recommended targets (1,200-1,500 mg per day). Calcium carbonate
is the most commonly used supplement and is recommended for those who do not
have achlorhydria (common in the very elderly and those on acid suppressing
Bone Densitometry 113
Estrogen
Hormone replacement therapy (HRT) is effective at preventing bone loss in
postmenopausal women and is associated with a bone density increase of 4-6% in
the first 2 years of therapy. In epidemiological studies HRT has been associated with
a 50% reduction in fracture risk, but protection is lost with cessation of therapy.
HRT is also associated with a 30% increase in the risk of breast cancer, increased risk
for thromboembolism and must be taken with a progestin if the patient has not had
a hysterectomy to avoid endometrial hyperplasia/carcinoma. The associated bleeding
is unacceptable for some women, especially in older age groups. HRT is associated
with a reduced incidence of cardiovascular events in cohort studies, but an increased
early event rate seen with a secondary prevention trial underscores the need for data
from prospective controlled trials of primary prevention such as the ongoing Womens
Health Initiative.
114 Nuclear Medicine
Table 5. Risk factors for fracture with and without adjustment for history of
previous fracture and calcaneal bone density. (Adapted from Cummings SR et al.
N Eng J Med 1995; 332:767-773.)
Relative Risk (95% confidence interval)
Risk factor Base Model Adjusted for
fracture history and
bone density
Age (per 5 years) 1.5 (1.3-1.7) 1.4 (1.2-1.6)
Maternal hip fracture 2.0 (1.4-2.9) 1.8 (1.2-2.7)
Weight increase since age 25 0.6 (0.5-0.7) 0.8 (0.6-0.9)
(per 20%)
Height at age 25 (per 6 cm) 1.2 (1,1-1.4) 1.3 (1.1-1.5)
Self-rated health (per 1-point 1.7 (1.3-2.2) 1.6 (1.2-2.1)
decrease)
6 Previous hyperthyroidism 1.8 (1.2-2.6) 1.7 (1.2-2.5)
Current use of long-acting 1.6 (1.1-2.4) 1.6 (1.1-2.4)
benzodiazepines
Current use of anticonvulsants 2.8 (1.2-6.3) 2.0 (0.8-4.9)
Current caffeine intake 1.3 (1.0-1.5) 1.2 (1.0-1.5)
(per 190 mg/day)
Walking for exercise 0.7 (0.5-0.9) 0.7 (0.5-1.0)
On feet less than 4 h per day 1.7 (1.2-2.4) 1.7 (1.2-2.4)
Inability to rise from chair without 2.1 (1.3-3.2) 1.7 (1.1-2.7)
using arms
Lowest quartile for visual depth 1.5 (1.1-2.0) 1.4 (1.0-1.5)
perception
Lowest quartile for visual low- 1.2 (1.0-1.5) 1.2 (1.0-1.5)
frequency contrast sensitivity
Resting pulse rate over 80 per min 1.8 (1.3-2.5) 1.7 (1.2-2.4)
Any fracture since age 50 - 1.5 (1.1-2.0)
Calcaneal bone density (per - 1.6 (1.3-1.9)
SD decrease)
SERMs
New agents have been developed that retain estrogens positive effects on bone,
while attempting to minimize the associated risks and side effects. This class of
drugs is termed Selective estrogen receptor modulators (SERMs) and include
raloxifene and tamoxifen. Raloxifene reduces the risk of vertebral fractures by
approximately 40%, and produces increases in bone density of approximately 3% at
the spine in its first three years. It also lowers total and LDL cholesterol, does not
stimulate the endometrium, and probably reduces the risk of breast cancer.
Calcitonin
Salmon calcitonin has been used subcutaneously for many years. The recently
released intranasal form is more acceptable for longterm use. Despite modest effects
on bone density, vertebral fracture risk is reduced by approximately 40%. Intranasal
calcitonin is well-tolerated and has minimal side-effects.
Bone Densitometry 115
Figure 18. Independent interaction between clinical risk factors and bone mineral
density. Risk factors were: age over 80, maternal history of hip fracture, any fracture
since age 50, reduced self-rated health (fair or worse), previous hyperthyroidism,
anticonvulsant therapy, current use of long-acting benzodiazepines, current weight
less than at age 25, height at age 25 over 168 cm, caffeine intake more than the
equivalent of two cups of coffee per day, on feet less than four hours per day, no
walking for exercise, inability to rise from a chair without using arms, lowest quartile
for visual depth perception, lowest quartile for visual contrast sensitivity, resting
pulse rate greater than 80 beats per minute. (Adapted from Cummings SR et al. N
Eng J Med 1995; 332:767-773.)
Bisphosphonates
This class of drugs inhibit osteoclast number and activity and have few extra-
skeletal effects. Etidronate was the first available compound. It has low potency and
inhibits bone mineralization when used continuously (though this is not seen with
intermittent regimens). Alendronate is a potent aminobisphosphonate that is capable
of increasing bone density at the lumbar spine by 8% over 3 years. More importantly,
it has been proven to reduce hip and spine fractures by 50% in postmenopausal
women with previous fractures. The main risk is erosive esophagitis, especially in
patients with prior esophageal disease, gastroesophageal reflux or when directions
are not carefully followed. Risedronate has also been demonstrated to produce similar
116 Nuclear Medicine
Figure 19. Risk for incident vertebral fractures (Vb Fx) based on current vertebral
fracture status. (Adapted from Ross PD et al. Ann Intern Med 1991; 114:919.)
Anabolic Agents
Strategies which directly stimulate osteoblast activity have the potential to produce
larger increases in bone density and potentially larger effects on fracture prevention.
One such strategy is the use of intermittent parathyroid hormone therapy which has
been demonstrated to produce large increases in bone mineral density in various
populations, and to reduce vertebral fractures by approximately 65% in
postmenopausal women with previous fractures.
Who to treat?
General recommendations that apply to the entire population include ensuring
adequate calcium and vitamin D intake, engaging in regular weight-bearing exercise,
avoidance of smoking and limitation of alcohol use. Although whole population
testing is not advocated, the National Osteoporosis Foundation (NOF) has defined
guidelines for targeted screening(Fig. 20). In addition, the NOF has identified four
clinical risk factors as being sufficiently common and predictive of osteoporosis in
postmenopausal women to be useful in the clinical setting for guiding decisions on
whether to recommend bone densitometry: (1) history of fracture after age 40; (2)
history of hip, wrist or vertebral fracture in a first-degree relative; (3) being in the
lowest quartile for body weight (less than 57.8 kg); and (4) current cigarette smoking.
The NOF established therapeutic cutpoints for antiresorptive therapy. According
to these guidelines, therapy to reduce the risk of fracture should be initiated in the
following groups: women with a T-score below -2; or women with a T-score -1.5 if
any of the four risk factors given above are present. In addition, women over age 70
Bone Densitometry 117
Figure 20. Guidelines for the targeted screening and treatment of osteoporosis.
Major risk factors include (1) history of fracture over age 40; (2) history of hip, wrist
or vertebral fracture in a first-degree relative; (3) being in the lowest quartile for
body weight (less than 57.8 kg); and (4) current cigarette smoking. (Adapted from
the National Osteoporosis Foundations Physicians Guide to Osteoporosis
Prevention and Treatment.)
with multiple risk factors or previous vertebral fracture should receive therapy. The
NOF cutpoints have not been universally accepted and do not take into account the
low absolute risk of fracture of younger women who may have osteopenia. In these
individuals, strategies to prevent bone loss should still be considered, such as HRT,
low-dose bisphosphonates or raloxifene. Decisions to start therapy must clearly take
into account the reversible and irreversible risk factors for fracture already described.
Frequently Asked Questions (FAQs)
Why dont all bone density instruments give the same readings?
Although DEXA instruments are ultimately calibrated against excised bone
samples, methodologic differences in how this is performed have led to large
discrepancies in patient measurements when performed on instruments from different
vendors. Efforts to reconcile these differences have led to a consensus on converting
DEXA measurements to standardized bone mineral density (sBMD). sBMD is
expressed in mg/cm2 to distinguish it from nonstandardized BMD which are in g/
118 Nuclear Medicine
cm2. Use of sBMD should ensure that average patient values are similar on different
instruments, but other methodologic differences exist and it should not be assumed
that patient sBMD values will be identical on all instruments. Due to these limitations
with current approaches to machine cross-calibration, small changes in bone density
may not be appreciated. Therefore it is strongly recommended that follow-up
measurements use the same instrument and scanning procedure.
Is it unusual to have a significant difference between a BMD
measurement of the spine and hip? Which is a more reliable
indicator of osteoporosis?
Apparent discrepancies between hip and spine BMD measurements are common
and emphasize the complexity of skeletal metabolism. As a systemic skeletal disorder,
osteoporosis affects all bones but the degree is modified by local determinants of
bone metabolism which include bone composition (trabecular bone undergoes more
6 rapid turnover and loss), mechanical loading (weight bearing enhances osteoblast
activity) and age-related artifacts (usually elevate spine BMD). Accuracy errors for
BMD measurements of the spine and hip are on the order of 5-7%. Together these
factors help to explain why differences between spine and hip BMD measurements
are so common.
The original World Health Organization (WHO) definition of osteoporosis based
on BMD (more than 2.5 standard deviations below young adult mean value) does
not specify a particular site. Although there is a tendency to use the lowest T-score in
classifying a patient, fracture rates are known to increase in proportion to the number
of osteoporotic sites. Therefore, all sites probably contribute independently to the
global fracture risk. When it comes to assessing the risk of specific fractures,
particularly hip fractures, a direct measurement of that site is preferred. The cost
and morbidity of hip fractures coupled with the strength of their association with
BMD has led to a recent proposal that the diagnosis of osteoporosis be reserved for
BMD measurements of the hip. This does not negate the important fracture prediction
information that can be obtained from other validated techniques and sites.
When evaluating fracture risk, it is the whole picture that counts rather than any
single measurement viewed in isolation. An accurate assessment needs to consider
information taken from both hip and spine, in addition to weighing non-BMD
predictors of fracture such as age and falls.
How do you diagnose osteoporosis in groups other than
postmenopausal Caucasian females?
The occurrence of fragility fractures, such as minimal trauma vertebral
compression fractures, is sufficient for a diagnosis of osteoporosis in any group
providing that other pathologic causes (such as tumour) have been excluded. The
criteria for diagnosing osteoporosis from bone density alone are more controversial
and caution is recommended before extrapolating the WHO recommendations.
For example, men have higher average bone density than women but also have a
lower fracture rate. The absolute likelihood of fracture in men seems to show the
same relationship to absolute bone density as in women. This has led to uncertainty
over whether the osteoporosis threshold in men should be based upon a male or
female reference group. Similar concerns arise over comparison to non-Caucasian
Bone Densitometry 119
Figure 21. Bone volume strongly affects measured bone density using areal
techniques such as dual energy x-ray absorptiometry (DEXA). Assume that two
cubes are constructed from hydroxyapatite (1 gram/cm3). The larger cube will have
twice the areal density of the smaller cube due to its greater depth since this is not
measured in DEXA.
groups. Black subjects have significantly higher bone density and lower fracture
rates than Caucasians, suggesting a similar relationship between absolute bone density
and absolute fracture risk. In contrast, orientals have lower hip bone density than
Caucasians but paradoxically have a much lower fracture likelihood. The reasons for
this are still debated but probably include shorter stature, likelihood of falling, and
possibly shorter femoral neck (also referred to as hip axis length). Normal children
represent a dramatic example of where low bone density has a surprisingly low rate
of fracture when one considers the frequency with which children are involved in
falls and other trauma. It is inappropriate to generate T-scores or apply WHO
diagnostic criteria to children since they have not yet achieved peak bone mass.
Many of the difficulties related to gender, ethnicity and children probably relate
to the effect of bone size on areal bone density measurements. Since most bone
density techniques give an areal measurement based upon a two-dimensional
projection of bone (in grams/cm2), larger bones will have a higher apparent bone
density than smaller bones due to the increased depth (Fig. 21). Although some
techniques have been developed to try to address this (such as calculations of skeletal
volume based upon modeling of the spine and femoral neck), only quantitative CT
(QCT) provides a true volumetric measurement.
120 Nuclear Medicine
Additional Reading
1. Blake GM, Wahner HW, Fogelman I. The Evaluation of Osteoporosis: Dual Energy
X-ray Absorptiometry and Ultrasound in Clinical Practice. London: Martin Dunitz
Ltd, 1999.
A detailed reference on the technology and clinical applications of bone density and
ultrasound.
2. Cummings SR, Nevitt MC, Browner WS, Stone K, Fox KM, Ensrud KE et al.
Risk factors for hip fracture in white women. New Engl J Med 1995; 332:767-773.
A true classic that describes the contribution of non-densitometric clinical risk factors
to hip fracture prediction in postmenopausal women.
3. Kanis JA. Gler CC. The Committee of Scientific Advisors, International
Osteoporosis Foundation. An update on the diagnosis and assessment of osteoporosis
with densitometry. Osteoporos Int 2000; 11:192-202.
A recent review of diagnostic criteria for osteoporosis from bone density measurements
with recommendations for change.
4. Kong YY, Penninger JM. Molecular control of bone remodeling and
6 osteoporosis.Exp Gerontol. 2000; 35(8):947-956.
A review of exciting new developments in the regulation of bone metabolism through
the osteoprotegerin system.
5. Marshall D, Johnell O, Wedel H. Meta-analysis of how well measures of bone
mineral density predict occurrence of osteoporotic fractures. BMJ 1996;
312:1254-1259.
A critical assessment of bone measurements and their ability to predict fractures.
6. National Osteoporosis Foundation. Physicians Guide to Prevention and Treatment
of Osteoporosis.
This clinical guideline is based upon a detailed report (Osteoporos Int 1998;8 Supplement
4) describing the evidence for the diagnosis, prevention, and treatment of osteoporosis in
postmenopausal healthy white women. The guideline can be viewed on-line at
www.nof.org/for_professionals/clinical/clinical.htm.
CHAPTER 1
CHAPTER 7
Skeletal Disorders
Leonard Rosenthall and Peter MacDonald
Introduction
The course of treatment for bone pain and skeletal abnormalities is effected by
an accurate diagnosis of the problem. Nuclear medicine has a role to play in both
the investigation of the problem and, in some cases, the treatment of bone and joint
disorders.
Skeletal Anatomy and Physiology
In the long bones of adults, blood primarily enters the diaphyseal cortex by
flowing outwards from the medulla rather than inwards from periosteal vessels. The
nutrient artery divides in the medullary cavity and anastomoses with the epiphyseal
and metaphyseal arteries which are direct branches of the regional systemic vessels
and enter through the numerous foramina penetrating the bones near their ends.
Blood flow from the epiphyseal and metaphyseal arteries is quantitatively greater
than that of the nutrient artery supply to the diaphysis. Periosteal arteries are part of
the network supplying the surrounding muscles. About two-thirds to three-fourths
of the inner cortex is sustained by the medullary arterial network. An obstructed
nutrient artery can result in a compensatory increase in centripetal periosteal flow to
support the full thickness of the cortex through interconnecting channels. Flat bones,
such as the cranium, and non-articular segments of short bones are supplied by
numerous periosteal vessels which nourish the cortex, cancellous bone and medulla.
The three main cells in bone are osteoclasts, osteoblasts and osteocytes. Osteoblasts
evolve from bone marrow-derived, pluripotent, stromal stem cells. Their role is to
form and mineralize bone matrix, and to synthesize skeletal growth factors. The
ultimate fate of the osteoblasts has not been clearly defined. Some are buried within
the bone matrix as osteocytes, while others become lining cells which cover quiescent
bone surfaces. Osteocytes are connected to each other and the lining cells through a
canalicular network that contains the bones extracellular fluid. They are believed to
release chemical messengers in response to physical strains in order to initiate the
appropriate modelling or remodelling response to mechanical stimuli. Osteoclasts
are derived from the hematopoietic precursors of the monocyte-macrophage lineage
and function in the bone resorption process. They are formed by the fusion of
mononuclear cells and are characterized as large multinucleated cells with a ruffled border.
Technical Considerations
Radiopharmaceuticals
Bone imaging is achieved with 99mTc-labelled phosphate and diphosphonate
complexes (collectively referred to as radiophosphate). The most commonly used
Figure 1. Normal whole body bone scan in the anterior (left) and posterior (right)
projections. (Case provided by Dr. W.D. Leslie.)
Figure 1. Normal whole body bone scan in the anterior (left) and posterior (right)
projections. (Case provided by Dr. W.D. Leslie.)
Figure 2. Fractures through the sacrum and both sacroiliac joint regions in an 82
year old osteoporotic women who fell from a standing position. The configuration
of the breaks is referred to as an H fracture.
simultaneously proceeds from the bony fracture surface, originating in the layers of
viable endosteum and periosteum. The new bone migrates centrally from both ends
and reinforces the ongoing ossification in the gap. The two processes form a solid
bony bridge, and when it is completed the fracture is considered clinically healed. In
the final phase of healing remodelling occurs as the woven bone of the callus is
converted to more adult lamellar bone, the external callus is smoothed by osteoclasts
and the medullary canal is reconstituted. Autoradiographic studies demonstrate that
Skeletal Disorders 125
Stage (duration)
Findings on three-phase Inflammatory Reparative (8 Remodeling
bone scan (3 to 4 weeks) to 12 weeks) (variable)
Delayed static phase broad and diffuse focal and intense diminishing
area of uptake at the uptake with
hyperconcentration fracture line time as healing
progresses
radiophosphate will deposit in the areas of new bone formation, with maximal up-
take where this process is predominant.
Three stages have been attributed to normal fracture healing (Table 1). During 7
the first 3 to 4 weeks after the break the site of trauma on the bone scan is characterized
by a relatively broad and diffuse area of hyperconcentration. In the subsequent phase
of 8 to 12 weeks duration the abnormal activity becomes more focal and intense
about the fracture line as mineralization of the osseous defects takes place. Thereafter,
the fracture exhibits a diminishing uptake with time as healing progresses. In the
three-phase radiophosphate study, the angiographic first phase is positive for about
the initial 3 weeks following the injury, whereas the blood pool phase can be positive
up to 10 weeks.
Radiophosphate imaging can disclose the presence of a fracture within 24 hours
of the injury in 95% of patients under the age of 65 years, but there may be a delay
in the elderly and debilitated patient. It is recommended that patients over 75 years
of age who have severe pain, but negative radiography and normal radiophosphate
imaging soon after injury, should have the bone scan repeated 72 hours later. CT
scan or plane tomograms may also aid in the diagnosis. The patient should be treated
as a hip fracture and kept non-weight bearing or on a bed-to-chair restriction during
this time. The negative predictive value for elderly patients suspected of hip fracture
who are imaged 72 hours after an earlier negative scan varies between 96% and 99%
(Fig. 3). The site of fracture is also a determinant of the degree of radiophosphate
accumulation. Those close to joints show the highest rate of uptake, whereas the
spine, pelvis and mid-shaft of the long bones have slower rates of uptake. This is
probably related to the regional blood flow.
About 90% of closed fractures that do not require surgical reduction will portray
normal radiophosphate uptake between 6 months and 2 years after injury. Intense
focal accumulation may persist longer in geriatric patients and in the presence of
compound and comminuted fractures. Orthopedic appliances can cause persistent
increased uptake, but it is usually low grade in the absence of infection or hardware
loosening. Persistent hardware uptake may be more common when plates and screws
are used versus an intramedullary rod system (the latter is load sharing versus the
former which is load bearing). Mechanical stresses at points of malalignment and
126 Nuclear Medicine
Figure 3. Left femoral neck fracture in a 92 year old woman who fell from a standing
position. (Left): In the bone scan 1 day after the injury there was a moderately
diffuse increased uptake in the hip region which could be attributed to regional
hyperemia. (Right): At day 4 there was definite evidence of a fracture (arrow).
7
angulation in weight bearing bones is another cause of prolonged enhanced
radiophosphate uptake. Less than 50% of the latter group will return to complete
quiescence within 3 years.
The bone scan can be helpful in dating vertebral fractures seen on the radiograph.
Old compression fractures are inactive by radiophosphate imaging, except when
complicated by degenerative disease, infection and tumor. Enhanced uptake in a
collapsed centrum suggests an injury of at most 2 years duration, but if the uptake is
intense the fracture probably occurred within a year.
Radiophosphate planar and SPECT imaging of the knee for the assessment of
meniscal and cruciate tears has received considerable attention, but where MRI
facilities exist there is little need for it today. Planar imaging can present a confusing
picture in the acute phase of internal knee injuries, because the major site of injury
is often hidden by diffuse uptake due to synovitis and generalized hyperemic reaction
to the hemarthrosis, trauma and stress reactions at musculo-ligamentous insertions.
SPECT imaging is able to unravel and define the various contributions to total
uptake, and thereby enhance the diagnostic yield. For injury to the meniscus, SPECT
sensitivity and specificity are reported to be 88% and 87%, respectively (Fig. 4).
Focal accumulation of radiophosphate at the medial femoral condyle peripherally
signals an avulsion of the medial collateral ligament, and pure anterior and posterior
cruciate ligament tears depict foci on the anterior and posterior tibial joint surface,
respectively. Cruciate tears may also be accompanied by focal intercondylar uptake.
Stress fractures
Repetitive episodes of inordinate minor stress in individuals engaged in
unaccustomed exercise induce an accelerated remodelling phase in the bones affected
by this strenuous activity. Bone remodelling consists of osteoclastic resorption which
reaches a peak bone loss at about 3 weeks. This is followed by a slower osteoblastic
filling of the osteoclastic cavities, reaching completion in about 90 days. As a
consequence there is an interval of imbalance when the bone is weakened and
Skeletal Disorders 127
7
128 Nuclear Medicine
Figure 5. Shin splints and stress fracture in a young male athlete. (Left): Posterior
bone scan showing a fusiform focus of increased activity in the left lower tibia
suggesting a stress fracture (arrow). (Right): Lateral view of the right tibia depicting
the linear increased cortical uptakes, more prominent posteriorly, which are
characteristic for shin splints.
uptake implies an absence of bone remodelling and inability to heal without surgi-
cal intervention.
Pseudoarthrosis
Pseudoarthrosis is the formation of a false joint cavity in the dense scar tissue
between the bone fragments. The hypertrophic type contains callus and a rich blood
supply, whereas the atrophic type does not have these features. Hypertrophic
pseudoarthroses show uptake of radiophosphate corresponding to the exuberant
sclerosis that borders the false joint, compared to a lack of accumulation in the
atrophic type. Difficulty may be encountered in discriminating between infected
and noninfected pseudoarthroses, because both conditions may depict substantial
uptake of 67Ga-citrate (gallium).
Osteomyelitis
In suspected acute osteomyelitis plain radiography should be the initial diagnostic
study. It usually takes 10 to 14 days after the onset of symptoms for the definitive
changes to develop. Three-phase radiophosphate bone imaging can disclose
uncomplicated acute osteomyelitis in adults within 3 days with a sensitivity and
specificity of about 95%. Septic arthritis alone and septic arthritis with extension
into adjacent bone can also be differentiated by three-phase radiophosphate
130 Nuclear Medicine
scintigraphy. Difficulties arise in the young pediatric group where the findings may
be falsely normal or present with a photon deficient (cold) defect. The latter finding
should be highly suspicious for osteomyelitis. Three-phase imaging becomes less
reliable when there is an underlying process that can induce enhanced bone
remodelling, such as healing fracture, pseudoarthrosis, neuropathic osteoarthropathy
and orthopedic appliances. 67Ga-citrate and 111In or 99mTc-labelled leukocytes are
used to determine whether osteomyelitis is superimposed on these conditions with
varying degrees of success.
Gallium localizes in infectious areas by granulocyte or bacterial uptake and binding
to the lactoferrin that accumulates in the region. It is disadvantaged by its tendency
for low to moderate concentration in non-infectious areas of enhanced bone
remodelling, e.g., pseudoarthrosis. Reliance in the diagnosis of osteomyelitis is placed
on an incongruence in the relative distributions of gallium and radiophosphate, or a
concentration of gallium that approaches or exceeds that of radiophosphate. 111In-
leukocytes, when available, are significantly better for the diagnosis of osteomyelitis
in the presence of complicating factors.
Chronic osteomyelitis is more difficult to diagnose because of its lower influx of
7 leukocytes and the deposition of 111In-leukocytes in areas of active marrow. Uptakes
are generally low, and there is a debate as to whether all uptakes are abnormal or a
threshold exists below which the accumulation is normal. In the central skeleton the
normal active marrow uptake reduces the sensitivity for chronic osteomyelitis.
Sensitivity and specificity are improved by combining 99mTc-colloid and 111In-
leukocytes. Any incongruity in bone marrow distribution of the two scanning agents
is considered significant (Fig. 6).
Diabetic foot
Osteomyelitis is a common complication in diabetes, occurring most frequently
in ischemic lower extremities. Its diagnosis can be impeded by an underlying
neuropathy with its attendant aseptic inflammation, bone destruction, abundant
periosteal and juxtaarticular bone proliferation and deformity. The radiographic
portrayal of acute progressive osteomyelitis and the aseptic neurotrophic foot can be
similar; both exhibit an avidity for radiophosphate, thereby confounding the diagnosis
of a superimposed focal osteomyelitis. The radionuclide diagnosis of osteomyelitis
in the diabetic foot requires a three-phase radiophosphate study followed by either a
gallium or radiolabelled leukocyte scan (preferably the latter because increased uptake
of gallium can occur in sterile inflammation and bone undergoing enhanced
remodelling). Absence or very low uptake of gallium excludes acute infection with
reasonable confidence, whether or not the radiophosphate bone scan is positive. If
the intensity of gallium uptake approaches that of radiophosphate and is congruent
with it in distribution, the presence of bone infection is suspected. A more widespread
distribution beyond the radiophosphate deposition is suggestive of cellulitis, but it
may mask an uptake of gallium in the infected bone, a condition that requires different
management. Similar criteria apply for the labelled leukocyte scans, except that the
degree of uptake in bone need not be as high as with gallium since it is not influenced
by enhanced bone remodelling and degenerative or destructive bone disease, and
considerably less by sterile inflammation. Deposition in soft tissues implies a cellulitis,
Skeletal Disorders 131
Figure 6. Left total knee arthroplasty. (Left): Radiophosphate image exhibits intense
periarticular uptake. (Middle): There are three areas of 99mTc-white blood cell
deposition (arrows). (Right): The 99mTc-colloid scan demonstrates uptake in the same
areas as the 99mTc-white blood cells, indicating active marrow and weighing against
infection.
but it may mask a co-existent osteomyelitis. Poor spatial resolution of 111In contrib- 7
utes to the false negative and false positive scan interpretations.
In comparisons between combined three-phase bone and 111In-leukocyte scan-
ning versus MRI, the latter was more sensitive in disclosing osteomyelitis in the
forefoot, but was less sensitive in the presence of a neuropathic (Charcot) joint. In a
study of 24 diabetic patients with foot ulcers, 13 had proven osteomyelitis. Sensitiv-
ity for clinical evaluation was 54%, 38% for plain radiography, 77% for combined
radiophosphate and 99mTc-leukocytes and 100% for MRI. Combined radionuclide
and MRI studies registered an 82% specificity.
Vascular Disorders
There are multiple causes of avascular necrosis (AVN), both traumatic and
nontraumatic. Fractures and dislocations can lead to AVN, the most common site
being the femoral epiphysis following femoral neck fractures. Nontraumatic causes
of epiphyseal AVN and intramedullary infarcts are thromboemboli to bone in
disorders such as hemoglobinopathies, sickle cell disease (Fig. 7), sickle cell trait and
sickle cell thalassemia. Decompression states (Caisson disease) that result from deep
sea diving and high-altitude flying or ballooning can lead to AVN due to nitrogen
emboli. Pancreatitis and alcoholism can be associated with fat necrosis and fat emboli
that may lead to AVN. Systemic lupus erythematosis, polyarteritis nodosa and giant
cell arteritis are connective tissue disorders involving small vessels which are
consequently susceptible to occlusion and AVN of dependent tissue. Gauchers disease,
Cushings disease and steroid therapy are associated with swollen fatty marrow cells
which may compromise vascularity and cause AVN.
Spontaneous AVN commonly occurs in the medial condyle of the femur, but it
may develop in the lateral condyle, tibial plateau and elsewhere. Its etiology is
unknown and usually middle aged men and women are affected, in contrast to
osteochondritis dissecans which is typically found in adolescent years. Other bones
that can be afflicted with AVN are the humeral epiphysis, metacarpal and metatarsal
132 Nuclear Medicine
Figure 7. Bone infarct in a patient with sickle cell disease. Acute pain developed in
the right thigh 2 days before the initial radiophosphate bone scan on April 25 (Left).
It showed a segment of photopenia in the right femoral shaft relative to the activity
in the bone above and below it (arrows). (Right) Four months later, August 20,
there is increased radiophosphate deposition in the area reflecting bone repair.
heads, lunate, capitellum, tarsal navicular and the dome of the talus. Osteonecrosis
can be a complication of frostbite (Fig. 8) and high voltage burns.
In AVN imaging with radiophosphate, a sequence of events are predicted, begin-
ning with an initial photopenia that corresponds to the avascular focus and followed
in time by an intense circumferential uptake of the reactive reparative phase. Upon
completion of revascularization and healing, the uptake returns to normal. Because
of the limitations of spatial resolution, the photopenia is not commonly seen in
small bones, but may be detected in the femoral capital epiphyses and about the
knees. To demonstrate this photopenia, which is characteristic of AVN, imaging
must take place early in the evolution of bone necrosis with high resolution SPECT
or pinhole collimation. In a comparison of imaging modalities, there were 14 patients
with proven AVN but normal plain radiographs, in whom 54% had positive CT
scans, 71% had characteristic radiophosphate bone scans and 86% had positive
MRIs. In 55 patients with various stages of AVN, positive results were obtained in
78% of the plain radiographs, 87% of the radiophosphate scans, 96% of the CTs
and 96% of the MRIs.
Where MRI is readily available it has largely replaced radiophosphate bone
imaging in the initial work-up of suspected AVN or infarction because it is more
sensitive and specific. AVN can be asymptomatic and be detected incidentally in
bone scans ordered for other reasons. Some investigators suggest both MRI and
Skeletal Disorders 133
Figure 8. Frostbite. This patient sustained frostbite osteonecrosis of the hands and
feet as a result of being stranded overnight in the mountains in sub-zero temperatures.
Radiophosphate bone scans show complete absence of blood flow to the hands
and feet, which were subsequently amputated. Arrows point to the demarcation 7
between viable and nonviable bone at the wrist and ankle.
will abate and may even exhibit activity levels below normal. In intermediate stages
the diagnosis rests with the finding of increased uptake in the delayed images, and
this has been found to be the most reliable sign for diagnosing and monitoring the
disease. Sensitivity and specificity of the delayed images in the hand are 96% and
97%, respectively. In the foot sensitivity approaches 100%, but specificity is only
80% owing to complicating factors such as diabetes mellitus, infection, etc. Positive
and negative predictive values are 54% and 100%, respectively. It has also been
found that patients with higher uptakes in the delayed phase have, on average, a
better response to therapy. The scan pattern in the pediatric age group may differ
from adults in that all three phases may be decreased in RSD.
Joint Prostheses
Pain is the most common presenting symptom in disorders of prosthetic joints
which include implant wear debris, component loosening, fractures of the component,
bone fracture and infection. Imaging with a combination of radiophosphate, gallium,
111 In or 99m Tc-labelled leukocytes, radiolabelled polyclonal antibodies or
Skeletal Disorders 135
7 Figure 10. Radionuclide arthrography. (Left): Loose right cemented femoral hip
prosthesis depicting increased focal uptake of radiophosphate at the stem tip and
in the intertrochanteric area. The intense deposit superior to the prosthesis neck is
due to heterotopic bone formation (arrow). (Right): Combined 99mTc-colloid and
radiocontrast injection into the joint cavity for athrography. 99mTc scan shows
radioactivity seeping into the space surrounding the femoral component indicative
of loosening (arrows).
are related to the hip, and there is general acknowledgement that radionuclide ar-
thrography is more sensitive for the femoral component but unreliable for the
acetabular component (Fig. 10).
Infection
Three-phase bone imaging is very sensitive for the disclosure of infected
arthroplasties but it is not specific. In a group of 98 patients presenting with painful
arthroplasties and suspected of having an infection, a sensitivity of 100% and
specificity of only 18% was confirmed by surgical documentation. A number of test
agents can be used to increase the specificity. Gallium was the first to be introduced.
If the distribution of gallium is incongruent with that of the radiophosphate, or if it
is intense and congruent, the likelihood of infection is high. Low grade congruent
uptakes are equivocal and since this is reported to occur in about half the cases, the
use of gallium is limited.
The results of labelled leukocytes are better, but vary with the intensity of the
inflammatory process. Normally the labelled leukocytes, unlike gallium, do not
accumulate at sites of increased bone turnover in the absence of infection. Sensitivity
was found to be poorest in chronic osteomyelitis, and this was attributed to the low
leukocyte attraction to the site. Specificity was compromised in chronic osteomyelitis
by the occasional low grade uptake from the aseptic inflammatory reaction associated
with healing fractures or by uptake in active marrow adjacent to the hip and knee
prostheses. With regard to the latter, it has been shown that in asymptomatic patients
Skeletal Disorders 137
with hip prostheses, 111In-leukocyte activity was present in 48% of the femoral tips
at 24 months, and 37% had significant uptakes in the region of the acetabular
component. The addition of a 99mTc-colloid scan will aid in identifying these areas
of disturbed marrow function adjacent to hip and knee implants so that their
concentration of labelled leukocytes will not be misinterpreted as infection (Fig. 6).
Results from using combined leukocyte/colloid marrow imaging are reported to
vary in accuracy from 89% to 98%, with the improvement being largely due to
increased specificity.
Radionuclide Synovectomy
Treatment of chronic synovitis by surgical synovectomy is not always successful
as recurrences occur with regeneration of the synovium and there may be an associated
prolonged rehabilitation due to joint stiffness and limitation of motion. In surgical
management of hemophilic synovitis and hemarthrosis the problem of maintaining
hemostasis can be daunting and expensive. Intraarticular injection of chemicals such
as nitrogen mustard, thiotepa and osmic acid, although less invasive, were not
consistently successful. Repeated intraarticular injections of corticosteroids pose a
risk for systemic toxicity. 7
Although modern arthroscopic synovectomy has made radionuclide synovectomy
less popular, it remains a valuable therapeutic option. When a beta emitting particulate
radiopharmaceutical is injected into the inflamed joint, the synovium is exposed to
a high radiation dose and atrophies. Desirable attributes for the agent include little
or no leakage from the joint cavity, high affinity binding of the beta emitter, uniform
distribution throughout the joint cavity, synovial uptake without initiating an
inflammatory reaction and a biological half-life within the joint that should not be
less than the physical half-life of the radionuclide. Leakage from the joint cavity via
the lymphatics can lead to deposition in the regional inguinal nodes. Leakage has
also been attributed to synovial inflammation and joint movement; therefore
premedication with intraarticular glucocorticoids (to reduce the synovial hyperemia)
and bed rest have been advocated to decrease the frequency and amount of
radiocolloid deposition in the regional nodes. Many suitable radiopharmaceuticals
have been developed (Table 2). The most frequently used are 32P-chromic phosphate
and 90Y-citrate which have pure beta emission with energy for good tisue penetration,
large particle sizes and are commercially available.
Ideally, the penetration of beta particles should be limited to the thickness of the
synovium to avoid radionecrosis of the cartilage and bone, but this may only be a
theoretical concern as cartilage is relatively resistant to radiation and no cases of
necrosis have been reported. Chromosomal abnormalities have been seen, but in all
the years of experience with intraarticular radiation synovectomy no instance of
induced malignancy have been reported.
Treatment Procedure
For the knee, an injection of a local anesthetic is obtained under aseptic conditions,
and through a lateral approach an 18 to 20 gauge needle is inserted into the joint
cavity. Smaller joints such as the wrist and elbow, may require x-ray fluoroscopic
control, perhaps with contrast injection to ensure absence of leakage and loculation.
At the knee, the ease in aspiration usually ensures proper needle placement. The
138 Nuclear Medicine
dose of 90Y-citrate instilled is 278 MBq (7.5 mCi) into the knee joint, 92.5 MBq
7 (2.5 mCi) into elbow joint and 74 MBq (2 mCi) into the wrist. Hydrocortisone is
generally instilled with the treatment dose to mollify a potential reactive radiation
synovitis i.e., increased pain and effusion, which may occur between a few hours
and 2 weeks after treatment. In order to assess placement and dispersion of the 90Y-
citrate, 37 to 74 MBq (1 to 2 mCi) of 99mTc-colloid is often included to permit
scanning (Figs. 11 and 12). The joint is bandaged snugly and the patient is sent
home with instructions to refrain from weight bearing and to rest the joint for 3
days and not undertake strenuous activities for at least 2 weeks to minimize leakage.
In rheumatoid joints, good results, seen as various degrees of pain relief, reduction
in joint effusion and increased range of motion, were obtained in 50%-70% of the
patients at 6 month to 12 month evaluations. This was independent of the
radiopharmaceutical used. Hemophilic hemarthroses demonstrated a reduction in
the frequency of hemorrhage in 80%-90% of the patients. Patients can be retreated
if there is reactivation of symptoms.
Frequently Asked Questions (FAQs)
Can the bone scan differentiate septic arthritis from sterile
synovitis (such as rheumatoid arthritis and gout)?
The application of radiophosphate and 67Ga-citrate to differentiate septic and
sterile rheumatoid synovitis has not been consistently successful. Joint-to-bone ratios
of gallium uptake tend to be higher for septic joints but the overlap with sterile
joints is quite large. Intense uptake suggests sepsis and a low concentration in an
untreated joint favors sterility. The use of radiolabelled leukocytes has also been
reported to yield false positive results for the presence of sepsis due to the variable
amounts leukocytes in sterile rheumatoid and gouty effusions. Leukocytes are also
present in the sterile effusions of osteoarthritis and may contribute to false positive
Skeletal Disorders 139
Figure 12. Radionuclide synovectomy. 99mTc-colloid scans of the knee following its
intracavitary injection at the same time as 32P-chromic phosphate for therapy. No
evidence of loculation at 30 minutes and 24 hours, except for a small amount
which entered a popliteal cyst (arrow).
leukocytes, a pattern that may also reflect chronic osteomyelitis. Predictably, a nega-
tive bone scan is highly predictive of a negative bacterial culture. The final arbiter of
cure is the clinical evaluation.
Additional Reading
1. Holder LE (guest editor). Orthopedic Nuclear Medicine (Part 1). Seminars in
Nuclear Medicine 1997; 27(4).
2. Holder LE (guest editor). Orthopedic Nuclear Medicine (Part 2). Seminars in
Nuclear Medicine 1998; 28(1).
A detailed review of the basic physiology, imaging techniques and clinical applications
of nuclear medicine as applied to orthopedics.
CHAPTER 1
CHAPTER 8
Skeletal Oncology
Leonard Rosenthall and Ralph Wong
Introduction
Radioisotopes play an important role in the diagnosis of benign and malignant
bone lesions. Bone scanning is commonly used to detect metastases from malignant
tumors because it is more sensitive than x-rays and allows the whole body to be
surveyed. The general principles of skeletal imaging are reviewed in Chapter 7. This
Chapter will concentrate on aspects specific to bone tumors.
Primary Benign Bone tumors
Benign and malignant primary bone tumors can be quite difficult to distinguish
with radiophosphate imaging. In general, benign lesions tend to have lower accretions
than malignant lesions, but in the individual case it is neither reliable nor predictive
of the histopathological outcome.
Adamantinoma
Adamantinoma is a rare tumor that occurs most commonly in the mid-diaphysis
of the tibia. In rare cases it can behave like a malignant tumor with metastases to the
lung. It appears as a large radiolucent lesion on the radiograph, and shows moderate
uptake of radiophosphate.
Aneurysmal Bone Cyst
Aneurysmal bone cyst (ABC) is a distinctive non-neoplastic solitary lesion of
bone consisting of a cystic cavity filled with nonendothelial-lined spaces containing
blood. Radiographically, there is a striking blow-out distention of part of the contour
of the affected bone, and it may simulate a malignant lesion. More than half of these
lesions occur in the metaphyseal regions of the long bones or vertebrae, and up to
one-third are linked to benign or malignant processes such as nonossifying fibroma,
chondroblastoma and giant cell tumor of bone. Due to their location and potentially
rapid, aggressive growth, they can be difficult to distinguish from a primary
malignancy. Despite the presence of blood filled cavities, the perfusion segment of
the three-phase radiophosphate bone study may be negative. More than half of the
delayed images depict a doughnut pattern of uptake (central photopenia encompassed
by a band of activity).
Bone island
A bone island is a common lesion that consists of a nest of compacted trabeculae
of mature lamellar bone. It may enlarge after puberty, but it does not have malignant
potential. Delayed bone scans are usually normal but may show a mild increased
uptake (Fig. 1).
Chondroblastoma
Chondroblastoma is an uncommon primary bone tumor of chondrogenic origin
that is seen in younger individuals before epiphyseal closure. It usually has its origin
from the region of the growth plate, most commonly in the proximal humerus, but
it can arise in flat bones. They are composed of dense concentrations of rounded or
polyhedral chondroblasts. Histological features are similar to chondromyxoid fibroma
and the two may be related. Malignant transformation rarely can occur.
Radiophosphate concentration in these lesions can be intense.
Desmoplastic Fibroma
Desmoplastic fibroma is an extremely rare benign intraosseous fibrous tumor
occurring mostly in the long bones. Three quarters of the cases are seen below the
age of 30 years, with peak incidence in the second decade of life. The tumor is
usually located centrally in the metaphysis where it symmetrically expands the bone
and thins the cortex. Radiographically it appears as an osteolytic lesion with well
circumscribed margins. Nonspecific increased uptake of radiophosphate can be seen.
Enchondroma
Enchondroma is a common benign lesion that appears in the medullary portion
of bone and is composed chiefly of mature hyaline cartilage. The solitary 8
enchondroma has a predilection for the phalanges and metacarpals of the hand. It
can also occur in large long and flat bones (pelvic and shoulder girdle) where they
have a potential for malignant transformation, and may be difficult to distinguish
from a low grade chondrosarcoma. The bone scan may demonstrate normal or mildly
increased uptake of radiophosphate (Fig. 2). A lesion which changes from normal to
high uptake over time, especially when associated with pain, is suspect for malignant
conversion.
Multiple enchondromas, or enchondromatosis, is an anomaly usually disclosed
in infancy. When one side of the body exhibits greater involvement than the other,
the condition is called Olliers disease. The combination of enchondromatosis and
soft tissue hemangiomata is referred to as Maffuccis syndrome. About 8% of the
cases of Maffuccis syndrome degenerate into chondrosarcomas.
Epidermoid Cyst
Epidermoid cysts are sharply delineated, squamous cell lined and filled with
desquamated keratinized debris. They occur in the calvarium, but traumatic
epidermoid cysts may develop in the phalanges. The scintigraphic presentation in
the calvarium is usually normal, but if the lesion is large enough a photopenic focus
may be appreciated.
Fibrous Defects
Fibrous defects are focal collections of nonspecific fibrous tissue located
predominantly in the cortical bone, but they may also be found in cancellous bone.
Sites of predilection are the long bones, primarily the femur and tibia. Several
classifications have been proffered for these fibrous lesions, but a simplified version
divides them into fibrous cortical defects and nonossifying fibromas. The former
occurs in about 30% of children over 2 years of age, and radiographically it presents
144 Nuclear Medicine
Figure 2A. Benign enchondroma. Anterior (left) and lateral (right) radiographs of
the left femur showing an area of mottled calcification in the distal femoral shaft
with slight expansion of the cortex. Differential diagnosis includes benign
enchondroma, central chondrosarcoma and infarct.
Skeletal Oncology 145
Figure 3. Giant cell tumor of bone. This 26 year old woman complained of increasing
right knee pain following minor trauma associated with progressive swelling and
inability to weight bear. Radiographs show a large lytic lesion in the lateral condyle
of the tibia (white arrow). A three-phase bone scan shows marked hyperemia
corresponding to the lytic lesion (black arrow) but reduced tracer accumulation centrally
(arrowheads). This appearance is seen with hypervascular destructive bone tumors. (Case
provided by Dr. W.D. Leslie.)
148 Nuclear Medicine
8
Figure 4 Osteoid osteoma. The arterial flow study is normal but there is focal
hyperemia in the left femoral head on the immediate blood pool image and intense
uptake on delayed images (arrows). (Case provided by Dr. W.D. Leslie.)
scan is usually normal, but there are isolated reports of increased uptake of
radiophosphate.
Unicameral or Simple Bone Cyst
A bone cyst is an intramedullary cavity lined by fibrous tissue and is usually filled
with a clear amber fluid, but it may be multi-chambered. The proximal metaphyseal
ends of the humerus and femur are the most frequent sites of occurrence. The
condition is seen in children and adolescents and thought to undergo involution as
it is much less common in adults. Radiographically, the appearance is that of a large
radiolucent lesion, broad at the metaphyseal end and narrow at the shaft end, and
does not cross the epiphyseal plate. Lower extremity cysts are more likely to fracture
than those in the upper extremity. Bone scans are usually negative, but a pattern of
a slight increase in radiophosphate uptake at the margins of the cyst and photopenia
in the center may be seen. A fracture will induce an avid concentration of the scanning agent.
Primary Malignant Bone tumors
Radiophosphate scanning is of limited use in the initial diagnosis of primary
malignant bone tumors, as these lesions are detected by radiography when the patient
first presents with local symptoms. Computed tomography (CT) and MRI are more
accurate in defining the extent of the tumor, both in bone and soft tissue. There is a
role for scintigraphy in the search for occult bone metastases, as may occur with
osteosarcoma and Ewings sarcoma.
150 Nuclear Medicine
Figure 5. Osteoid osteoma. Posterior planar image of the lumbar spine depicts
subtle increased uptake in the 5th lumbar vertebra (arrow). Coronal SPECT images
clearly define and localize the lesion.
Osteosarcoma
Osteosarcoma is a high grade tumor that accounts for about 20% of all primary
bone malignancies, and generally occurs during the second to third decades of life
with another peak in the sixth decade. Those that arise later in life are associated
with pre-existing conditions such as Pagets disease and previous bone irradiation. It
is the most common primary malignant bone tumour with the exception of multiple
myeloma. Most frequent sites of origin are the metaphyses of the long bones, and it
is more common in males. Diseases that are predisposed to the development of
osteosarcoma include osteochondromatosis (Olliers disease), Maffuccis syndrome,
multiple hereditary exostosis, osteogensis imperfecta and Pagets disease. The primary
lesions have abnormally high uptake of radiophosphate, but are best characterized
and defined by CT and MRI. Bone scintigraphy is a convenient modality for detecting
osseous and soft tissue metastases and, thereby, aids in patient management.
Osteosarcoma arising in Pagets disease carries a poor prognosis. Typically the
tumors present as large destructive unresectable lesions in the pelvis, proximal femur
and humerus. They usually present as a relatively photopenic area surrounded by
the high uptake of radiophosphate in the pagetic bone. In patients undergoing
chemotherapy, the decrease of radiophosphate uptake in the tumor has been shown
to parallel the degree of induced tumor necrosis. Thallium-201 and 99mTc-sestamibi
are also reported to render a good assessment of tumor viability and response to 8
chemotherapy.
Chondrosarcoma
Chondrosarcoma is the second most common primary malignant tumor of bone
with a peak incidence during the fifth decade of life. Typically, it is located in the
pelvis, femur and shoulder girdle. It is of cartilaginous origin and appears on the
radiograph as a radiolucent expansile lesion. It may destroy the cortex and invade
the adjacent soft tissue. It is best defined by CT and MRI. Avid uptake of
radiophosphate can be seen in both the intramedullary and exostotic forms of the
malignancy.
Ewings Sarcoma
Ewings sarcoma is the second most common primary malignant bone tumor of
children and adolescents, and most occur between 10 and 20 years of age. It is
thought to develop from a primary mesenchymal cell although the actual histogenesis
has not been conclusively determined. It is commonly seen in the femur and pelvis.
Unlike osteogenic sarcoma, it has a predilection for the flat bones. Plain radiography
followed by MRI to define tumor extent are the main imaging modalities. The
radiophosphate bone images can depict intense activity on all three phases of the
study, but it is not specific or sufficiently detailed for management, and is therefore
relegated mostly to the role of metastatic tumor search. Metastatic bone disease is
present in 10% of patients on the initial clinical examination, and about 50% of
patients eventually develop bone metastases. Some authors have recommended 99mTc-
colloid marrow imaging to disclose marrow involvement, but it is unreliable after
radiation and chemotherapy.
152 Nuclear Medicine
Multiple Myeloma
Multiple myeloma is a plasma cell dyscrasia of the bone marrow. It can be mistaken
radiographically for metastatic bone disease because it presents with multiple osteolytic
lesions or diffuse osteopenia. The other two elements of the triad are plasmacytosis
of the bone marrow (>10%) and a monocloncal gammopathy (in serum and/or
urine). Skeletal radiography is superior to bone scintigraphy in the evaluation of the
presence and extent of the disease. On a site-by-site comparison radiography is
reported to have a sensitivity of 82% relative to 42% by bone scanning. On a patient
basis radiography and scintigraphy have sensitivities of 94% and 78%, respectively.
Radiophosphate imaging often fails to demonstrate a hyperconcentration at sites of
skeletal involvement by myeloma despite extensive disease by radiography. This occurs
because reactive bone formation is not a feature of the disease; this might be due to
release of an inactivation factor by the tumor. The bone scan may be helpful when
there is focal pain and a negative radiograph.
Histiocytosis X or Langerhans Cell Histiocytosis
Histiocytosis X comprises a spectrum of conditions from an isolated benign bone
involvement (eosinophilic granuloma) to a benign dissemination (Hand-Schuller-
Christian disease) to a highly malignant form (Letterer-Siwe disease). The disease
8 can involve bone, soft tissue, or both. Bone scintigraphy is positive in many of the
lesions (Fig. 6), but it is not reliable as a screening procedure. Plain radiography and
CT are the recommended imaging methods for the initial assessment. Bone
scintigraphy can be useful in searching for metastases or investigating pain in the
context of a normal radiograph.
Diagnosis and Follow-Up of Skeletal Metastases
Radiophosphate bone scanning in metastatic disease is recognized for its high
sensitivity and low specificity. As a consequence of parallel developments of competing
diagnostic methods such as CT, MRI, ultrasonography and blood tests for tumor
associated antigens, the role of bone scanning has changed. Initially it was
recommended as a screening procedure for bone metastases in all newly diagnosed
malignancies, but recent evaluations of its overall efficacy with different tumor types
have qualified this routine use.
Pathophysiology of Skeletal Metastasis
The metastatic cascade begins with the release of a single cell, or a clump of cells,
as a result of proteolysis induced by tumor enzymes. These cells enter the vascular or
lymphatic systems where most are destroyed in transit by the cells of the immune
system. Those few potent cells that escape the initial defense and reach a distant site
require local conditions that promote implantation and growth. The cells adhere to
receptors on the vascular endothelium or basement membrane, and then pass through
the vessel wall to establish themselves in the organ. Local destruction is mediated by
growth factors elaborated both by tumor cells and the invaded tissue. This provides
space for tumor growth, and a concomitant stimulation of angiogenesis allows the
tumor to meet its nutritional requirements. The majority of tumor deposits occur in
the red marrow and this has been attributed to the slow transit through the marrow
sinusoids. A requirement for tumor growth is local osteoclastic bone resorption.
Tumor cell stimulation of osteoclastic activity is mediated by its release of such factors
Skeletal Oncology 153
Figure 7. Skeletal metastases from prostatic carcinoma. Anterior and posterior total
body scans show multiple focal sites of intense uptake involving both the axial and
appendicular skeleton. (Case provided by Dr. W.D. Leslie.)
varying contrast settings, many will reveal a heterogenous rather than homogeneous
distribution. With successful treatment the bone scan can revert to normal or a
presentation of multiple focal abnormalities and the usual soft tissue background
and visible kidneys.
Skeletal Oncology 155
Figure 8. Diffuse prostatic metastases producing a super scan. Anterior (left) and
posterior (right) total body scans show intense uniform concentration in the vertebral
column, pelvis and sternum. The usual soft tissue background and kidneys are not
visible by comparison. Enhanced uptake in the skull and ribs is also present, but in
this illustration is overwhelmed by the spine and pelvis.
Photopenic defects are an important sign of tumor presence and should be con-
sidered with the same gravity as focal increases in activity. They are probably due to
vascular compromise as a result of tumor compression of the small marrow vessels.
Since photopenic or cold lesions are harder to resolve than the typical hot
lesion, those cold lesions that are detected are relatively large (Fig. 9).
156 Nuclear Medicine
Figure 9. Photon deficient metastasis. This patient with carcinoma of the prostate
had a solitary asymptomatic lesion (arrow) in the thoracic spine at staging. Six
months later he developed upper back pain and the bone scan was normal except
for a photopenic focus in a thoracic vertebra (arrow), which corresponded to the
site of pain. A scan four months later revealed widespread metastatic disease.
8
Lung Carcinoma
Bone scans have been used to screen patients for metastases in newly diagnosed
non-small cell lung cancer as part of the staging process. With the advent of CT and
MRI, evidence of metastatic deposits in the hilar and mediastinal lymph nodes, liver
and adrenals can be identified in order to establish non-resectability of the primary.
The additional information gleaned from bone scan screening will not alter the
staging or management even though it has been reported that the bone scan may be
positive in about 40% of patients on initial presentation.
Most small cell lung carcinomas will have metastasized by the time the patient is
first seen and these are readily revealed by physical examination, laboratory findings
and radiographic techniques. Resectability is not an issue in this tumor category,
and although it is reported that about 50% will show bone lesions by scintigraphy it
has little influence on management in the initial phase.
Bone scanning is generally relegated to monitoring chemotherapy, assessing bone
pain for possible local radiation therapy or orthopedic intervention to preclude
fracture in the long bones or spinal cord compression, disclosing hypertrophic
pulmonary osteoarthropathy as a cause of lower extremity pain (Fig. 10) or identifying
nonmalignant causes of pain such as trauma, infection and osteonecrosis.
Breast Carcinoma
Present day evaluations of large clinical trials indicate that the frequency of bone
metastases in stage I and stage II breast cancer patients is less than 10%. Most
investigators are of the opinion that the use of bone scintigraphy for staging is
unnecessary in stage I and most stage II individuals in view of the newer diagnostic
radiographic methods that are available and the results of cost-effectiveness analyses.
Bone scans have a role if the primary tumor is large or histologically aggressive, if
Skeletal Oncology 157
Diagnostic Clues
Multiple fractures History of trauma and plain radiographs usually
suffice. Occasionally trauma may be minimal or
absent (severe osteoporosis) in which case a follow-up
scan should show resolution. In infants child abuse
should be questioned.
Figure 11. Pagets disease of bone (osteitis deformans). Multifocal areas of increased
uptake are scattered throughout the skeleton (arrows). Although some of these would be
difficult to distinguish from metastatic lesions, the appearance of the tibia (diffuse
involvement with bowing deformity) is characteristic for Pagets disease. The skull lesion
shows peripheral uptake and central photpenia, a form of pagetic involvement known as
osteoporosis circumscripta. (Case provided by Dr. W.D. Leslie.)
Figure 12. Metastatic flare reponse. (A) This 48 year old woman had breast cancer
metastatic to liver (left) and bone (right). She was treated with a conventional
doxorubicin-containing chemotherapeutic regimen and had follow-up bone/liver
scans four (B) and eight months (C) later. There is dramatic improvement on both of
the follow-up liver scans. In contrast, the bone scan shows transient worsening
followed by later improvement. There had not been any alteration in the patients
therapy, and the improvement in the liver metastases shows that the tumor was
sensitive to the chemotherapeutic regimen. (Case provided by Dr. W.D. Leslie.)
162 Nuclear Medicine
sign of tumor regression. Continued increase in the number and intensity of the
lesions beyond 6 months after treatment indicates loss of control.
How long should you wait before repeating a bone scan?
That depends on the reason for the scan. Occasionally the diagnosis of skeletal
metastases is still in doubt after a bone scan and conventional radiographs have been
completed. If a follow-up examination shows worsening, then metastatic disease is
very likely, whereas resolution (in the absence of anti-tumor treatment) confirms a
benign process. Although some high-grade tumors can progress rapidly, usually 3-6
months are required to detect a change. On the other hand, if the scan is being used
to evaluate response to treatment then at least 6 months should elapse to avoid
confusion with the flare phenomenon (see previous question).
Additional Reading
1. Jacobson AF. Primary and metastatic bone disease. In: Sandler MP, Coleman RE,
Wackers FJT, Patton JA, Gottschalk A, Hoffer PB, eds. Diagnostic Nuclear
Medicine. Baltimore: Williams & Wilkins, 1996:649-667.
2. Charron M, Brown ML. Bone scanning in metastatic disease. In: Collier BD, Jr.,
Fogelman I, Rosenthall L, eds. Skeletal Nuclear Medicine. St. Louis: Mosby,
1996:87-123.
8
CHAPTER 1
CHAPTER 9
Kidney
Michael Hoskinson and Keevin Bernstein
Introduction
Nuclear medicine techniques for evaluating the kidney and urinary tract have
evolved from pioneering studies with primitive radiopharmaceuticals and hand-held
probe detectors to a sophisticated technology capable of providing important
physiologic and anatomic information. The advent of the gamma camera placed
emphasis on morphological evaluation of the kidneys. However, with recent advances
in other anatomic imaging modalities, renal scintigraphy has rediscovered its roots
in the evaluation of function. This can be appreciated in the discussion of radionuclide
techniques that measure purely physiological processes such as glomerular filtration
rate (GFR). Other sections deal with the important contributions of nuclear renal
imaging in the evaluation of acute renal failure, renovascular hypertension, urinary
tract obstruction and renal transplants.
Renal physiology
The renal parenchyma receives approximately 20% of cardiac output. In the
normal person, 20% of renal plasma flow is filtered so that 180 litres of a protein-
poor ultrafiltrate is produced each day. This occurs across highly permeable glomeruli
which present a large surface area (1.6 m2). The ultrafiltrate, which includes
nitrogenous waste products and salts, traverses a complex tubular system where tubular
secretion, reabsorption and metabolism of specific solutes occurs. The proximal
nephron (proximal convoluted tubule) reabsorbs 55-70% of the salt and water
presented to it. Fine-tuning of salt and water reabsorption occurs in the distal nephron
(collecting tubules and ducts) so that less than 1% of the filtered load is ultimately
excreted. Other solutes, such as potassium and hydrogen, undergo reabsorption and
secretion at various sites along the nephron. The final urine constituents pass from
the collecting ducts into the renal pelvis and calyces.
Under non-stressed physiologic situations, renal perfusion is autoregulated by
smooth muscle in the afferent arteriole (myogenic reflex) to maintain a constant
perfusion pressure and GFR over a wide range of systemic blood pressures. Under
situations of compromised renal perfusion, there is balanced afferent and efferent
arteriolar constriction mediated predominantly by sympathetic stimulation and
angiotension II which helps to maintain glomerular filtration.
In the context of renal function, clearance can be conceptualized as the volume
of plasma cleared of a material by the kidneys per unit time (usually per minute).
If a substance is filtered at the glomerulus and neither secreted nor reabsorbed, its
clearance is a measure of GFR. 99mTc-DTPA is one such material. In a normal adult
patient, the GFR is approximately 120 ml/min. If a substance is completely cleared
from the plasma as it passes through the kidney by any combination of filtration
Nuclear Medicine, edited by William D. Leslie and I. David Greenberg.
2003 Landes Bioscience.
164 Nuclear Medicine
Figure 1. Diagram of the nephron showing GFR agents such as 99mTc-DTPA, 51Cr-
EDTA and 125I-iothalamate filtered at the glomerular level. While 99mTc-MAG3 and
123
I- or 131I-hippurate are filtered, their predominant method of renal excretion is by
tubular secretion at the level of the proximal convoluted tubule. 99mTc-DMSA and
99m
Tc-glucoheptonate become fixed within the renal tubules. 99mTc-glucoheptonate
is also filtered and secreted by the nephron. (Figure provided Ms. L.Ward.)
Kidney 165
and secretion, its clearance is a measure of effective renal plasma flow (ERPF). ERPF
is that plasma which flows through the kidney and is available for filtration and/or
secretion, excluding capsular and interstitial perfusion.
Technical Considerations
Radiopharmaceuticals
Several tracers are available for renal imaging and functional assessment (Fig. 1).
A brief description of some of the more common ones is presented below.
99m Tc-Diethylenetriaminepentaacetic Acid (99mTc-DTPA)
This tracer is a nearly ideal agent for assessment of GFR. It is a small chelate with
a molecular weight of 500, and its only mode of excretion is glomerular filtration. It
is neither secreted nor reabsorbed by the kidneys. Its renal extraction reflects the
filtration fraction of plasma (i.e., 20% first pass clearance). GFR determination using
99mTc-DTPA should take into account the possibility of binding of 99mTc-DTPA to
Figure 2. Coronal tomographic (SPECT) slices from a 99mTc-DMSA study. The slices
go from posterior to anterior through the kidney. Note that the display orientation
for SPECT is the same as for CT: the right side of the patient is on the left side of the
image. The defect in the upper pole of the left kidney (arrow) is due to either acute
pyelonephritis or a scar.
Non-Imaging Agents
125
I-iothalamate is used as a tracer for GFR assessment. The energy of the photons
emitted by iodine-125 is too low for imaging, and the 60 day half-life of this isotope
requires a very small injected dose. GFR assessment involves plasma sampling and
counting the radioactivity in a well counter.
51
Cr-EDTA is a chelate similar to 99mTc-DTPA and is likewise filtered by the
glomeruli. The compound is stable and the isotope long-lived. This agent is extensively
used in Europe but has limited availability in North America.
Kidney 167
Figure 3. Normal 99mTc-DPTA scan. In the early part of the flow phase at 3 seconds
per frame (A), the kidneys appear at the same time as the aorta. The left kidney is as
intense or more intense than the spleen. In the 1 minute per frame dynamic images
(B), activity is seen in the collecting systems and ureters by 2-3 minutes. Note that
the bladder is not in the field of view in most of the dynamic images. Better
positioning (bottom row on right: pre and post void images) allows visualization of
the bladder. Note the image of the injection site (bottom row, third from right)
showing no extravasation of tracer.
Procedure
The most important consideration in preparing a patient for renal scintigraphy
is hydration (specifics are dealt with in the individual sections). For GFR
quantification, it is suggested that the patient avoid eating a heavy meal for 4 hours
prior to the study as a heavy protein load can increase GFR. It is also important to
review the patients recent medication use and hemodynamic status, particularly if
pharmacological interventions (e.g., captopril stimulation) are anticipated.
Typically, the gamma camera is placed posterior to the patient. (Keep in mind
that the images are displayed from the cameras perspectivein a posterior acquisition,
the patients right side is on the right side of the image, just as if the eyes of the
168 Nuclear Medicine
Figure 4. 99mTc-MAG3 flow (A) and sequential images (B). These show much better
contrast between kidney and background than 99mTc-DTPA for two reasons: the
higher first pass extraction efficiency of 99mTc-MAG3 and the fact that 99mTc-MAG3
is highly protein bound, so very little of it diffuses into the extravascular extracellular
space, leaving a higher concentration available for excretion. Note that while the
study was acquired at 1 minute per frame, the images are displayed as sequential 2
minute images. The time activity curves on the bottom reflect the normal rapid
increase (uptake phase) and subsequent decrease (excretion phase) in renal tracer
9 concentration seen with 99mTc-MAG3.
observer were behind the patient). Transplanted kidneys are more anterior than native
kidneys, so the study is acquired with the camera anterior to the abdomen and pelvis.
As with any nuclear medicine procedure, renal scanning involves injecting a
radioactive tracer and watching where it goes. Renal perfusion can be assessed with
the various 99mTc-labeled tracers with the camera set to take rapid sequence images
at a rate of 1-3 seconds per frame during the first minute of the acquisition. The
remaining 20-30 minutes of the study is usually performed at a framing rate of 20-
60 seconds to evaluate the uptake and excretion phases (Figs. 3 and 4). A 30-second
image of the injection site is highly recommended at the end of the study. If a
substantial portion of the injection is interstitial instead of intravenous, this can
have a great effect on the renogram, with delayed excretion and a stretched-out
curve. Interstitial injections invalidate efforts to quantify renal function.
Quantification of Renal Function
Computer Analysis of Renal Scans
By placing appropriate regions of interest (ROIs) over the kidneys, various
parameters of renal function can be derived. These include differential renal function
measurements, renogram curves generated from ROIs placed around the entire kidney
(Fig. 5), renogram curves generated from ROIs placed over the cortex of the kidney
(Fig. 6), and pelvic emptying curves (Fig. 7).
Kidney 169
Figure 5. Renogram curves and differential renal function. Selected images from
the flow and clearance phases are shown in (A). Five regions of interest have been
drawn on an early image: left and right kidneys, background ROIs placed
inferolateral to the two kidneys and an ROI placed over the aorta. These are used
to derive a variety of quantitative measures (B) including each kidneys contribution
to total renal function expressed as a percentage of the sum of the background
corrected counts from the kidney ROIs. The calculation of differential renal function
(%DIFF) is performed on data from one of the early images, before any activity has
reached the collecting system. Flow (C) and renogram (D) curves are shown. The
latter are time activity curves of the net counts within the previously drawn renal
ROIs applied to the series of thirty 1 minute images and corrected for the background.
(Correction for non-renal background activity is accomplished by subtracting [counts
per pixel in background] x [number of pixels in kidney] from total counts in the
kidney ROI.) The time to peak activity (Tpeak or Tmax) can be read off the graph
(4 minutes in this example). This is the time when uptake in the kidney equals
excretion. The rate of parenchymal clearance is expressed as the half-time (T 1/2)
and the ratio of 20 minute to peak activity (20/MAX).
170 Nuclear Medicine
to measuring GFR (the Gates method) involves placing ROIs over both kidneys
and calculating the uptake in the kidneys as a percentage of the injected dose (after
correcting for scatter and soft tissue attenuation). A regression curve is used to relate
this percentage to GFR.
Clinical Role in Acute Renal failure
Clinical
A large number of disorders (Table 1) can precipitate acute renal failure (ARF)
and a detailed discussion is beyond the scope of this handbook. Pre-renal causes
predominate, followed by renal and post-renal causes. A systematic approach to
differential diagnosis and management is required since complete recovery is possible
in many cases.
Usually the differential diagnosis can be narrowed following a thorough clinical
assessment and the following limited investigations: urinary electrolytes (to calculate
fractional excretion of sodium as a marker of pre-renal azotemia), urinalysis (to look
for active sediment from glomerular or interstitial inflammation), and renal
ultrasonography (to exclude obstructive hydronephrosis). Biopsy is reserved for
patients in whom pre-renal and post-renal failure have been excluded and the cause
of intrinsic renal azotemia is unclear. Renal biopsy is particularly useful when clinical
Kidney 173
INTRINSIC RENAL
I. Renovascular obstruction (bilateral or unilateral with one functioning kidney)
Renal artery obstruction: thrombosis, embolism, dissection, vasculitis
Renal vein obstruction: thrombosis, compression
II. Disease of glomeruli or renal microvasculature
Glomerulonephritis and vasculitis
Hemolytic uremic syndrome, thrombotic thrombocytopenic purpura,
scleroderma
III. Acute tubular necrosis
Ischemia (as for prerenal azotemia) 9
Toxins: radiocontrast, cyclosporine, aminoglycosides, cisplatin,
ethylene glycol, acetaminophen, endogenous (rhabdomyolysis,
hemolysis, uric acid, oxalate)
IV. Interstitial nephritis
Allergic: antibiotics, nonsteroidal anti-inflammatory agents, diuretics,
captopril
Infection: acute pyelonephritis, viral (e.g., CMV), fungal
Infiltration: lymphoma, leukemia, sarcoidosis
Idiopathic
V. Intratubular deposition and obstruction
Myeloma proteins, uric acid, oxalate, acyclovir, methotrexate,
sulphonamides
POSTRENAL (OBSTRUCTION)
I. Ureter
Calculi, blood clot, sloughed papillae, cqncer, retroperitoneal
fibrosis
II. Bladder neck
Neurogenic bladder, prostatic hypertrophy, calculi, cancer, blood
clot
III. Urethra
Stricture, congenital valve, phimosis
from other forms of acute renal failure by intense renal uptake of gallium 48 hours
or more after injection (Fig. 9). Non-inflammatory ATN typically shows little or no
gallium accumulation. Unfortunately, gallium uptake is not a specific finding and is
seen in a variety of inflammatory (lupus and pyelonephritis) and non-inflammatory
states (proteinuric nephropathy and amyloidosis).
Renal infarction
Autopsy studies suggest that embolic renal infarction is under-diagnosed
antemortem. Flank pain may be a prominent symptom following occlusion of a
renal artery or vein, but clinical and laboratory evidence of renal infarction may be
completely lacking. Scintigraphy can help to differentiate renal infarction from renal
parenchymal disease (Fig. 10).
Clinical Role in Hydronephrosis
Clinical
Not all dilated renal collecting systems are obstructed (Fig. 11). The collecting
system can remain dilated indefinitely after a previous episode of obstruction is
relieved. Vesicoureteral reflux causes dilatation, and megaureter can be congenital.
Relative obstruction can occur at high outflow states (Fig. 12). The term obstructive
uropathy refers to dilatation of the renal pelvis and calyces caused by obstruction.
Obstructive nephropathy refers to functional impairment of the kidneys caused by
obstruction. Obstructive nephropathy can be as severe as absence of function in the
Kidney 175
Figure 10. Renal infarct. 99mTc-DTPA flow and early static images (A) show absent
flow and function in the lower pole of the left kidney (black arrows) in this patient
with recent onset of flank pain. The CT scan (B) shows lack of contrast enhancement
in the left kidney cortex (white arrowhead). A selective left renal angiogram (C,
shown in the same orientation as the renal scan) shows an aneurysm arising from
the left renal artery which is occluded distally (white arrow).
Kidney 177
Figure 12. Flow-related obstruction. Sequential 1 minute images (A) and pelvic
emptying curves (B) show rapid emptying from the right renal pelvis. On the left,
there is an initial good response to furosemide but the clearance curve reaches a
plateau (arrow) indicating flow-related obstruction (Homesys sign).
Figure 13. Recovery of renal function post pyeloplasty. Early 99mTc-MAG3 static
images on a 6 month old infant before (A) and 10 months after a right pyeloplasty
(B). There is minimal tracer accumulation in the kidney before the procedure and
normal uptake following relief of obstruction.
study. The downslope of the curve from the moment of first action of furosemide to
the end of the study is fitted to an exponential function from which is derived the 9
excretion half-time. An unobstructed system will show rapid emptying and a sharp
drop in the counts within the regions of interest (Fig. 11). The normal curve will
have a concave slope, with a half-time of 10 minutes or less. An obstructed system
will show a flat or even rising curve, or a half time of greater than 20 minutes (Fig.
14). In between these easy to recognize extremes are the intermediate curves that
require further consideration.
Factors that make interpretation difficult are renal impairment (in which the
kidney is incapable of increasing urine output sufficiently to wash out hot urine
from the renal collecting system) and large volume renal collecting systems (since
the rate of washout will be inversely proportional to the volume). If function is
preserved (single kidney GFR of 15 ml/min or greater) in the kidney with a
suboptimal response, then an intermediate pattern likely indicates partial obstruction.
If the single kidney GFR is less than 15 ml/min, then the study is truly indeterminate.
With large volume collecting systems, the F-15 approach may be helpful in a repeat
study. If the kidney has normal function, then even partial obstruction may not be
functionally significant, and a conservative approach may be warranted.
Clinical Role in Renovascular hypertension
Clinical
Renovascular disease is an important correctable cause of secondary hypertension.
While it accounts for less than 1% of mild hypertension, the incidence may be as
high as 10-45% in patients with severe hypertension. It is also an increasingly
recognized and potentially reversible cause of advanced renal failure. It may be
180 Nuclear Medicine
Figure 14. Obstructive hydronephrosis. The first phase (A) shows an enlarged right
kidney (on the right hand side of the posterior image) with poor initial cortical
uptake and slow filling of a dilated pelvis. The second phase (B) was then started
with injection of furosemide 40 mg intravenously. Note that there is negligible
clearance from the right renal pelvis visually and on the time activity curve (C). In
contrast, the left renal pelvis shows normal drainage even before the diuretic
injection.
9
Figure 15. Principle of ACE inhibition (captopril) renography. In a normal patient
(A), blood is delivered to the glomerulus via the afferent arteriole and leaves by the
efferent arteriole. Normally, there is a sufficiently high glomerular pressure to allow
for renal physiology (represented by arrows) across the capillary membrane and
Bowmans capsule. In a patient with renal artery stenosis and renovascular
hypertension (B), the initial drop in blood pressure in the glomerulus is counteracted
with angiotensin II-mediated vasoconstriction of the efferent arteriole (arrow).
Glomerular pressure and hence renal physiology is maintained. If captopril is
administered (C), the efferent arteriole relaxes with a subsequent drop in glomerular
pressure and filtration. This decrease in renal physiology can be detected with
renal scanning. (Figure provided by Dr. I.D. Greenberg.)
Table 9.3 Change in tracer handling after angiotensin converting enzyme (ACE)
inhibitor challenge for affected kidneys in renovascular hypertension
Uptake Transit
99m
Tc-DTPA Decreased Prolonged
99m
Tc-MAG3 No change Prolonged
unilateral disease in patients with GFR > 30 ml/min. An unstimulated renal scan
has a false negative rate of 20- 25%. The predictive value of scanning can be increased
by performing a captopril-stimulated renal scan in the patient with normal or
minimally impaired renal function (GFR > 30 ml/min) who is not on an ACE
(angiotensin converting enzyme) inhibitor or AT1 (angiotensin type 1) receptor
antagonist. In properly selected patients, the sensitivity and specificity of high
probability findings approach 90%. A normal study will show symmetric uptake
and prompt excretion on both the baseline and captopril studies and implies a low
(<10%) probability of renovascular hypertension.
184 Nuclear Medicine
Figure 18. Bilateral symmetrical captopril induced changes. Testing was performed
in this young woman with uncontrolled hypertension after interruption of ACE
inhibitor therapy for 72 hours. The baseline 99mTc-MAG3 study (A) is normal.
Following captopril (B), there is marked parenchymal retention symmetrically in
the kidneys. This is a non-specific response that is attributed to a modest drop in
the patients blood pressure (from 150/90 to 135/75). (Case provided by Dr. W.D.
Leslie.)
Kidney 185
Figure 19. 99mTc-DPTA scan in a normal renal transplant. Both the flow (A) and 1
minute sequential images (B) are normal.
Pathophysiology
Renal artery stenosis causes a drop in pressure in the glomerulus. This would be
expected to result in a loss of filtration were it not for compensatory mechanisms
instigated by the juxtaglomerular apparatus (JGA). In response to inadequate pressure
within the afferent arteriole of the glomerulus, the JGA produces renin which converts
angiotensinogen to angiotensin I. Angiotensin I is subsequently converted to
angiotensin II, a potent vasconstrictor. Angiotensin II causes constriction of the
efferent arteriole and results in increased pressure in the glomerulus and restoration
of glomerular filtration. Angiotensin II also stimulates aldosterone which potentiates
sodium retention and potassium secretion in the distal nephron. Both angiotensin
II and aldosterone contribute to an elevated blood pressure. ACE inhibitors are
pharmacologic agents which block the formation of angiotensin II, interrupting the
compensatory mechanism of efferent arteriolar vasoconstriction thus lowering GFR
in the affected kidney. This serves as the basis for ACE inhibition renography (Fig.
186 Nuclear Medicine
Figure 20. 99mTc-DTPA scan in a renal transplant with moderate acute tubular necrosis
(ATN). Perfusion (A) is normal, but on the sequential images (B) uptake in the kidney
is impaired and there is poor contrast between kidney and background.
15). Captopril is short acting, and is given orally for ACE inhibition renography. If
a parenteral agent is required, enalaprilat can be used.
Procedure
Hydration is very important to proper performance of this study. Patients should
be hydrated to the same extent as for diuretic renography. Captopril and especially
enalaprilat can cause hypotension in dehydrated patients and, therefore, blood
pressure should be monitored after administration of the ACE inhibitor. If
hypotension occurs, the patient should be placed supine with legs elevated and
rehydrated with an intravenous saline infusion. Patients for whom hypotension would
be dangerous, such as those with a history of stroke, transient ischemic attacks,
carotid stenosis or coronary artery disease, should have an intravenous line in place
before the ACE inhibitor is given. The patient should not leave the department
Kidney 187
until blood pressure has stabilized. There are some drugs which may reduce the
sensitivity of ACE inhibitor renography. In our laboratory we discontinue (if possible)
ACE inhibitors and AT1 receptor antagonists for 3 days prior to the examination.
Captopril is given orally and its absorption is reduced in the presence of food.
Therefore, patients should not eat a solid meal for four hours before the study,
though they should drink fluids freely. Captopril 25-50 mg is crushed to enhance
absorption and is administered 60 minutes before the tracer. Alternatively, enalaprilat
40 g/kg to a maximum of 2.5 mg is given intravenously over 3-5 minutes, 15
minutes before the tracer. The use of enalaprilat instead of captopril avoids potential
problems with poor gastric absorption, but is associated with a higher incidence of
hypotension.
Captopril renograms can often be performed without a baseline examination for
comparison because a normal captopril challenge study is sufficient to rule out
renovascular hypertension. A baseline study is useful in equivocal cases for detecting
subtle captopril-induced transit abnormalities. The baseline and captopril studies
can be performed on separate days or on the same day. In the latter situation, a
larger dose is given for the captopril study e.g., 50 MBq of 99mTc-MAG3 for the
baseline study and 250 MBq for the post captopril study. The same-day procedure
should be considered for patients who have had to discontinue interfering
medications, where a return visit would be difficult, and for in-patients since clinical
and hydration status can change rapidly from one day to the next.
Processing and Interpretation 9
If the ACE inhibitor causes a drop in GFR of the involved kidney, the scan
findings will depend on the tracer used (Table 3). Since GFR is reduced, 99mTc-
DTPA uptake (a marker of GFR) will be reduced. Indeed, in severe cases, GFR may
completely cease. In this case, blood pool activity will be seen in the affected kidney
and, as tracer is excreted by the opposite kidney, the blood pool activity will decrease
at the same rate as blood pool in other organs such as the liver and spleen. A reduced
GFR leads to slowing in bulk urine flow through the kidney and any 99mTc-DTPA
entering the nephron will likewise traverse it slowly i.e., transit through the kidney
is prolonged (Fig. 16).
When 99mTc-MAG3 is used, its initial uptake is unchanged, because even if there
is a substantial drop in filtration 99mTc-MAG3 is still delivered to the proximal
convoluted tubule and excreted into the lumen. Because bulk urine flow is reduced,
transit through the kidney is prolonged (as with 99mTc-DTPA). This can be assessed
by visual inspection of the images or analyzing renogram curve parameters such as
time to peak and 20 min/max ratio (Fig. 17). A normal time to peak with 99mTc-
MAG3 is between 2 and 4 minutes. A prolongation of at least 2-3 minutes from the
baseline to captopril study, or 40% of the baseline value, is considered significant.
The normal 20 min/max ratio is less than 0.30. An increase in this ratio of 0.15
from the baseline study suggests a high likelihood of renovascular hypertension
Bilateral symmetrical changes are usually not due to renovascular hypertension
(Fig. 18). Even if renal artery stenosis affects both kidneys, findings will be asymmetric
in the vast majority of cases. Symmetrical changes are more likely to be caused by
hypotension, dehydration or medications such as calcium channel blockers. A small
poorly functioning kidney (< 30% uptake) may not show any change with ACE
188 Nuclear Medicine
Figure 21. 99mTc-MAG3 scan in a normal renal transplant. Perfusion (A) is normal.
There is excellent uptake and rapid excretion seen in both the sequential images
(B) and the renogram curve (C).
Figure 22. 99mTc-MAG3 scan in a renal transplant with moderate acute tubular
necrosis (ATN). Perfusion (A) is normal. On the sequential images (B), initial uptake
is good and there is excretion by 3-4 minutes. Parenchymal retention in the kidney
is seen at the end of the study. The persistent activity can be appreciated on the
renogram curve (C).
190 Nuclear Medicine
Figure 23. 99mTc-MAG3 scan in patient with severe acute tubular necrosis (ATN).
Perfusion (A) is mildly imparied. The sequential images (B) show good uptake but
excretion is greatly delayed and there is progressive accumulation of tracer in the
kidney throughout the study. Ureteric obstruction is unlikely, because some activity
does reach the bladder (though many transplant recipients have some residual
function in their native kidneys which might be responsible for the bladder activity).
If there is concern for obstruction, an ultrasound is usually helpful. The renogram
curve (C) shows increasing activity in the transplant.
Kidney 191
Figure 24. Black hole sign. 99mTc-DTPA 1 minute sequential images of a left iliac
fossa renal transplant 2 weeks post-op. This 44 year old woman initially had
moderate acute tubular necrosis (ATN) after a cadaveric transplant. She developed
acute rejection 1 week post-op and despite therapy progressed to renal infarction.
The black hole sign (arrow) is almost always indicative of a non-viable kidney.
Figure 26. 99mTc-MAG3 scan obtained 4 hours post injection and showing gut activity.
The transplant (arrowhead) is seen in the left iliac fossa. Irregular activity seen in
the mid and lower abdomen on the right represents biliary excretion of the tracer
into the gut and is usually present to some degree on delayed 99mTc-MAG3 views.
Occasionally it can be so intense as to mimic a urine leak.
collecting system activity, and ultrasound must be relied upon to rule out ureteric
obstruction. The most severe cases of harvesting injury may show poor perfusion,
little uptake and high background levels even with 99mTc-MAG3.
Follow up studies in uncomplicated harvesting injury show progressive
improvement over days to weeks. Beware of the case that shows improvement during
the first week followed by a drop in perfusion or function. This may signal an
additional complication such as rejection or cyclosporine nephrotoxicity.
Other complications in the early post-operative period include vascular
catastrophe, hematoma, and urine leak. Renal artery or vein thrombosis presents as
little or no perfusion or function, and may show the black hole sign, in which the
transplant can be seen as a negative image with less activity than non-renal background
(Fig. 24). This implies infarction of the graft. Hematoma presents as a photopenic
area surrounding or adjacent to the graft. Hematoma can cause pressure on the
kidney, impairing function directly or by causing ureteric obstruction. Urine leak
194 Nuclear Medicine
Figure 27. Renal trauma. 99mTc-glucoheptonate flow (A) and sequential images (B)
on a young adult male injured in a car accident. There is very little flow and function
in the right kidney. The large photopenic area inferior to the functioning upper
pole represents infarcted kidney, hematoma and a small urinoma. The same patient
was rescanned subsequently. Delayed images (C) show a normal left kidney, a
right urine leak and a small amount of functioning right kidney.
often arises from necrosis of the ureter because of its tenuous blood supply. In the
early post-op period, the patient is usually catheterized, so intense urine activity
outside of the bladder is easy to detect (Fig. 25). If the patient is not catheterized, it
is important to get an adequate post void view to assess for leak. Delayed views at 3-
4 hours after injection can be helpful, but be aware that some 99mTc-MAG3 is excreted
Kidney 195
by the liver, so that there is normally some activity in the gut on these delayed views
(Fig. 26).
Frequently Asked Questions (FAQs)
Should renal scanning be used in the assessment of renal
trauma?
The imaging investigation of choice for renal trauma is computed tomography
(CT). CT allows investigation of the whole abdomen, whereas renography evaluates
only the kidneys. If renography is required, it may be best done with 99mTc-
glucoheptonate which allows assessment of renal perfusion, function, cortical integrity
and urine leaks (Fig. 27). Hemorrhage can also be detected, especially during the
blood flow phase of the study. Complete absence of activity, such as in the black
hole sign mentioned above, indicates devascularization of the kidney.
Additional Reading
1. Conway JJ, Maizels M. The well tempered diuretic renogram: A standard method
to examine the asymptomatic neonate with hydronephrosis or
hydroureteronephrosis. J Nucl Med 1992; 33(11):2047-2051.
A report from combined meetings of The Society for Fetal Urology and members of The
Pediatric Nuclear Medicine Council, Society of Nuclear Medicine.
2. Dubovsky EV, Russel CD, Bischof-Delaloye A et al. Report of the radionuclides in
nephrourology committee for evaluation of transplanted kidney (review of
techniques). Semin Nucl Med 1999; 29:175-188. 9
A review of the use of radionuclide techniques in the assessment of renal transplants.
3. Fine EJ. Interventions in renal scintirenography. Semin Nucl Med 1999;
29:128-145.
This article, as evident from the title, concentrates on diuretic and ACE inhibitor
renography and touches upon other possible interventions (aspirin, exercise).
4. Tayor A. Radionuclide renography: A personal approach. Semin Nucl Med 1999;
29:102-127.
A nice review with emphasis on diuretic and ACE inhibitor renography.
CHAPTER 10
Gastrointestinal
Peter Hollett and Ford Bursey
Introduction
A wide variety of nuclear medicine examinations are available for studying the
function of the gastrointestinal tract. Tests include assessment of salivary gland func-
tion; motility studies of the esophagus, stomach, small and large bowel; measuring
substrate absorption; localization of gastrointestinal bleeding; detection of Meckels
diverticulum (covered in Chapter 17) and tests for inflammatory bowel disease (cov-
ered in Chapter 12). This Chapter is not meant to be exhaustive but rather will
concentrate on some of the more important and commonly requested tests.
Clinical Role in Esophageal Motility Disorders
The esophagus must transport swallowed material from the mouth to the stom-
ach while minimizing reflux of stomach contents. Symptoms associated with im-
paired function are dysphagia, pain or regurgitation. Motility disorders can be primary
or secondary (Table 1) and result from smooth muscle disorders (e.g., scleroderma)
or disorganization of the intrinsic nervous system (e.g., achalasia, Chagas disease,
and diabetes). The identification of these disorders can be aided by nuclear esoph-
ageal motility studies though classification is still conventionally described in rela-
tion to esophageal manometry:
Normal: After a swallow, pressure in the upper esophageal sphincter (UES) and
lower esophageal sphincter (LES) falls. A contraction wave starts in the pharynx and
progresses down the esophagus.
Scleroderma: The lower part of the esophagus (smooth muscle) shows a reduced
amplitude of contractions with hypotension of the LES.
Achalasia: The LES is usually hypertensive and fails to relax in response to a
swallow. The body of the esophagus shows contractions that are reduced in ampli-
tude and simultaneous in onset.
Diffuse esophageal spasm (DES): The lower part of the esophagus shows simulta-
neous-onset, large-amplitude, prolonged, repetitive contractions.
Hypercontractile esophagus (nutcracker esophagus): The major dysfunction is ab-
normal increase of contraction amplitude and normal propagation through the
esophagus.
Non-specific motility disorder: Abnormalities that do not conform to any of the
preceding patterns.
Technical Considerations
Most laboratories perform esophageal motility assessments with a liquid bolus,
although some groups have claimed that semisolid materials are more sensitive. If
10
Figure 1. Normal esophageal study. To assess esophageal transit, ROIs are drawn
over the proximal, middle and distal thirds of the esophagus avoiding the stomach
(arrow) and hypopharynx (E). Time activity curves (TACs) are generated for the
three esophageal segments. The transit or emptying times for the individual seg-
ments are calculated as the time from peak to 10% of peak (A,B,C). The RF4 is the
residual fraction remaining in the esohagus after 4 swallows and is calculated from
a TAC generated from a ROI placed around the entire esophagus (D).
Gastrointestinal 199
for both liquids and solids. Often dysphagia is non-progressive and intermittent.
Occasionally reflux induced esophageal spasm will present with dysphagia or atypi-
cal chest pain. Many patients may have non-specific disorders of motility related to
unrecognized or untreated gastroesophageal reflux, though occasionally classic ab-
normalities such as achalasia and nutcracker esophagus are detected. Abnormal esoph-
ageal motility can also be seen with aging (presbyesophagus). Esophageal motility
disorders are seen in many systemic diseases including systemic sclerosis (scleroderma),
Parkinsons disease and diabetes.
Investigation of a patient with dysphagia begins with a careful history and physi-
cal examination. The physical examination should include an evaluation of the neck,
mouth, oropharynx, larynx, and neurologic system. Supplementary testing with
endoscopy and/or barium videoesophagography is usually required. The results from
these investigations should indicate those with a mechanical or inflammatory etiol-
ogy. Esophageal manometry should be arranged for those in whom no mechanical
or inflammatory etiology is detected and in whom dysphagia persists despite therapy
(e.g., dilatation and acid suppression for benign stricture from GERD).
Radionuclide transit studies are less sensitive and specific than manometry and
cannot indicate which specific motility disorder is responsible for the abnormal transit.
As a result the routine use of esophageal transit studies is not recommended in the
initial investigation of dysphagia. Direct comparison of radionuclide transit mea-
surements and videoesophagography indicates similar sensitivity (68-74%) and speci-
ficity (68-70%) for specific esophageal disorders, though both are less accurate than
manometry in distinguishing nonspecific esophageal motility disorders from nor-
mal. Radionuclide transit measurement is a cheap, noninvasive and rapid supple-
mentary examination which, when positive, can support further invasive studies of 10
the esophagus in patients with unexplained chest pain.
Clinical Role in Gastric motility Disorders
Gastroparesis, defined as delayed gastric emptying because of abnormal gastric
motility in the absence of mechanical outlet obstruction, is a common problem
causing significant morbidity. Less commonly, accelerated gastric emptying can pro-
duce symptoms known as dumping syndrome. The current gold standard for quan-
tifying gastric emptying is nuclear scintigraphy.
Physiology
The stomach has two distinct physiologic motor areas: the proximal stomach
and the distal stomach. The proximal stomach, with its slow, sustained contractions,
plays a key role in regulating intragastric pressure and emptying of liquids, while the
distal stomach, with its peristaltic contractions, has a major role in mixing, grinding
and emptying of solids. Gastric emptying is highly dependent on the meals volume,
caloric density, composition and concentration of fatty acids, carbohydrates and
proteins. Receptors in the small bowel sense these elements and can release hor-
mones or activate neural pathways that can greatly influence the rate of emptying.
The emptying of solid food from the stomach consists of three distinct phases.
First, there is an initial lag phase during which little or no solid food is evacuated
from the stomach into the duodenum as the meal is ground into small pieces (tritu-
ration), mixed and distributed by the muscular contractions of the stomach. During
200 Nuclear Medicine
the second phase there is a constant rate of emptying and then a much slower rate
during the final third phase as the stomach approaches empty. Liquids do not need
to be triturated and therefore emptying of liquids is more rapid than for solids and
tends to follow an exponential rate of emptying. (Liquid phase studies are now
rarely performed due to the wide variation in the normal handling of liquids within
the stomach and lower sensitivity when compared to the solid phase examination.)
Technical Considerations
A wide variety of drugs can affect gastric emptying and therefore most medica-
tions should be discontinued, if possible, prior to a scintigraphic evaluation (Table
2). Evaluation is best done after a 12 hour fast. Diabetic patients should be studied
in the morning 20-30 minutes after their normal insulin dosage. Smoking delays
gastric emptying and should be avoided.
There are many variations on the testing procedure but the basic principles re-
main the same. Solid phase studies typically use 99mTc-sulfur colloid cooked with an
egg preparation (though some centers use other protein sources). The prepared egg
product is then administered orally, often with a small carbohydrate load (for ex-
ample, with toasted bread as an egg sandwich). Image acquisition begins soon after
the meal is eaten. The stomach lies obliquely within the abdomen and as food moves
from the more posterior fundus to the more anterior antrum there is an apparent
increase in anterior counts due to lesser depth (attenuation). For accurate quantitation,
images in both the anterior and posterior projections are obtained with averaging
(geometric mean). Recent work suggests that the use of a single LAO projection also
minimizes geometric effects.
10 Data acquisition is performed for 60 to 120 minutes using either continuous or
intermittent imaging. In cases where there is prolonged retention of material within
the stomach even more delayed images up to 3 hours may be required. Continuous
imaging permits complete visualization of the lag phase when food redistributes
from fundus to pylorus. Imaging can be performed in the semi-upright, sitting or
standing position. It is very important to establish normal ranges for the technique
employed in each individual laboratory, as the results are highly dependent on the
acquisition parameters and the test meal used.
The most commonly derived quantitative parameters for gastric emptying are
the lag phase, fractional emptying and the emptying half time (Fig. 2).
Clinical
Gastroparesis can be seen as an idiopathic disorder or secondary to many condi-
tions (Table 2). The autonomic neuropathy of long-standing diabetes mellitus is the
most common identifiable cause and develops in about 20-30% of poorly con-
trolled patients (Fig. 3). Neither the type of diabetes nor the age of onset has been
shown to predict the risk of developing diabetic gastroparesis.
Symptoms that evoke the possibility of gastroparesis include early satiety, post-
prandial fullness, halitosis and frequent belching. The sensitivity and specificity of
these symptoms is poor as they also occur in the setting of non-ulcer dyspepsia and
mechanical gastric outlet obstruction. It is important to rule out a mechanical cause,
such as partial gastric outlet obstruction secondary to chronic peptic ulcer disease,
with endoscopy or a barium contrast series.
Gastrointestinal 201
Rapid gastric emptying (dumping) may occur after gastric surgery as a result of
altered vagal activity or in the Zollinger-Ellison syndrome as a result of acid
hypersecretion. Hyperthyroidism may also increase the rapidity of gastric emptying.
202 Nuclear Medicine
10
10
Figure 3. Abnormal gastric emptying study in a 43 year old woman with diabetes
and suspected gastroparesis. (A) The bulk of activity remains in the stomach at the
end of the study. (B) The amount having left the stomach by 60 minutes and 120
minutes is minimal.
Separating liquids in the diet from solids and ingesting the solids first may help to
minimize symptoms.
Clinical Role in the Localization
204 Nuclear Medicine
of Gastrointestinal bleeding
The rapid localization of potentially life threatening GI hemorrhage is an urgent
medical and surgical problem. The history and physical examination, while neces-
sary, rarely reveal an exact cause of bleeding and further testing is usually required.
Nuclear medicine scanning may help in the evaluation of both acute and chronic
gastrointestinal bleeding.
Technical Considerations
Most centers use a labeled red blood cell technique for visualizing GI bleeds. A
variety of red blood cell labeling technique techniques have been described (dis-
cussed in Chapter 4). It is critical that minimal free 99mTc-pertechnetate be present
in the preparation since pertechnetate, which accumulates in the gastric wall and is
secreted into the lumen, can cause a false positive examination. To minimize the
amount of free pertechnetate, the in vitro labeling technique is preferred.
Dynamic imaging is performed for at least one hour or until a site of active GI
bleeding is clearly identified. Many centers employ further delayed imaging if no
active GI bleeding is encountered. One point cannot be overstated: it is essential to
acquire dynamic images. GI transit can be rapid, particularly in the small bowel,
and intermittent imaging can lead to erroneous localization of the bleeding. Review
of the cine images greatly enhances detection of a bleeding site and following the
direction of activity as it traverses the bowel further improves localization. For in-
stance, a site of bleeding may be identified in the left upper quadrant of the abdo-
men. If activity descends along the left flank, the bleeding site is likely to be within
the region of the splenic flexure. If activity moves inferiorly and to the right, the
10 bleeding site is probably within the jejunum. Blood in the colon will generally move
antegradely but occasionally it can move in a retrograde fashion (Fig. 4).
In patients who have had multiple negative diagnostic procedures, some centers
use intravenous heparin to unmask the bleeding site. This should only be under-
taken after less risky approaches have failed, when bleeding is recurrent and life
threatening, and with preparation to administer blood products or perform other
resuscitation as required.
Clinical
Gastrointestinal bleeding is a common clinical problem and a major cause of
morbidity and mortality. An urgent assessment of hemodynamic stability is the foun-
dation of the clinical approach to such patients. Obtaining venous access and ensur-
ing that oxygenation and tissue perfusion is adequate are essential first steps. At the
same time as supportive measures are taking place, an attempt should be made to
determine the source of the bleeding and whether the bleeding is ongoing. A com-
plete blood count and coagulation profile should be checked and blood cross-matched
in the event that transfusion is required.
In most patients it is clinically apparent if the bleeding is originating from the
upper or lower gastrointestinal tract, though occasionally this presents a diagnostic
challenge. Endoscopy will determine the source of bleeding in almost all who present
with hematemesis and most who present with melena. Passage of a nasogastric tube
may help in the evaluation of GI bleeding but a negative gastric aspirate does not
exclude an upper GI source. Acid peptic disorders account for up to 70 percent of
Gastrointestinal 205
Figure 4. 99mTc-RBC study showing a bleeding site in the colon. This 82 year old
woman with a history of a recent sigmoid resection for carcinoma of the colon
presented with three bouts of dark red bleeding into the colostomy bag. Sequential 10
images displayed at 3 minutes/frame show active bleeding originating from the
transverse colon near the splenic flexure (arrow) with activity moving in retrograde
fashion into the transverse colon (arrowhead) and in antegrade direction into the
colostomy bag (dual arrowhead). Surgery confirmed a bleeding diverticulum within
the splenic flexure.
of suspected lower GI bleeding. Nuclear medicine scanning can detect smaller rates
10 of bleeding than angiography, is less invasive and may help to focus the efforts of the
angiographer. As angiography offers the potential of therapeutic intervention, the
decision as to whether a GI bleeding scan is performed in conjunction with angiog-
raphy or not is dictated by the urgency of the clinical situation (Fig. 5).
Clinical Role of Urea breath testing
The discovery of the causal relationship between the gram-negative spiral rod,
Helicobacter pylori, and peptic ulcer disease (PUD) has dramatically changed the
investigation and approach to therapy. While only a small number of patients that
harbor H. pylori develop PUD, most patients with PUD are infected with H. pylori.
Although the diagnosis of H. pylori infection can be established by histological
examination of gastric biopsy specimens, urea breath testing is non-invasive, inex-
pensive and highly accurate (over 95% sensitivity and specificity).
Technical Considerations
Urea breath testing relies upon the fact that the urease enzyme is not present in
mammalian cells and H. pylori is the only bacteria commonly found in the stomach
that contains the enzyme. Following an overnight fast the patient swallows a small
amount of 14C-urea. Gastric H. pylori cleaves the urea into ammonium and 14C-
bicarbonate which is absorbed into the bloodstream (Fig. 6). Under the action of
carbonic anhydrase, bicarbonate is transformed into 14C-carbon dioxide, which is
exhaled in the breath. Carbon-14 activity in the breath is therefore an indicator of
H. pylori urease activity.
Gastrointestinal 207
10
tive results are rare, but occasionally bacterial overgrowth from species other than H.
pylori will metabolize the labeled urea.
Techniques are now available for using urea labeled with non-radioactive car-
bon-13, though the radioactive format is still the most widely available. The non-
radioactive method requires a standardized meal, a larger dose of urea and more
specialized equipment (mass spectrometer or laser spectroscope).
Clinical
The accuracy of diagnosing PUD from history is poor. The diagnosis is con-
firmed in only 50% of those in whom it is suspected. After excluding ulcers related
to nonsteroidal antiiflammatory drugs (NSAIDs), the vast majority of duodenal
and gastric ulcers are caused by chronic infection with H. pylori. With standard
treatment regimens such as a one-month treatment with an H2 receptor antagonist
or a proton pump inhibitor, there is a high rate of ulcer relapse (50 to 90%) after
acute ulcer healing. Eradication of the organism dramatically alters the natural history
of PUD with a marked decrease in ulcer recurrence. Highly effective (>90%) eradi-
cation of H. pylori is achieved with 7-10 days of a triple drug regimen (bismuth or a
proton pump inhibitor and two antibioticsclarithromycin and either metronida-
zole or amoxicillin). Adherence to such complex regimens is variable and side effects
are not infrequent.
The traditional approach to the diagnosis of PUD has included either an upper
GI series or endoscopy to document an ulcer. Determination of H. pylori status
followed by eradication therapy of patients found to be H. pylori positive may be a
rational and cost-effective approach to the investigation and management of pa-
10 tients with suspected PUD (Fig. 7). With a test and treat approach, a positive H.
pylori breath test removes the need for endoscopy if no alarm features (anemia,
evidence of bleeding, weight loss or early satiety) are present. Due to the increased
risk of malignancy in older patients, this approach should be confined to those less
than 50 years of age. A small number (20-30%) of patients with H. pylori infection
and non-ulcer dyspepsia also appear to benefit from eradication therapy, though
this remains controversial.
H. pylori has been associated with a two- to threefold increase in risk for gastric
cancer. Studies are in progress to determine if H. pylori eradication will alter the risk
of malignancy. Gastric maltoma, a lymphoproliferative disorder of gastric mucosal
associated lymphoid tissue (MALT), is related to infection with H. pylori. Eradication
of H. pylori in early stages of this low-grade tumor may lead to complete remission.
Surgical resection and chemotherapy are also effective treatment modalities. If a
non-operative approach to management is chosen then these patients need to be
followed very closely.
If peptic ulcer disease is uncomplicated, i.e., no evidence of bleeding, perfora-
tion or obstruction, then documentation that eradication has been successful is felt
to be unnecessary. The success of eradication should be confirmed in all those with
complications, taking care to avoid medications that can give a false negative examination.
Gastrointestinal 209
10 Additional Reading
1. Kim CK, Worsley DF, Machac J. Interventions in gastrointestinal nuclear medi-
cine. In: Freeman, LM, ed. Nuclear Medicine Annual 1996. Philadelphia: Lippincott
Williams and Wilkins, 1996:213-258.
A well-rounded collection of tidbits for imaging and helpful hints for interpretation of
examinations.
2. Klein HA and Wald A. Esophageal transit scintigraphy. In: Freeman, LM, Weissman
HS, eds. Nuclear Medicine Annual 1988. Philadelphia: Lippincott Williams and
Wilkins, 1988:79-124.
An older but still excellent discussion of scintigraphic approaches to measuring esoph-
ageal motility.
3. Malmud LS, Fisher RS, Knight LC, Rock E. Scintigraphic evaluation of gastric
emptying. Semin Nuc Med 1982; 12(2):116-125.
While an older reference, it is still relevant and written by one of the founders of this
field.
4. Thomson ABR, Shaffer EA eds. First Principles of Gastroenterology: The Basis of
Disease and an Approach to Management. Fourth edition. Mississauga: Astra Zeneca
Canada.
An excellent resource for students or busy clinicians. This book was published as a result
of an unrestricted educational grant from Astra Zeneca Canada Inc. which is listed as
the publisher. The book is endorsed by The Canadian Association of Gastroenterology
and the Canadian Association for the Study of the Liver.
5. Van Santen S, Flook N, Chiba N, Armstrong D et al. An evidence-based approach
to the management of uninvestigated dyspepsia in the era of Helicobacter pylori.
Can Med Assoc Journal 2000; 162 (12 Suppl):S3-23.
A thorough review of dyspepsia and its management. Discusses the role of a test and
treat approach to dyspepsia.
CHAPTER 1
CHAPTER 11
Hepatobiliary Imaging
Reinhard Kloiber and Gary R. May
Introduction
Historically, liver imaging using radionuclides focused on the detection of liver
masses. More recently, anatomic imaging modalities including ultrasound, CT and
MRI have replaced colloid imaging because of their greater sensitivity and ability to
distinguish cystic from solid lesions. Solid masses may nevertheless have a nonspecific
appearance on anatomic imaging requiring the use of radiopharmaceuticals that
identify certain cell lines, metabolic properties or surface characteristics of specific
lesions. The modern technetium-99m labeled hepatobiliary agents retain a pivotal
role in studying biliary flow and dynamics.
Radiopharmaceuticals
IDA Analogues
Analogues of technetium-99m labeled iminodiacetic acid (99mTc-IDA) first
became available for clinical use in the 1970s and remain the most widely used
radiopharmaceuticals for hepatobiliary imaging (Fig. 1). After intravenous injection,
these organic anions are taken up by hepatocytes in a manner similar to bilirubin.
They are secreted into bile without conjugation. 99mTc-IDA uniformly mixes with
bile, thereby becoming a marker of bile flow to the gallbladder and bowel. No
significant intestinal reabsorption or enterohepatic cycling occurs. Ideal agents exhibit
high extraction efficiency by liver (even in the presence of liver dysfunction), rapid
transit, and high concentration in bile. Commercially available 99mTc-disofenin
(Hepatolite) and 99mTc-mebrofenin (Choletec) meet these criteria.
Colloids
99mTc-sulfur colloid and a variety of other colloids are phagocytosed from the
blood stream by Kupffer cells in the liver and other reticuloendothelial cells. Under
normal circumstances, 85% of the dose accumulates in liver, 10% in the spleen and
the remainder in bone marrow (Fig. 2). Metastases, cysts and abscesses all displace
normal Kupffer cells and appear as cold defects. Some benign focal liver lesions
contain Kupffer cells and the benign nature of the mass can be confirmed by showing
colloid uptake within it.
Red Blood Cells
Red blood cells can be labelled with technetium-99m (99mTc-RBC) and are used
to identify cavernous hemangiomas within the liver. These consist of large blood
filled spaces. Blood flow is generally low in relation to the volume of blood present
and filling after intravenous injection of a substance may require a significant period
of time. Radiographic contrast material leaks rapidly from the intravascular space
Nuclear Medicine, edited by William D. Leslie and I. David Greenberg.
2003 Landes Bioscience.
212 Nuclear Medicine
Figure 1. Structure of iminodiacetic acid. The groups on the benzene ring labelled
R can be hydrogen or short organic chains. They determine the specific attributes
of the analogue such as avidity of uptake by the liver and rates of excretion.
into the extracellular fluid space. Late enhancement of a mass on CT therefore does
not prove that the lesion is filled with blood but can also be seen in malignant
masses with an expanded extracellular fluid space. 99mTc-RBCs remain confined to
the circulation and even hemangiomas that require hours to equilibrate can be
correctly diagnosed. (For more on the technique of labelling RBCs, refer to Chapter 4.)
Other tracers
Antibodies
Antibodies directed at characteristic surface antigens selectively expressed by
specific neoplasms, such as prostate specific antigen (PSA) or carcinoembryonic
antigen (CEA), appear in theory to be the ideal agents. Unfortunately, all antibodies
or fragments are metabolized in liver yielding a high background. Mapping of normal
11 tissue with 99mTc-sulfur colloid and subtraction from the antibody images to highlight
areas of abnormal binding can improve detectability. Antibody imaging of liver lesions
holds theoretical promise but, after years of investigation, is still of limited clinical
utility.
Receptor-Specific Agents
The indium-111 labelled pentapeptide analogue of somatostatin (111In-
pentetreotide or octreoscan binds to receptor-positive carcinoids and a variety of
pancreatic islet cell and neuroendocrine tumours. Lesion to liver ratios are often
sufficiently high to confirm the nature of metastases (Fig. 3). Extrahepatic sites are
demonstrated with even greater sensitivity. The technique is complementary to CT.
Gallium Citrate
Certain neoplasms, including lymphomas, carcinoma of the lung, sarcomas and
hepatomas, may exhibit avid uptake of gallium-67 citrate (67Ga). Gallium acts as an
iron analogue. It binds to transferrin after intravenous injection and is stored in liver
and bone marrow. Small quantities are excreted in bile. Normal liver therefore contains
moderately high background activity. Many adenocarcinomas originating in the
gastrointestinal and genitourinary tracts show only low grade uptake and metastases
appear as photopenic defects relative to normal liver. Imaging characteristics of 67Ga
are poor compared with 99mTc and the masses must be substantial in size to be
Hepatobiliary Imaging 213
Figure 2. Normal planar 99mTc-sulfur colloid liver spleen imaging. Activity throughout 11
the liver and spleen are uniform. Splenic activity is less than liver activity. The
bone marrow is only faintly seen.
Fluorodeoxyglucose
Fluorine-18 labelled fluorodeoxyglucose (18F-FDG) is a glucose analogue taken
up by cells through the same transport system as glucose. Following cellular uptake,
FDG is phosphorylated. Since FDG-phosphate cannot be metabolized further, it is
effectively trapped in the cell and can be used as a marker of glucose metabolism.
As with gallium, uptake is nonspecific. Because of the high sensitivity of anatomic
imaging modalities in detecting liver masses, the greatest utility of FDG imaging is
in the detection of extrahepatic sites for tumour staging (see Chapter 15). Availability
of positron emitting radioisotopes, including 18F, has been limited to date but is
improving.
214 Nuclear Medicine
Figure 3. Metastases from malignant carcinoid. A 46 year old woman with past
resection of a small bowel carcinoid presents with carcinoid syndrome. A CT image
during CT arterial portography (a) shows multiple hypodense masses (arrows). The
corresponding 111In-pentetreotide SPECT transaxial image (b) confirms that these
are somatostatin receptor positive metastases (arrows).
EF =
(A < Bg a ) < (B < Bg b ) 100%
(A < Bg a )
Hepatobiliary Imaging 215
11
Figure 4. Hepatocellular carcinoma (HCC) in a cirrhotic liver. A 75 year old man
with alcoholic cirrhosis presents with worsening liver failure. The enhanced CT
image (a) shows the small cirrhotic liver with cavernous transformation of the portal
vein and surrounding ascites. The posterior aspect of the right lobe of the liver
shows a bulbous projection suggesting a mass (white arrow). The planar (b) and
transaxial SPECT (c) 99mTc-sulfur colloid images show the mass to be devoid of
Kupffer cells indicating that it is not regenerating liver tissue (black arrows). A more
anterior defect on 99mTc-sulfur colloid SPECT images represents scarring. 67Ga activity
within the mass on planar (d) and transaxial SPECT images (e) is consistent with
HCC (black arrowheads).
11
Figure 5. Normal 99mTc-IDA study. Anterior and right lateral images immediately
after injection (a) show uniform activity throughout the liver. Activity in the heart is
low-grade with normal liver function. By 30 minutes (b) the biliary tree and bowel
are seen. The normal gallbladder (arrow) is seen within one hour (c). A right lateral
view is important to prevent potential confusion between activity in the gallbladder
(anteriorly placed) and duodenum (posteriorly placed) on the anterior view alone.
post contraction activity is corrected for decay. The method assumes complete mixing
of the radiotracer with bile. The EF can be overestimated if abundant stones or
sludge are present because only the portion of the gallbladder volume filled with
radiolabelled bile would be measured. The EF may have a role in evaluating patients
Hepatobiliary Imaging 217
11
with acute or chronic biliary type pain in the absence of demonstrable calculi (see
section on Biliary-Type Pain in the Absence of Calculi). Emptying rates and patterns
of emptying have not been sufficiently studied to determine their relevance.
Time activity curves generated from regions over the heart, liver, and bile ducts
offer no advantage over biochemical indices in assessing liver function but may be of
value in partial obstruction or sphincter of Oddi dysfunction (see section on
Postcholecystectomy Pain).
Acute Calculous Cholecystitis: The Emergency Room Setting
The vast majority of outpatients presenting with symptoms of acute cholecystitis
have cystic duct obstruction by calculi. Ultrasound is typically the initial imaging
218 Nuclear Medicine
Figure 7. Rim sign in acute cholecystitis. 15 minute (a) and 15 minute post morphine
(b) images fail to demonstrate any filling of the gallbladder indicating cystic duct
obstruction in this 42 year old man with acute right upper quadrant pain. The
inflamed liver adjacent to the gallbladder fossa shows delayed excretion compared
with the remainder of the liver (arrowheads). This finding adds specificity to the
diagnosis of acute cholecystitis.
procedure. It is usually readily available in the emergency room setting, can assess
multiple abdominal organs and yields results quickly. It is the procedure of choice to
detect calculi in the gallbladder with a sensitivity approaching 98%. In the presence
of acute inflammation it can show thickening of the gallbladder wall, pericholecystic
11 fluid, and elicit localized tenderness (the sonographic Murphys sign). Unfortunately,
calculi are a common problem and unless they are located in the gallbladder neck or
cystic duct, they imply acute cholecystitis in only 50 to 70% of cases. Gallbladder
wall thickening and fluid are infrequent findings and are also nonspecific, being
found with ascites and other edematous states. CT has similar advantages and
disadvantages. It is better at demonstrating the distal portion of the common bile
duct, which may be obscured by gas on US, but clearly cannot be used to assess
tenderness over the gallbladder.
In this setting, nonvisualization of the gallbladder within 60 minutes on
scintigraphy has a sensitivity for acute cholecystitis approaching 100%. The finding
is unreliable if biliary excretion is reduced as a result of hepatocellular disease or
cholestasis. Of patients with proven chronic cholecystitis, 80% will show normal
gallbladder filling and the remainder will have delayed visualisation (after 1 hour) or
persistent nonvisualization. The latter results in an unavoidable number of false
positive studies. The proportion depends on the prevalence of chronic cholecystitis
in the population studied.
Infusion of morphine at a dose of 0.04 to 0.10 mg/kg is used to accelerate filling
of the gallbladder by causing spasm of the sphincter of Oddi and increasing intrabiliary
pressure. Morphine-induced gallbladder visualization usually occurs within 30
minutes and imaging beyond that time is unnecessary unless there is delayed excretion.
Hepatobiliary Imaging 219
fossa. The inflamed gallbladder was enlarged on ultrasound (US) and is only partially
filled by the radiotracer (arrow). A faint rim sign is also present (arrowheads).
Considering the vagaries of clinical assessment, selection biases (likely skewed to-
wards those cases with the most significant pain), and the small number and pro-
portion of patients who have surgery, it is impossible to calculate meaningful
sensitivity and specificity.
Careful patient selection and meticulous technique are essential to avoid a falsely
reduced gallbladder EF:
Use of a meal to stimulate gallbladder contraction is unreliable. CCK is released
from the duodenal mucosa in response to the presence of nutrients in the duodenal
lumen, particularly fats and proteins. The composition of the meal will affect the
rate of delivery to the duodenum because of the effect on gastric emptying. Delayed
gastric emptying resulting from diseases such as diabetes or renal failure will impair
CCK release. Release of CCK is also dependent on normal pancreatic exocrine
function and an intact duodenal mucosa which may be affected by conditions such
as celiac disease or enteritis.
Infusion of CCK-8 is the optimum method to stimulate gallbladder emptying.
Maximal and reproducible gallbladder contraction is achieved by a physiologic
infusion at a rate of 0.020 g/kg per hour over a one hour period. Rapid infusion or
bolus injection may result in submaximal contraction and nonspecific cramping
abdominal pain. Pain reproduction is a poor indicator of disease and should not be
used as a criterion.
Gallbladder contraction is also dependent on cholinergic innervation, calcium
ion concentration and pressure within the biliary tree. A wide variety of medications
including those with cholinomimetic or inhibitory action, calcium channel blockers,
opiates, and possibly H2 receptor antagonists and progestins may alter the normal
response to CCK-8. Vagotomy, advanced neuropathy of diabetes and generalized
smooth muscle abnormalities as may be seen with pseudo-obstruction syndromes
can lead to reduced contractility (Table 1). The effect of body habitus, phase of 11
menstrual cycle, oral contraceptives remain uncertain.
Reduced gallbladder contraction provides supporting evidence of a biliary cause
for pain and for potential benefit from cholecystectomy. Scintigraphy should be
reserved for patients thought to have biliary type pain by an experienced clinician.
The patients must be free of serious concurrent illness and should not be taking
medications which may impair gallbladder contractility. The use of scintigraphy as a
screening procedure in patients with nonspecific upper abdominal pain is not
recommended.
Assessment of the nature of the pain by an experienced clinician is probably the
key criterion. Reduced gallbladder contractility in a patient with typical biliary type
pain may provide confirmatory evidence of gallbladder dysfunction.
Post-Cholecystectomy Pain
Cholecystectomy results in remission or improvement in biliary type pain in
most patients with calculi and pathologic findings of chronic cholecystitis. Ten to
40 % of patients may continue to experience symptoms. Most of these can be traced
to nonbiliary causes such as pseudo-obstruction syndrome, gastroesophageal reflux,
pancreatitis, and other unrelated conditions. Others have anatomic abnormalities of
the biliary tree such as retained calculi, strictures or neoplasms which are best
diagnosed by anatomic studies such as magnetic resonance cholangiopancreatography
222 Nuclear Medicine
11
bile duct curve (a). Symptoms resolved after endoscopic sphincterotomy and bile
duct emptying also returned to normal (b).
Hepatobiliary Imaging 225
Figure 11. Bile leak after endoscopic cholecystectomy. A 73 year old man suffered
increasing pain two days after endoscopic cholecystectomy. The immediate image
(a) shows normal liver activity. Retention of radiotracer is seen in the biliary tree at
30 minutes (b). A defect in the distal CBD (arrow) is a retained calculus. Bile
extravasation surrounding a biloma in the gallbladder fossa (arrowheads) is seen at
60 minutes (c).
99m
Tc-IDA is a marker for hepatocytes. Masses containing hepatocytes such as
FNH, adenoma or hepatoma do not contain bile ducts connected to the biliary tree.
Images performed shortly after injection may or may not demonstrate the mass. If
visible, it may appear more or less intense than normal liver depending on the relative
uptake. As normal liver excretes the tracer, the masses become relatively more intense
than normal liver. The technique is most useful in localizing the masses on delayed
SPECT images as the precise location may be difficult to determine from US
examinations where images can be obtained in any orientation. Unfortunately, benign
and malignant masses cannot be separated with complete confidence as well-
differentiated hepatomas will also accumulate 99mTc-IDA.
11
Cavernous Hemangioma
Cavernous hemagiomas are single or multiple well-circumscribed clusters of large
blood filled sinuses (Figs. 12 and 13). They occur predominantly in women. Most
are asymptomatic and are discovered as incidental findings during CT, MRI or
ultrasound performed for other reasons. Rarely pain may result from hemorrhage or
thrombosis.
On US most cavernous hemagiomas are well-defined and hyperechoic because
of reflections from multiple internal septa and vessel walls. Mixed echogenicity can
result from thrombosis, hemorrhage, necrosis and scarring. Findings are nonspecific
and hemangiomas cannot be confidently distinguished from vascular metastases,
FNH, adenoma, or hepatoma.
On CT hemangiomas are hypodense without contrast enhancement. After
contrast injection there is initial peripheral or globular enhancement (puddling)
which progresses toward the center. Small lesions may enhance rapidly and may be
indistinguishable from other vascular masses. Large hemangiomas may not enhance
completely or fill only later when most of the contrast is in the extracellular fluid
space.
On MRI most hemangiomas are well defined and hyperintense on T-2 weighted
images. Large lesions and those with hemorrhage, thrombosis or scarring tend to
Hepatobiliary Imaging 227
have atypical appearance. Differentiation from other vascular masses may be diffi- 11
cult.
The classic appearance on 99mTc-RBC imaging is an area of reduced perfusion
on a flow study and reduced activity on immediate planar or SPECT because blood
flow is low in relation to blood volume in the hemangioma. After 1 to 2 hours the
radiolabelled cells have equilibrated and hemangiomas become uniformly intense
and similar in intensity to the heart and major vessels. With this pattern, specificity
approaches 100% making 99m Tc-RBC imaging the procedure of choice.
Hemangiomas as small as 1 cm can be visualized. When they are 1.5 cm or more,
the sensitivity for detection exceeds 80%. Hemangiomas with extensive hemorrhage,
necrosis, or scarring may not have the typical pattern, and small lesions adjacent to
major vessels can be missed. Small suspected hemangiomas that are deep within the
liver or adjacent to the heart are best studied with MRI.
Focal nodular hyperplasia
Focal nodular hyperplasia (FNH) is a hamartoma containing cords of hepatocytes
and Kupffer cells separated by sinusoids and arranged into nodules. The nodules are
separated by fibrous septa which radiate from a central scar and contain rudimentary
bile ducts. The arterial supply originates at the center and radiates outward from the
228 Nuclear Medicine
11
Figure 12. Multiple cavernous hemangiomas. Utrasound performed for pain in this
41 year old woman showed multiple large liver masses. Early and late images from
an infused CT (a and b) show progressive centripetal (lesion 1), globular (lesion 2)
and inhomogeneous (lesion 3) enhancement. The larger lesions are incompletely
opacified. The masses are hypovascular on the flow and immediate images (c and
d) and uniformly filled with 99mTc-RBC on transaxial (e) and coronal (f) delayed
SPECT images. Activity in the hemangiomas is equivalent to the cardiac chambers
(H).
outward from the center through the septa, unlike the peripheral origin and centrip-
etal flow seen with adenomas and metastases.
Hepatobiliary Imaging 229
Figure 13. Multiple hemangiomas. The infused CT (a) done for staging of lung
cancer in this 70 year old man shows non-specific small peripherally enhancing
lesions in the liver (white arrows) with a larger lesion centrally (black arrow). 99mTc-
RBC activity is equivalent to that in major vessels on the transaxial (b) delayed
SPECT image (arrows).
that is diagnostic for FNH. Twenty to 30% show no demonstrable activity or are 11
too small to assess confidently. In the remainder, there is variable activity confirming
the benign nature of the mass. A more specific diagnosis can be made by also
demonstrating 99mTc-IDA uptake in the lesion.
Hepatocellular adenoma
Heptatocellular adenoma (HA) is a rare neoplasm that occurs almost exclusively
in young women or those receiving ovarian hormone replacement therapy. Unlike
FNH, HA has a high risk of hemorrhage, necrosis, infarction and even rupture with
significant morbidity and mortality. HA consists of normal appearing hepatocytes.
Arterial supply is from the periphery. The appearance on anatomic imaging is
nonspecific.
Kupffer cells may be present but not in sufficient quantities for the masses to
demonstrate 99mTc-sulfur colloid uptake. The pattern on 99mTc-IDA imaging is similar
to FNH. In the absence of 99mTc-sulfur colloid uptake, this appearance is nonspecific
and can also be seen with well-differentiated hepatomas.
Hepatocellular Carcinoma
Hepatocellular carcinoma (HCC) in North America typically occurs in the
presence of cirrhosis making diagnosis difficult using the standard anatomic imaging
modalities. Patients exposed to aflatoxins or anabolic steroids, or those with the
230 Nuclear Medicine
Figure 14. Focal nodular hyperplasia (FNH). The infused CT (a) performed for
abdominal trauma in this 32 year old woman shows an isodense mass in the lateral
segment of the left lobe of the liver (arrows). A central hypodense area suggests the
central scar of FNH. Transverse (b) and coronal (c) 99mTc-sulfur colloid SPECT images
show greater activity in the lesion (arrowheads) compared with normal liver
confirming the diagnosis of FNH.
fibrolamellar variant of HCC, may have an otherwise normal liver. In that situation
HCC often presents as a nonspecific vascular mass. A high proportion of advanced
tumours will have portal vein thrombosis in the absence of portal vein compression.
99m
Tc-sulfur colloid imaging is useful in evaluation of a liver distorted by cirrhosis.
Mass like areas that accumulate colloid represent localized areas of hypertrophy or
regenerating nodules. Areas of absent colloid activity that accumulate gallium or
some other marker of metabolic activity likely represent hepatoma (Fig. 4). Since
11 close to 90% of all hepatomas accumulate gallium to some degree, simultaneous
99mTc-sulfur colloid subtraction yields the greatest sensitivity. In the absence of
cirrhosis, the technique is less specific and other gallium-avid tumours cannot be
differentiated.
On 99mTc-IDA imaging, the level of uptake by the HCC reflects the level of
differentiation and greater accumulation carries a better prognosis. Well-differentiated
HCC however cannot be differentiated from HA with confidence. Elevated alpha-
fetoprotein supports a diagnosis of HCC.
Focal Fatty infiltration and Sparing
Fatty infiltration of the liver is seen with alcohol abuse, obesity, diabetes mellitus
and other disorders of metabolism. Increased fat content results in hyperechogenicity
on US and decreased density on CT. In some cases the fat deposition is nonuniform
with areas of focal fatty infiltration or focal fatty sparing. The focally altered
echogenicity on US or density on CT may be difficult to distinguish from metastases.
Hepatobiliary Imaging 231
Figure 15. Focal fatty infiltration. Ultrasound performed for abdominal pain in this
27 year old woman showed a 4 x 5 cm. echogenic focus in the liver. A noninfused
CT (a) shows a localized area of hypodensity in the medial segment of the left lobe
of the liver (arrow). A traversing vessel appeared undisplaced suggesting focal fatty
infiltration rather than a mass, but the appearance was not definitive. Transaxial (b)
and coronal (c) 99mTc-sulfur colloid images show uniform activity confirming a benign
process.
Lack of displacement of vessels and the characteristics of fat on MRI are helpful in
the diagnosis in both situations. Kupffer cells are unaffected and the liver appears 11
homogeneous on 99mTc-sulfur colloid imaging (Fig 15).
Frequently Asked Questions (FAQs)
Can scintigraphy predict whether or not a patients atypical
chronic symptoms are related to the presence of known
gallstones?
No. Gallstones are common and do not necessarily result in symptoms. Delayed
visualization or nonvisualization of the gallbladder or abnormal emptying can
certainly be seen in asymptomatic patients. The value of these scintigraphic findings
in predicting who will have a favorable response to cholecystectomy is unknown.
Clinical assessment as to the nature of the pain is the most reliable predictor.
Does nonvisualization of the gallbladder in a patient with acute
RUQ pain confirm the presence of acute cholecystitis?
No. Visualization of the gallbladder effectively excludes acute calculous
cholecystitis. Cystic duct obstruction in the absence of acute inflammation, a
contracted gallbladder filled with stones, inadequate fasting, impaired liver function,
and other acute intra-abdominal disease can also result in nonvisualization. Focal
232 Nuclear Medicine
in the gallbladder fossa or a rim sign add specificity to the diagnosis of acute
cholecystitis but are relatively uncommon findings.
Are all masses that accumulate 99mTc-sulfur colloid benign?
Yes. Anatomic variations in the shape of the liver, regenerating tissue, focal fatty
infiltration or sparing and FNH all contain Kupffer cell populations. If one of the
above named conditions is suspected, 99mTc-sulfur colloid imaging is the initial
procedure of choice. The only caveat is that the mass must be large enough to be
resolved as a photopenic area in the absence of 99mTc-sulfur colloid uptake.
Are all masses that accumulate 99mTc-IDA benign?
No. Although the level of uptake reflects the degree of differentiation of
hepatocytes in the mass, well-differentiated HCC will show activity comparable to
some cases of HA and FNH. 99mTc-IDA imaging is an ancillary procedure to localize
a mass or to make a specific diagnosis of FNH in a mass already proven to be benign
by 99mTc-sulfur colloid uptake.
Does failure to visualize a hemangioma suspected on US or CT
using 99mTc-RBC indicate a malignant mass instead?
No. Hemangiomas may not be visible because of their small size or proximity to
the heart or larger vessels deep within the liver. MRI may be useful in that situation.
Hemorrhage, thrombosis, or fibrosis may obliterate the blood filled spaces. Serial
anatomic imaging, to ensure stability in size, or biopsy may be necessary.
Additional Reading
1. Heiken P. The liver. In: Lee J, Segal S, Stanley R, Heiken J, eds. Computed Body
Tomography with MRI Correlation. 3rd ed. Philadelphia: Lippincott-Raven,
11 1998:701-778.
A detailed review of the role of CT and MRI in the diagnosis of liver masses.
2. Kinnard M, Alavi A, Rubin RA, Lichtenstein GR. Nuclear imaging of solid hepatic
masses. Seminars in Roentgenology 1995; 30:375-395.
A detailed description of the role of scintigraphy in characterization of liver masses.
3. Katsuyoshi I, Kazumitsu H, Takeshi F et al. Liver neoplasms: Diagnostic pitfalls in
cross-sectional imaging. Radiographics 1996; 16:273-293.
A pictorial overview of problems in diagnosis of liver lesions with CT and MRI.
CHAPTER 1
CHAPTER 12
Inflammatory Disorders
William D. Leslie and Pierre Plourde
Pathophysiology of Inflammation
Inflammation is the organized reaction of the microcirculation to tissue injury.
Tissue injury sets in motion a carefully orchestrated complex cascade of events in an
attempt to eliminate the offending agent and repair any damage. Acute inflammation
is typically brief (lasting minutes to days) and features the neutrophil as the dominant
cell. In contrast, chronic inflammation persists much longer (from weeks to months)
and the cellular response is dominated by tissue macrophages and lymphocytes. It is
important to recognize that there is a continuum in these two processes, and most
inflammatory responses consist of a measure of both.
Acute inflammation leads to three basic tissue responses: (a) vasodilation and
increased blood flow (hyperemia) with slowing of the microcirculation (stasis), (b)
loss of endothelial integrity with leakage of fluid and plasma components, and (c)
emigration of leukocytes into the extravascular tissue. The clinical counterparts of
these pathologic events are local hyperemia, swelling and pain. Acute inflammation
is mediated by plasma-derived and cellular components (Fig. 1). The importance of
adhesion molecules has recently been recognized. These are membrane glycoproteins
found on the surface of activated endothelial cells, leukocytes and platelets. The
complementary expression of these molecules on leukocyte and endothelial surfaces
induces an interaction characterized initially by rolling, subsequent adhesion and
finally transmigration of the leukocyte between adjacent endothelial cells.
Granulocytes combat invading microorganisms in several ways. If possible, the
organisms are phagocytosed and then destroyed with powerful oxidants such as
superoxide and hydrogen peroxide generated through a respiratory burst. Other
microorganisms are rendered susceptible to phagocytosis through opsonization with
immunoglobulins or activated complement. Lactoferrin is discharged from the
granulocytes secondary granules and binds iron, an essential growth factor for many
bacteria and fungi. Neutrophils predominate in the inflammatory infiltrate during
the initial 6-24 hours but have a short life span. They spend a brief period of time in
the circulation (half-life about 6 hours) and, having migrated out of the vascular
space, are then unable to reenter it. In the extracellular space neutrophils undergo
programmed cell death (apoptosis) after only 24-48 hours.
In some situations chronic inflammation evolves from an acute inflammatory
process, but in other scenarios (including viral, parasitic, autoimmune, foreign body,
and malignant inflammatory processes) it represents the primary immune response.
The chronic tissue response is often characterized by fibrosis, cellular proliferation
and granulation tissue formation. The activated T-lymphocyte is responsible for the
production of chemotactic and growth factors which act to recruit monocytes and
induce tissue proliferation of macrophages. Activated tissue macrophages may fuse
Nuclear Medicine, edited by William D. Leslie and I. David Greenberg.
2003 Landes Bioscience.
234 Nuclear Medicine
Preparation
In order to cross the leukocytes cell membrane, indium must be chelated to a
lipophilic carrier molecule such as oxine or tropolone. The labeling protocol is
outlined in Table 1. Not all departments have the ability to perform cell labeling as
this requires experienced personnel, laminar flow hoods and impeccable quality 12
control to ensure a safe and satisfactory product. Although there is no specific patient
preparation for the test, it is important to ensure that the patient has an adequate
number of circulating leukocytes (at least 3x109/L). A large quantity of anticoagulated
blood (40-50 ml) must be withdrawn from the patient to obtain sufficient numbers
of leukocytes. Autologous leukocyte scanning is not feasible in infants due to the
large volume of blood required. Similarly, neutropenic patients are unsuitable subjects.
In rare circumstances, donor leukocytes can be used.
Leukocytes must be handled gently to avoid activation as this compromises their
viability and ability to migrate to sites of inflammation. In general, the period of
time during which leukocytes remain ex vivo should be as short as possible and
ideally less than 2 hours.
Imaging Considerations
Imaging typically involves both early and late scans. Early imaging is performed
between one and four hours post-injection and the delayed images are performed at
24 hours. Due to the long half-life of 111In (approximately 3 days), further images
up to 48 hours are also possible. With inflammatory bowel disease leukocyte
236 Nuclear Medicine
12
Figure 3. Normal 111In-leukocyte scan. Note the normal accumulation within bone
marrow, liver and spleen. As in this patient, the spleen typically shows the most
intense accumulation of 111In-leukocytes.
Due to the low injected activities, imaging is prolonged (10-20 minutes per
static view) and the low counts result in limited image detail. This can make it
difficult to localize a focus of uptake, particularly in the peripheral extremities which
normally lack hematopoietic marrow and where leukocyte uptake can appear as an
isolated point in space. The gamma camera is able to distinguish isotopes photon
energies (111In energies 247 and 173 keV versus 99mTc 140 keV). This makes it
possible to scan two different tracers simultaneously: the 111In- leukocytes detect
inflammation and a 99mTc-labeled tracer is used for localization and diagnostic
correlation. Examples of dual-isotope strategies are as follows: 99mTc-bone tracers
for localization and simultaneous assessment of suspected osteomyelitis (Fig. 4);
99mTc-colloid to map marrow in suspected infection of an orthopedic prosthesis;
technetium-99m also leads to greater elution of the label from the WBC and therefore
the fractional uptake at 18-24 hours is usually less than with 111In-leukocytes.
Nonetheless, in direct comparisons the improved image quality seen with 99mTc-
leukocytes offsets any reduced retention such that the two agents are equivalent in
terms of their ability to detect inflammatory lesions.
Imaging Considerations
As with 111In-leukocytes, early and delayed imaging should be performed. Due
to the shorter half life, however, this is usually performed one and four hours post-
injection. Images can also be performed at 24 hours, although isotope decay and
normal bowel activity limit image quality.
12
Gallium
Gallium is a heavy metal that is handled by the body in a manner very similar to
iron. It is injected in the form of 67Ga-gallium citrate, the gallium component of
which then binds avidly to transferrin. The circulating metal-protein complex is
thought to localize at sites of inflammation through the following mechanisms:
increased delivery due to hyperemia, increased pooling due to vasodilation, passive
diffusion into the interstitium through leaky capillaries, and binding of gallium to
other iron-avid proteins such as granulocyte-derived lactoferrin or bacterial
siderophores. Lactoferrin is also responsible for the normal visualization of lactoferrin-
secreting glands such as the lacrimal glands and breasts. It is currently believed that
the most important factor is diffusion across inflamed capillaries with retention
within the interstitial space. There is negligible binding of gallium to viable
granulocytes and localization can occur in the absence of leukocytes. Gallium has
therefore proven to be an effective agent in immunosuppressed and leukopenic
patients.
Gallium normally concentrates in sites of iron storage (liver, erythropoetic bone
marrow and to a lesser degree spleen), in secretory organs that produce lactoferrin
Inflammatory Disorders 239
12
(lacrimal glands, salivary glands, breast) and in the skeleton (Fig. 6). In normal
persons, gallium is taken up by the liver (5%), spleen (1%), kidneys (2%), bone
marrow (5%), and skeleton (13%). During the initial 24 hours the urine accounts
for a small amount of gallium excretion (25%), and subsequent clearance proceeds
240 Nuclear Medicine
12
Figure 6. Normal 67Ga-gallium citrate whole body scan (left anterior, right posterior).
There is normal accumulation in the liver, bone marrow and soft tissues. Uptake in
the right upper quadrant of the abdomen reflects normal gut excretion.
very slowly (at 7 days post-injection the decay-corrected whole body retention is
still 65%). Variable uptake in the nasopharyngeal lymphoid tissue, lacrimal glands
and salivary glands can also be a normal finding. Posterior images of the abdomen
may reveal faint renal uptake up to 24 hours post-injection. Thymus and spleen
uptake may be prominent, particularly in children or following chemotherapy.
Abnormal patterns of gallium accumulation can be seen when the plasma carrier
protein, transferrin, becomes saturated e.g., following chemotherapy, blood
transfusion, hemochromatosis or iron therapy. The scan pattern tends to resemble a
bone scan, with prominent skeletal uptake, little soft tissue uptake, and increased
renal and bladder activity. If this is suspected then measuring serum iron and
transferrin should reveal the nature of the problem.
Inflammatory Disorders 241
12
Imaging Considerations
The usual dose of 67Ga-gallium citrate is 111-185 MBq (3-5 mCi) injected
intravenously. Higher doses up to 370 MBq (10 mCi) are used if tomographic SPECT
imaging is planned. Gallium localizes relatively slowly and images are usually first
obtained at 24-48 hours which is a disadvantage in acutely ill patients in whom a
rapid diagnosis is required. There is variable gallium excretion by the gastrointestinal
tract. Many groups routinely use cathartic bowel preparations to clear gastrointestinal
activity. In most situations, performing delayed views (up to one week post-injection)
will clarify if abdominal activity is physiologic excretion (which typically clears over
time) or inflammatory in nature (which usually stays fixed).
Monoclonal Antibodies
Monoclonal antibodies have long held the promise of increased diagnostic
specificity. In theory, such agents could be directed at antigens expressed on
immunologically competent host cells or at epitopes on infectious organisms. The
greatest experience has been with mouse-derived (murine) antigranulocyte antibodies
directed at cell surface antigens. Granulocytes in the circulation, as well as those
already localized at the inflammatory site, become labeled by isotope bound to the
injected monocloncal antibody. Some contribution from passive diffusion is probably
present though this has not been quantified. The use of monoclonal antibodies
directed towards leukocyte cell membrane antigens thus has the potential of
combining the best of both worlds. In addition to the excellent inflammatory
targeting of leukocytes, these agents can be administered through simple intravenous
injection. This avoids the inconvenience and risk of cell handling, cell activation
and reinjection of a blood product.
Soon after injection the tracer is confined to the vascular spaces, but then gradually
accumulates in liver, spleen and bone marrow (Fig. 7). As with 99mTc-leukocytes,
some of the 99mTc label separates from the antibody and can appear in the bowel
after 3 hours. Antibody fragments are nearly completely eliminated by the kidneys
12 and accumulate in renal tubular cells, which limit its use for evaluating the urinary tract.
Preparation
Several antibody preparations and radionuclide labels have been investigated,
including 111In, 99mTc and 123I. A convenient 99mTc-based kit formulation of Fab
fragments, known as sulesomab (LeukoScan), is currently approved in some
countries and undergoing clinical trials in others. A 99mTc-labeled intact IgG antibody
is available in Europe for (Granulozint). Both products target antigens shared by
carcinoembryonic antigen (CEA) and granulocytic normal cross-reacting antigens
(NCA-90 and NCA-95, respectively). Early clinical trials with 99mTc-anti- CD15
(LeuTech) also appear promising.
Although doses of antibody exceeding 10% antigen saturation can lead to cell
activation, this can be avoided by keeping the administered dose less than 1 mg.
Lower doses and the use of antibody fragments deficient in the Fc-terminal also
avoids the induction of human anti-mouse antibody (HAMA), an important
consideration since HAMA leads to accelerated immunoglobulin clearance and may
preclude additional administrations. Monoclonal antibodies are used in a variety of
laboratory tests such as serum CEA, a tumor marker, and results may be affected
(usually falsely low).
Inflammatory Disorders 243
Imaging Considerations
Imaging is similar to labeled leukocytes. In general, Fab fragments are much
smaller than intact antibodies (50 kD in comparison to 180 kD) leading to more
rapid clearance from the circulation and earlier diagnosis. Images as early as one
hour can show diagnostic findings and delayed images up to 24 hours can also be
performed. These agents appear to be comparable to 111In- leukocyte imaging for
identifying acute inflammation with sensitivity and specificity in the range of 77-
91% and 67-75%, respectively (Fig. 8).
Polyclonal Immunoglobulins
The ability of polyclonal human immunoglobulin IgG (HIG) to detect
inflammatory foci was a serendipitous discovery. Passive diffusion is probably the
primary mechanism of uptake with retention in the interstitium due to the large
molecular size (169 kD). HIG can be labeled with either 99mTc or 111In. However,
interest in radiolabeled immunoglobulins has waned considerably after reports that
viral infections (most notably hepatitis C) have been transmitted with therapeutic doses.
Nanocolloids
Nanocolloid refers to a human albumin colloid (Nanocoll TN). The 99mTc-
labeled particles (mean particle size of 30 nm and virtually no particles larger than
80 nm) are small enough to extravasate through leaky inflamed capillaries. Plasma
clearance by the reticuloendothelial system (RES) is extremely rapid which permits
imaging between 30 minutes and 2 hours post-injection. Normal uptake by liver,
spleen and hematopoetic tissue limits assessments of the abdomen, spine and areas
with surgically disturbed marrow. Good results have been reported in peripheral
bone and joint inflammation but sensitivity for soft tissue infections is only 59%
(compared to 97% for labeled leukocytes in the same series).
Newer Agents
A variety of newer radiopharmaceutical agents are being developed, and many of 12
these have already undergone clinical trials. Promising results have been reported
with a two-step streptavidin-biotin approach. The unlabeled streptavidin is injected
and allowed to localize through passive diffusion. Subsequently, the patient is injected
with radiolabeled 111In-biotin, which is much smaller and has high binding affinity
for streptavidin. Streptavidin-biotin imaging appears to be clinically useful for
orthopedic and vascular graft infections. Advances in the understanding of the
pathophysiologic mechanisms of diseases and the associated molecular biology may
lead to exciting developments in radiotracers that target endothelial selectin receptors
expressed in response to local inflammatory stimuli. In areas where positron emission
tomography (PET) is available, 18F-fluoro-deoxyglucose (FDG) appears to be a
sensitive technique.
All of the previous radiolabeling techniques fail to distinguish between bacterial-
mediated infection and non-bacterial inflammation. Recently, 99mTc-Infecton, a novel
ciprofloxacin based imaging agent, has been proposed to be specific for bacterial
infection. In preliminary reports this agent has shown a sensitivity of 70% and a
specificity of over 90% for detecting infective foci, with better imaging results than
even radiolabeled leucocytes. False negative results were attributed to previous or
current antibiotic treatment and/or infection with organisms that do not take up
244 Nuclear Medicine
Infecton. If this agent lives up to the initial promise then Infecton will have major
advantages over current imaging techniques, including radiolabeled leucocytes, and
may prove to be a superior method for localizing bacterial infections.
Clinical Role: General Principles
Inflammation reflects a pathophysiologic process whereas the gross anatomic
changes may lag far behind. Conventional imaging typically provides information
on only one part of the body, and detects disturbance to the normal anatomy. In
contrast, nuclear medicine can scan the whole body and localize occult disease
processes. In many cases, anatomy may be severely distorted by previous disease,
surgery or prosthetic materials with the result that conventional imaging is severely
degraded. For these reasons, nuclear medicine occupies an important niche in
diagnosing and localizing inflammation when conventional imaging is unsuccessful.
Inflammatory Disorders 245
12
12
the upper half of the panel are anterior views of the liver and in the lower half right
lateral views.
248 Nuclear Medicine
largely devoid of gallium uptake, pulmonary processes are well depicted with gallium.
The sensitivity is high for acute bacterial pneumonia (91%), pulmonary fungal
infection or tuberculosis (97%) and sarcoidosis (75%). Certain patterns of uptake
strongly suggest the diagnosis of sarcoidosis; these are known as the panda sign
(symmetrical uptake in the lacrimal and salivary glands suggesting the face of a
panda) and lambda pattern (bilateral symmetrical uptake in the paratracheal,
mediastinal and bronchopulmonary lymph nodes in the shape of the Greek letter).
As would be predicted from the lymphocytic nature of sarcoidosis, labeled leukocytes
are much less sensitive than gallium. Labeled leukocytes also show low sensitivity for
chronic pneumonias (30%) and pulmonary fibrosis (30-50%), where gallium is
again the preferred inflammatory imaging agent. Diffuse bilateral pulmonary uptake
can be a striking and unexpected finding (Fig. 13). The most common causes include
sarcoidosis, opportunistic infections such as Pneumocystis carinii, and drug-induced
pneumonitis. The most common drugs implicated are bleomycin, busulphan,
nitroureas, amiodarone, cyclophosphamide, methotrexate and nitrofurantoin. The
chest radiograph is frequently normal in these cases.
Normal gastrointestinal activity does not usually obscure renal abnormalities.
Focal renal gallium uptake can be seen with tumors such as hypernephroma or
lymphoma. Renal abscess or acute focal bacterial nephritis can also be responsible.
Sensitivity of gallium imaging for renal infection is 75% with a specificity of 87%.
As with pulmonary uptake, diffuse bilateral renal uptake of gallium can reflect bilateral
pyelonephritis, collagen vascular disease producing glomerulonephritis, and certain
forms of renal failure. Among the latter, allergic interstitial nephritis tends to show
more intense renal uptake than other causes. Many drugs have been implicated, and
it is important to establish the diagnosis as the renal failure is potentially reversible if
the drug is withdrawn and the patient treated with systemic steroids. Since gallium
is frequently used as a tumor tracer, it has reduced specificity in patients with
underlying malignancies.
Clinical Role in Fever of unknown origin (FUO) 12
Clinical Information
Not infrequently, patients will experience prolonged or recurrent febrile episodes
that defy diagnosis despite careful clinical assessment and conventional imaging. In
choosing a nuclear medicine technique, it is important to distinguish two broad
groups: those with a suspected acute pyogenic process (typically developing in
hospitalized patients or when there has been recent surgery) and a chronic process
(usually developing in a nonhospitalized patient). In the former, a pyogenic bacterial
cause is likely whereas in the latter the differential is much broader and includes
non-infectious etiologies such as lymphoma, solid tumors and sarcoidosis.
Classical FUO refers to persistent fever (at least 2-3 weeks) of 38.3 C or higher
in generally healthy adults that eludes diagnosis after at least three days of hospital
investigation (or three out-patient visits). Under this strict definition, an infectious
etiology is actually infrequent. This contrasts with persistent fever in hospitalized
(nosocomial) or neutropenic (fewer than 500 neutrophils/mm3) patients in whom
bacterial and fungal infections are the dominant cause. Persistent fever in an individual
with confirmed HIV infection is intimidating due to the extensive differential that
includes common infections (bacterial, viral, mycobacterial, fungal), non- infectious
250 Nuclear Medicine
12
12
12
Figure 14. Small cell lung cancer presenting as fever of unknown origin (FUO). An
81 year old woman had a past history of smoking and remote breast carcinoma
treated with mastectomy and post-operative radiotherapy. She fell at home and
began experiencing back and left chest pain, weakness, dyspnea and a ten pound
weight loss. Examination was unremarkable except for persistent fever and reduced
air entry in the right hemothorax. Extensive investigations in hospital (hematology,
biochemistry, blood cultures, abdominal ultrasound, echocardiogram) were normal.
The chest x-ray revelaed left upper lobe fibrosis from the previous radiotherapy. X-
rays of the lumbar spine and bone scan showed compression fractures of T11 and
L1. The possibility of lumbar osteomyelitis was entertained based upon the fever,
back pain, x-rays and bone scan. A gallium scan (left: anterior view, right: posterior
view), performed to confirm focal spine inflammation, showed minimal gallium
vertebral accumulation but marked mediastinal and left perihilar uptake (arrow). A
chest CT scan confirmed extensive mediastinal and perihilar adenopathy.
Subsequently scalene and supraclavicular nodes appeared and biopsy demonstrated
metastatic small cell carcinoma of the lung.
Inflammatory Disorders 255
Radionuclide Testing
Leukocyte scintigraphy is the preferred study for the investigation of possible
graft infection, particularly within the abdomen. The finding of focally intense
accumulation strongly suggests infection (Fig . 15) whereas low-grade diffuse uptake
by the graft is more in keeping with non-specific adherence to the pseudointima.
The diagnostic sensitivity with leukocyte scintigraphy is high, and specificity can be
improved by assessing the pattern of accumulation. The sensitivity of labeled
leukocytes consistently exceeds 90%, whereas the comparative sensitivity of CT scan
is 37-50%. The specificity of labeled leukocytes is more variable, ranging from 53%
to 100%. Causes of false-positive graft uptake include perigraft hematoma, graft
thrombosis, pseudoaneurysm and wound infections that do not involve the graft
directly. Nonspecific platelet adherence may persist for many weeks (or even months)
following surgery. The inadvertent labeling of platelets may contribute to nonspecific
graft uptake, and some groups routinely extract a platelet-depleted leukocyte
population using a gradient sedimentation method. Other agents, such as gallium,
polyclonal human immunoglobulin and nanocolloids, have not been as successful
as leukocytes.
Dialysis Fistulas
Dialysis fistulas are also prone to infection, particularly if they include implanted
prosthetic material, while native vessel arteriovenous fistulas are relatively resistant
to infection. Staphylococcus aureus accounts for 60-90% of access site infections,
suggesting direct contamination from skin at the time of access. As with other graft
infections, clinical findings can be extremely subtle and are completely absent in
approximately one-third of cases.
When an infectious process is limited to a small area away from the suture line
antimicrobial therapy and dbridement may be sufficient to eradicate the infection.
Unfortunately, in many cases this is insufficient and the arteriovenous fistula must
be completely removed.
Nuclear imaging of a dialysis fistula can be difficult to perform and challenging 12
to interpret. Frontal and profile views of the fistula are required, but it can be difficult
to control for the degree of forearm pronation. Furthermore, some camera designs
are cumbersome to position for areas close to the antecubital fossa. The large amount
of blood contained within the fistula can often be seen as low-grade activity, especially
on early images or when significant erythrocyte cross-labeling has occurred.
Occasionally, normal bone marrow can be a source of confusion and highlights the
importance of mapping any uptake to the fistula itself. Clinical assessment of the
patient while still under the gamma camera is an essential step. With these precautions
nuclear imaging can be highly accurate in the diagnosis of an infected fistula (Fig. 16).
Frequently Asked Questions (FAQs)
How do I determine which nuclear medicine technique to use?
Knowledge of the normal biodistribution and excretion patterns of the tracers
described above can lead to a rational approach to selecting one method over another
(Table 3). Soft tissue infections are usually bacterial and generate an acute pyogenic
inflammatory response. Labeled leukocytes are preferred due to their high levels of
256 Nuclear Medicine
Figure 15. Infected left femoral-popliteal graft before and after prolonged antibiotic
therapy. An 81 year old man with severe vascular insufficiency and gangrene of
the great toe underwent a distal left femoral-dorsalis pedis bypass graft using a
segment of saphenous vein. The left great toe was quite necrotic, with ulceration
extending to bone and purulent discharge. The thigh wound showed continued
12 serous drainage and with accumulation of a lymphocele which required exploration
and drainage three weeks after the original surgery. Throughout this time the patient
had been on antibiotics as draining fluid grew multiple organisms, although the
patient remained afebrile with only very slight local erythema. (A) The graft was
clearly localized on the 99mTc-erythrocyte blood pool image (left) and the striking
111In-leukocyte (right) accumulation in the left mid-thigh medially corresponded to
uptake, but passive diffusion agents can also be valuable. Assessment of the abdomen
requires delayed imaging and is best achieved with agents which do not show
physiologic bowel excretion. Overall, 111In- leukocytes are ideally suited to this
function (Fig. 17). Neither 111In- leukocytes nor 67Gallium show appreciable urinary
tract excretion. These agents can be valuable for confirming pyelonephritis or
Inflammatory Disorders 257
Figure 16. Infected right brachial-axillary graft. A 61 year old woman with end
stage renal disease due to diabetic nephropathy and hypertension was receiving
hemodialysis through a right brachial-axillary graft. She developed acute onset of
fever (up to 40.1 C) associated with non-specific abdominal and low back pain.
Blood cultures were positive for two different gram positive cocci: Staphylococcus
aureus and alpha-hemolytic Streptococcus. Clinical examination was unhelpful as
were many investigations (chest x-ray, chest/abdominal CT scan, echocardiogram
and lumbar spine CT) and consultations conducted in an attempt to determine the
cause of the fever. Dual-isotope scaning with 99mTc-erythrocytes (left) localized the
fistula while the 111In-leukocyte scan (right) demonstrated intense leukocyte
accumulation in the right brachial-axillary graft from a graft infection (arrow).
111 99m 67
In-Leukocytes Tc-Leukocytes Ga-Gallium citrate
Classic FUO ++ ++ +++
Nosocomial FUO +++ +++ ++
Soft tissue ++ +++ ++
Intra-abdominal +++ ++ ++
Inflammatory bowel +++ +++ +
disease
Urinary tract +++ + +++
Vascular grafts +++ ++ ++
such as compression fracture, tumor and Pagets disease. Prior antibiotic therapy
may be correlated with a higher incidence of false-negative scans and photon-deficient
uptake. Of patients with clinically resolved infection, many who initially present
with increased activity will show decreased vertebral activity on follow up studies.
In contrast to the poor imaging performance of 111In-leukocytes, 67Ga-gallium
citrate usually shows markedly increased uptake and is the preferred agent for
suspected vertebral osteomyelitis and/or discitis, frequently revealing a butterfly
pattern from paravertebral spread. Fungal and tuberculous vertebral infections can
also be detected with gallium. Gallium can be diagnostic where radiographic
abnormalities are originally lacking, and confirm a current and/or ongoing septic
process at sites where vertebral destruction is already present on x-ray.
Although the bone scan is also very sensitive for skeletal infection, in the spine
and sacroiliac joints it may lag behind the gallium scan for several days. Therefore, if
12 infectious spondylitis-diskitis is strongly suspected on clinical grounds, a normal
bone scan should still be followed by a gallium scan.
Can nuclear medicine methods determine when it is safe to stop
antibiotic therapy?
Since nuclear medicine methods are designed to identify sites of active
inflammation, whether acute or chronic, absence of inflammation can be a very
useful tool in determining when antimicrobial therapy should be discontinued. For
example, it is often difficult to know when to stop treating chronic diabetic
osteomyelitis or a prosthetic joint infection that has been managed with several
weeks or months of antibiotics. Some markers that can be used to monitor progress
include acute phase reactants such as erythrocyte sedimentation rate and repeat
radiographs to look for improvement in anatomical defects. However, nuclear
medicine methods can add a level of reassurance by confirming the absence of
inflammatory activity, thus aiding in the decision to discontinue antibiotic therapy.
Inflammatory Disorders 259
Figure 17. Infected pancreatitis with fistulous internal communication to the gut
imaged with 111In-leukocytes. The early images (left) show intense accumulation in
the pancreas (arrow). By 24 hours (right), activity now is clearly evident in the
cecum and ascending colon (arrowheads).
Additional Reading
1. Becker W. The contribution of nuclear medicine to the patient with infection. Eur
J Nuclear Medicine 1995; 22(10):1195-1211.
Reviews the basic pathophysiology of inflammatory imaging.
2. Datz FL. Abdominal abscess detection: Gallium, 111In-, and 99mTc-labeled
leukocytes, and polycloncal and monoclonal antibodies. Seminars in Nuclear
Medicine 1996; Vol XXVI, No. 1:51-64.
Reviews abdominal imaging, including mechanisms of uptake, labeling procedures, 12
normal patterns and artifacts.
3. Datz FL. Infection imaging. Seminars in Nuclear Medicine 1994; Vol XXIV, No.
2.
This volume covers the full range of nuclear techniques, including mechanisms of uptake,
labeling procedures, normal patterns and artifacts. There are individual articles on
111In- leukocytes, 99mTc-leukocytes, gallium, antibodies and chemotactic peptides.
4. Mandell GL, Bennett JE, Dolin R. Principles and practice of infectious diseases,
Fifth Ed. Philadelphia: Churchill Livingstone, 2000.
One of the classics on infectious diseases and well worth the trouble of consulting for
both common and exotic disorders. Detailed discussion on the clinical approach to
undiagnosed fever.
5. Peters AM. Localising the cause of an undiagnosed fever. Eur J Nucl Med 1996;
23(3):239-242.
An excellent discussion of the differences between out-patient and nosocomial fever and
how these influence differential diagnosis and testing.
6. Rennen HJ, Boerman OC, Oyen WJ, Corstens FH. Imaging infection/
inflammation in the new millennium. Eur J Nucl Med 2001; 28(2):241-252.
A survey of different nuclear medicine approaches to the imaging of inflammatory diseases:
past, present and future.
CHAPTER 13
Thyroid Disorders
Albert A. Driedger and Thomas J. McDonald
Thyroid Anatomy and Physiology
Embryology and Anatomy
The normal thyroid gland begins as a primitive diverticulum at the base of the
tongue in the third week of gestation and grows caudally toward its ultimate position
anterior to the thyroid cartilage. In the adult it forms a butterfly-shaped structure
weighing about 20 grams. Some thyroid tissue may remain along the thyroglossal
duct to form a pyramidal lobe or a thyroglossal thyroid remnant in post-natal life. In
rare cases the thyroid fails to descend and forms at the base of the tongue (lingual
thyroid) where it may present as a mass. The thyroid originates near the formation
site of the aortic sac, and for this reason accessory thyroid tissue may be found in the
mediastinum.
The normal thyroid is made up predominantly of follicular cells whose function
is the production, storage and secretion of thyroid hormone. These cells are able to
trap iodine and produce thyroxine by about the 11th week of gestation. The follicular
cells, possibly under the influence of TSH, arrange themselves to form follicles into
which they secrete thyroglobulin. A second population of cells, known as parafollicular
or C cells, is derived from the neural crest. These appear about midgestation and are
responsible for the secretion of calcitonin. The C cells do not form follicles and their
secretory polarity is oriented toward the capillaries.
There are relatively few lymphatics in the thyroid compared to capillaries and
these are concentrated in the pericapsular areas. Thyroidal lymphatics drain to nodes
on the thyroid, trachea, larynx and in the tracheo-esophageal groove as well as to the
deep cervical nodes along the internal jugular vein. Drainage to nodes in the upper
mediastinum is also common (Fig. 1).
Normal Physiology and Metabolism
The follicular cells reversibly take up iodide ions from the bloodstream against a
30-fold concentration gradient and irreversibly convert them to an organic form,
which is secreted into the follicles (Fig. 2). The steps of uptake and organification
are stimulated by thyroid stimulating hormone (TSH). Uptake is competitively
inhibited by large anions such as thiocyanate and perchlorate. Organification is
blocked by the thyroid-blocking drugs, propylthiouracil (PTU) and methimazole.
Oxidized iodine is secreted into the colloid where synthesis of mono- and
diiodotyrosines occurs, going on to the production of thyroxine (T4) and tri-
iodothyronine (T3). These steps are also blocked by anti-thyroid drugs. Within the
follicle, thyroid hormones are bound to thyroglobulin (Tg). The storage capacity of
the thyroid as an endocrine gland is exceptional: an average gland contains about 3
Figure 1. Anatomic distribution of lymph nodes that drain the thyroid gland. (From
Agur AMR. Grants Atlas of Anatomy, 9th Ed., 1991. Reproduced with permission
from Williams & Wilkins.)
13
months supply of hormone. The amount of iodine contained in a normal gland,
subject to an adequate iodine intake, is about 7-10mg. Release of hormone from the
follicle is stimulated by TSH and inhibited by iodide and lithium.
Tg is a large protein (660kd) that serves as the matrix for thyroid hormone
synthesis in the follicles and also as the storage vehicle. Tg synthesis is stimulated by
TSH via the cyclic AMP mechanism. Of the 134 tyrosine residues in the molecule,
no more than 15-17 are usually iodinated, including no more than 2-4 molecules of
T4 and T3. The release of the hormone seems to result in the destruction of the Tg
molecule, seemingly an inefficient process when the energy costs of synthesizing
such a large molecule are considered. Normally, Tg is largely retained within the
follicles, but it may be released into the blood stream in disease states. Tg is often
produced by neoplasms derived from follicular cells and is a useful marker for recur-
rence of tumor.
Production of thyroid hormone is regulated by the pituitary gland through the
secretion of TSH, which in turn is regulated by hypothalamic production of
262 Nuclear Medicine
Iodine-123
123
I has a half-life of 13.2 hours and a 159 kev gamma emission which, for external
imaging purposes, is nearly ideal. The radiation dose to the patient is low and the
emission is well suited for gamma camera detection. However, 123I is not available
everywhere and it is more expensive than alternatives such as 131I or 99mTc-
petechnetate.
Iodine-131
131I has a half-life of 8 days. It has a complex spectrum of both beta and gamma
emissions with a dominant gamma at 364 kev, a high energy requiring thick
collimation to exclude scatter and thick scintillation crystals for efficient detection.
The beta emissions contribute heavily to the thyroidal radiation dose, which may be
in the range of 0.5-1Gy from a diagnostic imaging study. The advantage of 131I is
that it is inexpensive and its half-life is long enough to permit keeping it in inventory
for use at short notice. One practical solution to the dose-price quandary is to perform
a radioactive iodine uptake (RAIU) with a small dose of 131I and use 99mTc-
petechnetate for thyroid imaging.
99m Tc-Pertechnetate
99m Tc-pertechnetate is a universally available and inexpensive radiopharmaceutical
whose half-life of 6 hours and monoenergetic gamma emission of 140 kev provide
high quality images with a low radiation dose. Pertechnetate follows iodide uptake
by follicular cells, but is not organified, and the uptake remains reversible. 13
Follicular Tumor Imaging
201 Tl-thallium chloride and 99mTc-sestamibi may be useful in rare situations where
needle biopsy has not provided a definitive diagnosis of a thyroid nodule and where
evidence is needed to support a decision for surgical excision. Thallium uptake and
washout has been characterized recently and is claimed to have better sensitivity and
specificity than fine needle aspiration. 99mTc-sestamibi is concentrated by many
neoplasms, not only those originating from the thyroid. This agent is better suited
for imaging than thallium: it is useful to image the post-surgical neck for evidence of
recurrent malignancy and as an adjunct to 131I imaging in the follow-up of treated
thyroid cancer patients.
18
F-fluorodeoxyglucose (FDG) is assuming increasing importance for the imaging
of suspected malignancies and for pre-operative staging. In relation to thyroid cancers,
FDG is concentrated by many tumors that do not concentrate 131I. The usefulness
264 Nuclear Medicine
of FDG lies in follow-up of patients with elevated serum thyroglobulin and negative
131
I scans.
Many follicular-derived cancers have somatostatin receptors on their cell surfaces
and can be imaged with radiolabelled somatostatin analogues such as 111In-
pentetreotide (octreotide, Octreoscan).
medullary carcinomas. For diagnostic purposes, the sensitivity of 30% is too low to
be very useful, but some symptomatic patients with MIBG uptake may be treatable
with 131I-MIBG. 201Tl-thallium and 99mTc-sestamibi are non-specific tumor agents
and, as with follicular tumors, may be helpful when other approaches have proven
unsuccessful.
Scanning of the Thyroid
Radionuclide imaging of the thyroid is a direct extension of the clinical
examination. It is important for the nuclear physician to be familiar with the setting
in which the examination is being requested. Patients may be self-medicating with
vitamins, kelp or other substances containing iodine and it is recommended to defer
the examination for several weeks after discontinuation of these substances (Table
1). Clinical examination of the patient while under the camera allows correlation of
palpable features with those of the scan. This ought always to be done by the physician
who will report the examination and who should be aware of the presenting
13 complaint, the relevant clinical history and laboratory data. A positive family history
will increase the pre-test likelihood for multinodular goitre and Gravess disease. A
complaint of pain predisposes toward thyroiditis. A history of radiation exposure
with a symptomatic mass increases the probability of a malignancy.
Thyroid imaging is best performed on a camera fitted with a pinhole collimator
in order to achieve the highest possible spatial resolution. The disadvantages of pinhole
collimation are those of image distortion in depth caused by parallax and of the low
sensitivity. Parallax increases as one moves from the center of the field to the periphery
and will distort an off-centered image. Externally placed anatomical markers may
seem misplaced on account of parallax.
The uptake probe is a nonimaging device that has high sensitivity for counting
radiation and is routinely used to measure RAIU by the thyroid. The uptake
measurement requires calibration of the probe with the patient dose. When the
uptake has stabilized (about 24 hours later) the patients neck is counted and the
percent uptake is calculated with a correction for decay during the interval. In North
Thyroid Disorders 265
Graves disease may appear at any age but occurs most often in females aged 20-
50 years. The presenting signs and symptoms are secondary to: an increased metabolic
rate, increased adrenergic activity and manifestations of infiltrative dermopathy and
ophthalmopathy. Patients may present with complaints of heat intolerance, weight
loss in spite of an increased appetite, loose stools, tachycardia, tremor, emotional
lability, fatigue and insomnia. Children may present with hyperactivity. In contrast,
elderly patients may appear depressed and apathetic, having withdrawn from
customary activities.
On examination, untreated patients may appear nervous and hyperactive or
fatigued and listless. With Graves disease, there is commonly mild to severe diffuse
thyroid enlargement of the gland which has been described as having a beefy
consistency. Bruits may be audible over the gland, the skin is warm and moist. There
may be tremor of the outstretched hands and tachycardia is commonly observed.
Thyrotoxic stare, lid retraction and lid lag may be present. Clinical signs of infiltrative
ophthalmopathy are seen in fewer than one-half of patients on presentation. Pretibial
myxedema and thyroid acropachy are relatively rare findings. Treatment with beta
blocking drugs may abolish the signs of adrenergic overactivity (eg. tremor,
13 tachycardia, stare and lid lag). The use of beta blockers does not compromise
diagnostic testing with radionuclides.
In Figure 3 the 99mTc-pertechnetate thyroid scans of common thyroid disorders,
including a typical patient with Graves disease, are displayed. On occasion, patients
with other forms of chronic thyroid disease may develop TSH receptor antibodies
producing a thyrotoxic state. The distinction is not very important except that the
latter may be at lower risk of developing eye complications. In these cases, the
radionuclide pattern may be somewhat patchy rather than the diffuse homogenous
pattern of uptake seen with Graves disease.
unexpected illness together with the presence of a goitre requires the exclusion of
thyrotoxicosis.
The thyroid scan in a toxic multinodular goitre has an irregular distribution of
hot and cold regions and the RAIU may be normal or only mildly elevated. In
patients who have iodine-induced thyrotoxicosis, the uptake may be low initially
13
but will increase on serial uptakes once the source of iodine is eliminated. With
sufficient uptake, radioiodine therapy provides an effective treatment mode.
Toxic Adenoma
Autonomously functioning nodules are benign adenomas, which operate
independently of the thyroid-pituitary feedback control mechanisms. Recent evidence
suggests that a number of these adenomas may result from somatic mutations
occurring in the TSH receptor, resulting in constitutive activation of the receptor
and autonomous cellular function. Thyrotoxicosis results when the output of thyroid
hormone from such a nodule exceeds the bodys requirements. As the symptoms of
thyrotoxicosis are often mild, an initial work up may, in certain cases, begin with the
exclusion of malignancy by a fine needle aspiration biopsy of a clinically detected
nodule. The finding of a low serum TSH level in association with elevated thyroid
hormone levels should be followed by an 123I or 99mTc-pertechnetate scan to confirm
increased uptake occurring in the nodule. In this setting, a preference for radioiodine
268 Nuclear Medicine
scans has been expressed by some authors because of the observation that some
malignancies concentrate pertechnetate but fail to concentrate iodine. In our view,
an aspiration biopsy, and not a radionuclide scan, is the better arbiter of the presence
or absence of malignancy; hence, there is little need to prefer one radionuclide over
the other for imaging.
Subacute Thyroiditis
13 Subacute thyroiditis (or de Quervains disease) is a relatively uncommon form of
thyroiditis characterized pathologically by the presence of extensive follicular cell
destruction, extravasation of colloid and aggregation of histiocytes around colloid,
coalescing into giant cells. There is strong but indirect evidence that this pathology
may result as a reaction to a preceding viral infection, but immune mechanisms may
also play a role. Patients often present with cervical pain, tenderness of the thyroid
(at times exquisite), symptoms of thyrotoxicosis and in some cases, systemic symptoms
of an inflammatory illness. Thyrotoxicosis results from an unregulated release of
preformed thyroid hormone from an inflamed gland. As the thyroid gland is depleted
of preformed hormone and the thyroiditis subsides, the serum T4 and T3
concentrations fall to normal and in some cases to subnormal levels, before recovery
of the gland with normalization of function. The entire course of the illness may last
as long as 6-9 months, at times longer.
Thyroid Disorders 269
Silent Thyroiditis
This form of thyroiditis is dominated by the pathological picture of a prominent
lymphocytic infiltration of the gland. The evidence to date suggests that silent
thyroiditis is a variant of lymphocytic thyroiditis. Both have a predilection to occur
in the post-partum period but are also common in the non-pregnant state. In contrast
to subacute thyroiditis, the clinical features are usually (but not invariably) milder.
There is usually less or no pain in the cervical region and less or no thyroid tenderness,
despite, at times, acute enlargement of the thyroid. There are minimal if any systemic
symptoms and the thyrotoxic symptoms, if present, may be quite mild. The
subsequent evolution in the thyroid biochemical abnormalities has a similar clinical
course to that of subacute thyroiditis. However, in contrast to subacute thyroiditis,
patients with silent thyroiditis often have recurrences and up to one-half of the
patients may develop permanent hypothyroidism in the future.
Factitious Hyperthyroidism
The clinical history is usually sufficient for the diagnosis of excessive exogenous
thyroid hormone ingestion. Occasionally this may be deliberate and surreptitious.
Thyroid Nodules
Goitres occur both endemically in regions characterized by iodine deficiency as
well as sporadically in all populations. Common to all goitrogenic processes is an
increase in the number of follicles. A number of factors, of which TSH is a major
contributor, can stimulate growth of new follicles. Many patients with nontoxic
goitres have thyroid-stimulating antibodies similar to those seen in Graves disease.
It is possible that at the outset of goitrogenesis the process is always, or nearly always,
diffuse. The newly generated follicular cells may display heterogeneous growth
potential and nodules develop from the most rapidly growing clones. The process is
usually slow, requiring years or decades. The properties of the original cells, which
give rise to nodules, tend to be maintained in their daughter cells: thus, individual
nodules in a multinodular goitre will display distinct behaviours. It is this
morphological heterogeneity that gives rise to the presence of hot, warm and cold
nodules coexisting in the same goitrous gland. Secondary events, such as central
necrosis in a rapidly growing nodule or loss of functional capacity over time, may
also contribute to the conversion of functioning adenomas into cold nodules.
The incidence of goitre increases with age and the reported rate of increase is a
function of how the gland is examined. Thyroid enlargement most often comes to
light through direct observation or palpation. The use of ultrasound imaging has
greatly facilitated the characterization of goitres and its use permits a confident
determination of whether the gland is diffusely enlarged or contains one or more
nodules. RAIU and imaging can play a useful supplementary role to determine the
functional status of goitres and of palpable nodules. However, nuclear studies no
longer play the major role in goitre evaluation that they did prior to the availability
of high-resolution ultrasound imaging.
Nodules that are large and that are enlarging rapidly, or that have other suspicious
features such as fine calcification, must be evaluated to exclude the possibility of
malignancy. The primary means of doing this is with needle aspiration biopsy,
preferably ultrasound-guided. In the event that several repeated needle biopsy attempts
are non-diagnostic, radionuclide imaging should be used to determine the functional
status of suspicious nodules. Nodules that concentrate iodine or pertechnetate are
13 very rarely malignant and can be safely followed while cold nodules that are suspicious
of malignancy should be resected. The approach to the investigation of a thyroid
nodule is given in Figure 4.
Thyroid Cancer
Although thyroid cancer accounts for less than 1% of all clinical cancers, it is the
commonest form of endocrine malignancy and accounts for more deaths than all
other endocrine cancers combined, excluding cancers of the reproductive tract.
Thyroid neoplasms may arise from either the follicular endothelium or from the C
cells (Table 3). The parafollicular C cells give rise to medullary carcinomas. Rarely,
the thyroid may be the site of origin of a lymphoma or sarcoma. Primary tumors of
kidney, breast, lung and melanomas can metastasize to the thyroid gland.
Follicular-Derived Thyroid Cancers
In North America the incidence of differentiated thyroid cancers has been
increasing throughout much of the 20th century, certainly since the 1930s. However,
Thyroid Disorders 271
the mortality rate from thyroid cancers has remained stable, a reflection of improved
therapies. Women are affected about three times more often than men. Unlike 13
medullary carcinomas, which may be sporadic or occur in families, the follicular-
derived neoplasms are said to occur mainly sporadically. Nonetheless, familial
groupings of papillary cancer do occur with greater than chance frequency. There is
a familial predisposition to thyroid cancer in Gardners syndrome (multiple colonic
and rectal polyposis) and in Cowdens disease (multiple hamartoma syndrome).
The only environmental factor known to increase the incidence of thyroid cancers
is exposure to ionizing radiation. In the early part of the 20th century, x-rays were
used to treat facial and neck hemangiomas of infants and lymphoid hyperplasia
(tonsils, adenoids and thymus) of young children. In the following decades there
was an increased number of thyroid cancers in these patients, which began to appear
about a decade after the exposure, reached a peak incidence at about 25 years and
continued to appear at greater than the sporadic population incidence throughout
the lives of these people. In the post WWII years, the state of Israel used x-irradiation
to treat immigrant children for scalp ringworm. These children developed an excess
number of thyroid cancers in adult life. The experience of radioactive fallout, with a
272 Nuclear Medicine
Parafollicular-cell derived
medullary carcinoma of the thyroid 5-10%
Nonepithelial tumors:
Lymphoma <1%
Sarcoma
Epidermoid carcinoma
Metastatic <1%
Adapted from: Biddinger P and Nikiforov YE. Pathologic features of thyroid tumors.
In: Thyroid Cancer. Fagin JA (ed.); Kluwer Academic Publishers. 1998.
of MTC will take up MIBG, though uptake is highly specific for tumors with an
active amine uptake mechanism. This radiopharmaceutical may serve a dual purpose
as about 85-90% of associated pheochromocytomas will also take up MIBG. Many
MTC have somatostatin receptors and can be imaged with 111In-octreotide or other
somatostatin radiopharmaceutical analogs (Fig. 5). Metastases in or near the liver,
spleen or kidneys may sometimes be difficult to image on account of the normal
high concentration of somatostatin receptors in these organs.
Frequently Asked Questions (FAQs)
Can the thyroid scan be used to follow the size of a goitre or
nodule?
That would not be the best use of the technology: ultrasound is better. The size
of nodules is possibly less important than their cytology. If you need to know what
a nodule is about, biopsy it.
Does everybody with thyrotoxicosis need a thyroid scan and
RAIU?
Not if you already know what you are dealing with or if RAI treatment is out of
the question. For instance, a pregnant thyrotoxic woman would be managed without
recourse to these tests. These tests serve to distinguish among the forms of
hyperthyroxinemia and, in thyrotoxic patients, to select the therapy dose.
Why isnt the thyroid scan a first-line test for thyroid nodules
any longer?
The thyroid scan is able to determine the functional status of nodules that are
larger than about 1.5 cm. It lacks the ability to exclude malignancy in cold nodules.
A fine needle aspiration biopsy does that.
Additional Reading
1. Cases JA, Jurks MI. The changing role of scintigraphy in the evaluation of thyroid
nodules. Seminars in Nuclear Medicine 2000; 30:81-87.
This review article develops the current algorithms for the investigation of thyroid nodules. 13
The roles of the laboratory, biopsy, radionuclide and ultrasound scanning and surgery
are reviewed.
2. Mazzaferri E. Management of a solitary nodule. New Engl J Med 1993;
328:553-559.
This short review article illustrates an efficient approach to the characterization of
thyroid nodules.
3. Berne RM, Levy MN, eds. The thyroid gland. In: Physiology. Mosby Year Book,
1993:932-948.
This is a very readable account of thyroid gland physiology.
4. Tuttle RM, Becker DV. The Chernobyl accident and its consequences: Update at
the millenium. Seminars in Nuclear Medicine 2000; 30:133-140.
This is a review of the state of knowledge concerning radiation-induced thyroid cancers
and, especially, of what we learned from the Chernobyl accident.
5. Braverman LE, Utiger RD, eds. The Thyroid. Eighth ed. Philadelphia: Lippincott
Williams & Wilkins, 2000.
A comprehensive, heavily referenced work concerning thyroid diseases.
CHAPTER 14
Radionuclide therapy
of Thyroid Disorders
Albert A. Driedger and Thomas J. McDonald
Introduction
Of the diseases discussed in Chapter 13, some are treatable with radioactive
iodine (RAI). In diagnostic applications it is optimal that as little energy as possible
be deposited in tissue so that the signal deposited in the detector can be maximized.
That is why diagnostic imaging utilizes gamma-emitting isotopes. However, in therapy
the intent is to deposit as much energy as possible in a target tissue in order to kill
those cells. For this application, nuclides whose emissions have a short path length
in tissue are essential. Historically, these have been beta emitters with path lengths
of a few mm in soft tissue. The most frequently used therapeutic nuclide of this class
is 131I even though its associated gamma emissions give rise to problems in radiation
protection. In future it is likely that Auger electron and alpha-emitting nuclides will
also find applications in therapy. With these emissions the energy of the decay will
be largely concentrated in the individual cells that concentrated the isotope.
Benign Thyroid Disorders
Indications for Radioiodine therapy
Graves Disease
All treatments for Graves disease disable or destroy the thyroid by one means or
another as we have no means to repair the intrinsic immunologic abnormality leading
to thyrotoxicosis. The three therapeutic modalities are partial thyroidectomy, long
term antithyroid drugs or RAI.
Surgery is rarely used in western practice but may be the preferred treatment in
a few instances such as:
1. thyrotoxicosis during pregnancy or amiodarone-induced thyrotoxicosis
where there is intolerance of antithyroid medications
2. serious adverse reaction to antithyroid medications in patients who refuse
RAI
3. concern about a co-existing malignancy.
Antithyroid medications, such as propylthiouracil or methimazole, will achieve
remission in some patients and are the preferred treatments during pregnancy. It is
the view of some endocrinologists, especially in Europe, that all patients ought to
have a trial on antithyroid drugs and that RAI should be reserved for those who do
not respond or who develop side effects. These drugs are associated with adverse
effects, which may range from skin rashes to reversible agranulocytosis and hepatitis.
Elderly patients with cardiovascular disease are best treated with a course of antithy-
roid medications to deplete the gland of thyroid hormone prior to receiving RAI.
Without this precaution there is a possiblity of an abrupt symptom exacerbation as
hormone is released from the damaged follicles. Rarely thyrotoxic patients are treated
long term with antithyroid medications alone.
In the western world, RAI is a common treatment of choice, even in the young
and in those with small goitres, although there are differences of opinion about the
goals of therapy. Many endocrinologists prefer ablation of the thyroid while some
attempt to minimize the likelihood of ensuing hypothyroidism by use of smaller 131I
doses. The attempt to avoid hypothyroidism may be futile, since immune thyroiditis
to some degree is a component of every case of Graves disease. The immunologic
assault on the thyroid will continue despite RAI therapy and may ultimately lead to
hypothyroidism when the residual follicles cease to function. Euthyroidism, if it is
achieved, is unlikely to be permanent. It is a commonly held view that, for most
patients, ablation is the proper endpoint of RAI therapy.
The only absolute contra-indication to RAI therapy is pregnancy as it may ablate
the fetal thyroid gland and could contribute to mental retardation if given during
the first trimester when cortical cell migration in the fetal brain is very sensitive to
radiation. RAI should also be deferred in a breast-feeding patient, as iodine is secreted
into breast milk. In all other patients RAI can be considered. The only proven long
term risk is that of permanent hypothyroidism. There is no increased risk of cancer
or leukemia following RAI therapy nor is there any risk demonstrable to children of
subsequent conceptions. It is usually recommended that conception be delayed for
6-12 months post-therapy by which time a stable euthyroid state ought to have
been achieved. There is no absolute contraindication to the treatment of thyrotoxic
adolescents and children with RAI.
In the short term, 131I administration may result in a large release of thyroid
hormone and, rarely, may produce thyroid storm unless precautions are taken. The
precautions can be achieved either through pre-treatment for a few weeks with
antithyroid drugs or by the concomittant use of beta blockers. Patients who have
thyrotoxic ophthalmopathy may experience transient worsening of their eye problems
in the weeks following treatment and this can be avoided by administration of a
short course of high dose steroids.
from nodule to nodule secondary to the variable uptake. The under-treated nodules
tend to become more autonomous over time, with possible recurrence of the toxic
state. These patients should be made aware that treatment may need to be repeated.
Treatment with larger doses of 131I will help to minimize the risk of recurrence.
Autonomous Nodules
Autonomous single nodules may occur in young adults but do so more commonly
in older people. Definitive treatment is achieved with either surgery or RAI. With
RAI therapy there is little risk of ensuing hypothyroidism as the function of normal
tissue is suppressed at the time of treatment and it will recover when the thyroid
stimulating hormone (TSH) levels subsequently normalize. There is no direct
evidence, but an argument can be made for a theoretical risk of carcinogenesis from
RAI therapy in this setting as the normal tissue adjacent to the nodule will receive a
significant, but non-lethal, radiation dose. Thus, it is common practice to resect
these nodules from younger people and to reserve RAI for older patients in whom
surgery carries a greater risk.
Non-Toxic Goitres
Non-toxic goitres with compressive symptoms are, in the main, best treated by
surgery. However, in circumstances where surgery is risky, reduction of goitre size
can be achieved with RAI. In the past it was thought that RAI treatment might be
dangerous if the goitre was compressing the trachea as the induction of swelling by
the action of radiation might compromise the airway. However, evidence has emerged
that significant airway compression does not occur and that RAI therapy is safe even
if the pretherapy CT scan demonstrates up to 50% narrowing of the trachea.
The clinical assessments of large goitres systematically underestimate their size
and this is a factor responsible for failure of RAI therapy. An ultrasound examination
defines the gland size better than palpation. If there is a large intrathoracic goitre
then CT may be the preferred way to determine gland size. The doses of 131I often
need to be quite large, up to the range of doses used in cancer therapy, as the RAI
uptake tends to be low. Pretreatment of the patients with a low-iodine diet for 10-14
days tends to increase the uptake and permit reduction of the dose. The former
practice of restricting outpatient doses to 1.11 GBq (30 mCi) or less may have
promoted a practice of under-treating large goitres of this type if there were no
14 hospital beds available for isolation purposes. In these circumstances, it was common
practice to administer the maximum outpatient dose without regard for the fact
that this might not achieve a cytocidal concentration in any part of a very large
goitre.
Preparing the Patient for Therapy
As with all other therapies, it is critical that the patient be informed and able to
consent to the proposed administration of RAI. The prevailing high public level of
negativity about nuclear technology, combined with the difficulty that thyrotoxic
people may have in focusing their thoughts, can make it difficult to obtain an
informed consent. In such cases it may be best to institute a short course of antithyroid
medication to provide a hesitant patient some time to consider the treatment options.
At a minimum, patients need to understand the nature of their illness, its natural
course, the proposed treatment and its probable consequences. They also need to
Radionuclide therapy of Thyroid Disorders 279
understand that although RAI treatment is safe for them, there will be a period
during which they must take care to avoid radiation exposure of their closest family,
social and workplace contacts. Women of childbearing age and who are sexually
active need beta-HCG documentation to ensure that one does not treat a pregnant
woman.
Methimazole and propylthiouracil should be discontinued for 3-4 days prior to
administration of 131I. If they are not discontinued beforehand, the treatment will
be less effective as the radioiodine will not be organified and will be rapidly excreted.
Most often these drugs need not be reinstituted. If required to control symptoms
then they should be withheld for 5-7 days. They may exert an unwanted
radioprotective effect on the thyroid since the thionamide drugs are free radical
scavengers. PTU given prior to therapy has a radioprotective effect on the gland that
may last up to a month. Thyrotoxic recurrences are more common in PTU pre-
treated people and one way to minimize recurrences is to increase the 131I dose by
10-15%.
It is important to provide all patients, but thyrotoxic ones in particular, with
written instructions about radiation safety and the protection of their families and
close associates in the days or weeks following treatment (Table 1). The International
Commission on Radiological Protection (ICRP) has recommended that members
of the public should receive no more than 1mSv of ionizing radiation exposure per
year from man-made sources, including incidental contact with recently treated
patients. The ICRP has also declared that non-pregnant adult family members and
close associates of the patient who are involved in provision of care and comfort may
be allowed a larger exposure, provided that they are informed and do so knowingly
and willingly. In order to advise the patient and family appropriately, it is necessary
to answer the questions posed in Table 2.
In the main, outpatient treatment of thyroid diseases with RAI is safe. Patient
and family must understand the restriction on intimate social behaviours in the days
following treatment. The physician can most efficiently provide them with behavioral
advice while describing the risks and benefits of treatment and can thereby alleviate
everyones anxiety about the significance of low dose radiation exposure. There are
several software products that can be used to model the likely exposure of others
from RAI and other radionuclides. (Once such program is available without charge
from rajc@flinders.edu.au.) They do not take the possibility of contamination into
account but this is of little consequence as about 99% of the dose to others will be 14
gamma radiation. These software programs take dose, uptake and the projected
proximity of contact into account and can be used to support the instructions given
to patients.
There are circumstances in which the assurance of radiation protection may be
challenging, such as in the cases of very active young children, mothers with
responsibility for the care of young children, intellectually challenged adults or the
frail elderly. In these complex situations, management requires coordination of a
multidisciplinary team whose skills include technical expertise as well as a variable
range of social resources. It is essential for nuclear physicians to educate families,
nurses, social workers and others about the salient aspects of radiation safety when
such situations arise. It should always be possible to provide all the care that the
280 Nuclear Medicine
The radioactive medicine that you have received will expose others to radiation
and a small part of the radioactive medicine can potentially be transferred to other
people through close contact. The radiation dose to them will be low but, by
observing the following precautions, you can reduce it still further.
patient needs without undue exposure of the staff if attention is paid to the time-
distance rule and to the minimization of contamination.
The administration of a large therapy dose is not in itself an indication for ad-
mission to hospital. In-hospital therapy may be necessary for a patient who requires
concurrent medical care, comes from a long distance by public transportation or is
unable to comply with instructions. The patient must be admitted to a single room
with a non-absorbent floor covering and equipped with a private shower and toilet.
If the facility is not designed for radiopharmaceutical therapy, then care must be
taken to avoid exposure of patients in adjacent rooms. Adult visitors should be
restricted to brief visits (15-30 minutes). It is important that patients not be made to
feel that they are locked up: the door to the room may be open, the room should
be cleaned as needed and staff should not feel so pressed to leave the room as to have
no time for social interaction with the patient.
Discharge may occur when the patient no longer needs concurrent medical care
and when it can be reasonably assumed that no member of the public will receive a
dose in excess of the public limit from the patient.
In many countries there are national advisory committees that have developed
useful documents to promote quality assurance and safety in radionuclide therapies.
These are often up to date and should be consulted in the preparation of a local
therapy protocol.
Dosimetry 14
The dose of 131I is derived from consideration of the gland size, the iodine uptake
and the intrathyroidal transit time. This is easier to do with diffuse Graves disease
than with nodular glands whose uptake is irregular. Gland size is usually estimated
by palpation, although when a retrosternal goitre is suspected, ultrasound or CT
examination may be required. Iodine uptake is measured with 131I or 123I. The
intrathyroidal transit time is not routinely measured but may be assumed to be
about 4 days. In some thyrotoxics the transit time may be much shorter and this will
result in underdosing of those patients unless there is appropriate dose compensation.
To achieve thyroid ablation requires administration of a radiation dose of 70-100
Gy to the gland. Thus, in principle, the 131I dose may be calculated as:
282 Nuclear Medicine
Table 3. Simplified low-iodine diet typically used for 10-14 days prior
to radioiodine scanning or therapy for thyroid cancer
These are high iodine foods and should be avoided.
Table salt or sea salt (do NOT use salt in food preparation or at the table)
All fish, sea food and shellfish
Potato chips, pizza, salted nuts and other heavily salted foods
Tea and instant coffee
Canned soups, canned vegetables/fruits and canned vegetable juices (unless
unsalted)
Milk and other dairy products (see below)
Eggs (maximum one per day) (see below)
Foods or beverages containing red food coloring, chocolate or molasses
Some medicines (multiple vitamins, Buckleys mixture, cod liver oil, amiodarone)
such as AGES (Age, Grade, Metastasis and Size) and TNM (Tumor, Node, Metasta-
sis). There is no universal agreement about which scheme is preferable. Clinical
validation for the AMES variables has come from follow up of large series, with
recurrence rates ranging from 6.8% in stage I to 50% in stage IV (Mazzaferri criteria).
For purposes of illustration, the AMES criteria will be discussed here.
A Adults older than 45 years of age tend to have disease with more aggressive
behaviour. Older patients should all be considered to be at increased risk of recurrence,
regardless of tumor characteristics.
M The most common sites of metastases from papillary carcinoma are the regional
lymph nodes and the lungs. The presence of lymph node involvement of itself has
little effect on the prognosis as these are usually easy to treat by resection and RAI.
However, bilateral neck or mediastinal lymph node disease may adversely affect the 14
prognosis. Lung metastases signal a high risk for disease persistence/recurrence.
Hematogenous dissemination occurs rarely, typically to bone and brain, and is
associated with a grim prognosis. Follicular carcinoma, on the other hand, tends to
metastasize via the hematogenous route with involvement of bones, lungs, brain
and other viscera, yet metastatic disease may be compatible with prolonged survival.
Accurate preoperative detection of distant metastases is typically not possible
with these cancers underscoring the importance of RAI imaging to detect extent of
metastatic disease and the difficulty inherent in imaging them while the thyroid
gland is in place. Thus, the definitive discussion of prognosis must often await ablation
of remnants with RAI and post-therapy scanning.
E The local extent of disease includes consideration of extrathyroidal extension
of tumor into adjacent tissues, vascular invasion, the finding of tumor at the resection
284 Nuclear Medicine
margin and lymph node involvement. In the case of papillary carcinoma, multifocality
ought to be included as a risk factor, especially if the gland resection has been
incomplete.
S The probability of disease recurrence rises with increasing size of the primary
tumor and may reach 40% for lesions greater than 4 cm. Small subcentimetre lesions
discovered incidental to other surgery probably do not require ablation, but there is
still a small risk of metastatic disease arising even from 2-3mm primary tumors.
Multicentricity, a history of exposure to ionizing radiation (e.g., treatment for
Hodgkins disease) or aggressive-appearing histology may be risk factors that would
justify total thyroidectomy and RAI ablation, even with a small primary lesion.
Patients who are negative on all parameters of the AMES assessment can be
considered as being at low risk (~5%) for future recurrence. In contrast, those who
have major risk factors can have up to a 50% recurrence rate. Regardless of the
grading, all patients with previous thyroid cancer should have lifelong follow up as
recurrences may occur even decades after the initial diagnosis.
Indications for RAI Therapy
RAI treatment may be considered under two indications:
1. ablation of thyroid remnants and
2. treatment of iodine-avid metastases.
Prerequisites for RAI treatment are that all necessary surgery has been completed
and there are no urgent clinical issues that mandate immediate intervention (e.g.,
airway obstruction, spinal cord compression or symptomatic brain metastases). It is
in the interest of patients that there be a smooth working relationship between
surgeons, endocrinologists and nuclear physicians to ensure that all clinical issues
are addressed in an efficient manner.
The nuclear physician should ideally meet the patient after surgery but prior to
discharge from hospital to discuss the purpose of radioiodine therapy and to prepare
the patient for the ensuing temporary hypothyroid symptoms. A timing of about 5-
6 weeks post-surgery works out well in most cases. There are several reasons for not
proceeding more quickly to ablation:
it allows time for patient recovery from surgery prior to RAI treatment
endogenous levels of iodine will be high for several weeks after surgery as
a result of transcutaneous absorption of iodine from the pre-operative
skin preparation
it will take up to 5 weeks to elevate TSH levels adequately (greater than
30 mU/L) as the circulating half-time of thyroxine is about one week
a low iodine diet for the last 10-14 days will help to maximize the iodine
uptake when the treatment is given (Table 3).
Not all patients will become symptomatically hypothyroid on this protocol as
they will be supported to variable degrees by the remnants of normal thyroid, which
nearly always remain as a result of the surgeons attempts to spare the parathyroid
glands. If the patient has a large remnant with a high uptake, it may prevent an
adequate rise of TSH and obscure the presence of metastases in the post-ablation
scan. At the other extreme are some patients for whom hypothyroidism may be
intolerable by virtue of their frail medical state. There have been many attempts to
minimize the symptoms of hypothyroidism by shortening the hypothyroid interval
as much as possible. While the most common protocol is to withhold all forms of
thyroid hormone for 5-6 weeks, some centers replace the patient with short-acting
liothyronine (T3, Cytomel) for three weeks followed by 2-3 weeks off all thyroid
hormone. This protocol may not always elevate TSH to the same degree as complete
withdrawal and liothyronine may not be readily available in some countries.
In some cases, it may be unwise to make frail patients hypothyroid. In addition,
rare patients may have diminished pituitary function and be unable to raise their
TSH levels in response to hypothyroidism. With recombinant human TSH
(Thyrogen) it is possible to treat such patients while on full dose thyroid hormone
replacement following several daily intramuscular injections of rh-TSH (Fig. 2). (At
the time of this writing Thyrogen is only approved for diagnostic scanning but can
be obtained for treatment on a case by case basis from Genzyme Corporation.)
14 A diagnostic scan prior to therapy is helpful in several ways, provided that therapy
will follow soon after so as to minimize the potential for stunning (see section on
Dosimetry). The uptake at 48 hours is useful to estimate the retention of the treatment
dose and to support discussions with the patient about the duration of precautions
when outpatient treatment is given. If the pathology is that of a single small lesion
without aggressive characteristics and limited to one lobe in a young person along
with an undetectable thyroglobulin (Tg) and the uptake is nil, then ablation arguably
may be abandoned and the patient placed on long term follow up. Those with an
elevated Tg and a negative scan need a therapy dose and post-therapy imaging.
One of the most important indicators of residual cancer is the Tg level and this
is most sensitive if performed when the patient is hypothyroid. It is important to
also measure antithyroglobulin antibodies at the same time as their presence may
suppress the Tg value. Further, in follow up, a rise in anti-Tg titre may indicate
disease recurrence.
Radionuclide therapy of Thyroid Disorders 287
Figure 2. A frail elderly man with congestive heart failure and metastatic papillary
carcinoma who became paraplegic during thyroid hormone withdrawal. He
recovered on thyroid hormone. Subsequently, while on thyroid hormone
replacement he was given recombinant thyroid stimulating hormone (rh-TSH) prior
to 131I administration. His post-therapy scan performed a week after treatment with
3.7 GBq of 131I demonstrates uptake into multiple pulmonary metastases.
For those patients with recurrent but iodine-negative cancer and for those with
metastatic medullary carcinoma, there is an emerging prospect of new therapies
using somatostatin receptor-avid agents, such as 90Y-octreotide or 90Y-lanreotide. 14
These therapies are experimental at this time.
Dosimetry
In recent years there has been concern expressed over the possibility of thyroid
stunning by diagnostic doses resulting in less 131I uptake from the treatment dose
than was anticipated on the diagnostic scan. This phenomenon is attributed to
sublethal radiation damage by the diagnostic dose, which could prejudice the outcome
of subsequent therapy. To avoid stunning, some centers have abandoned pretherapy
scanning, choosing to base the decision to treat on other criteria and proceed directly
to administration of the therapy dose. On closer examination, it seems that the
stunning is observed when there is a prolonged interval (up to several weeks) between
the diagnostic and therapeutic administrations. In centers where the interval between
the two procedures is minimized, stunning is not observed. With coordination, it is
288 Nuclear Medicine
possible to complete any desired RAI imaging and to proceed to therapy within 48-
72 hours. For patients presenting for initial ablation, the imaging and uptake
measurements of the remnants can be done with 50-70 MBq of 131I. If the patient is
presenting for consideration of retreatment of known metastases, then higher scanning
doses up to 350 MBq may be used without prejudicing an immediately following
therapy. Metastases typically have much lower 131I uptake per gram than normal
remnants.
Unlike external beam therapy, where administered doses can be accurately
calculated, radionuclide therapies do not lend themselves to similar dosimetric
precision. The goal of therapy is to ablate normal thyroid tissue and tumor while
avoiding toxicity to the bone marrow and lung. There are two possible approaches
to dosing:
use of a standard dose or
pretreatment dosimetry study with a low 131I dose to approximate target
organ and remnant or tumor radiation doses.
The outcomes are similar between the two approaches, although there are
circumstances, such as concurrent renal failure, in which accurate dosimetry is
necessary to avoid toxicity within the customary range of doses.
The objectives of RAI ablation remain a vexing source of controversy. If the
objective is only to reduce the volume of normal tissue then ablation is possible with
a low dose of 131I e.g., 1.1 GBq. However, if one wishes to truly ablate the remnants
with a single administration of RAI, then a mean dose of about 3 GBq seems to be
required. If one also intends to destroy malignant cells then a still higher dose in the
range of 3.5-5.5 GBq administered to a hypothyroid patient may be necessary.
Typically, one cannot know which situation applies until the post-therapy scan (at
the earliest). The Tg level may help one to decide the range of dose that is warranted.
In order to avoid missing the opportunity to destroy tumor, most practitioners opt
for a larger 131I dose. Single doses of 1.1 GBq (30 mCi) do not reliably achieve
remnant ablation and patients so treated tend to be lost to follow up, possibly as a
result of becoming disaffected by serial bouts of hypothyroidism needed to complete
the ablation. In our view, the goal should be to achieve ablation with a single treatment
and this justifies doses of 3-4 GBq. If the goal is to also treat known residual tumor
then the initial dose will be higher. If the Tg is elevated, a follow up scan is required
in 6-12 months with retreatment if tumor persists.
14 Patients who have previously been ablated may require retreatment of metastatic
disease. In these cases there is an advantage to the dosimetric method, which will
allow administration of the largest non-toxic dose. If standard, non-dosimetric
approaches are used, the administered dose should be in the range of 5.5-7.5 GBq
and sometimes higher, provided that the blood counts and renal function are normal.
If multiple treatments are required, they may follow at intervals of 6-12 months
depending upon the clinical urgency (Fig. 3).
External beam radiotherapy of the neck has little place in the management of
iodine-avid thyroid cancer. The dosage that can be delivered to tumor is generally
lower than with RAI, except in those that are not iodine-avid. External beam therapy
should be reserved for treatment of non-iodine-avid disease and for management of
emergencies when RAI cannot be mobilized quickly enough.
Radionuclide therapy of Thyroid Disorders 289
Figure 3. A 20 year old female who presented with locally invasive and surgically
unresectable papillary carcinoma. Each image was obtained 48 hours after
administration of 70-370 MBq of 131I. In (a) the pretreatment image shows intense
uptake into thyroid tissue; (b) partial response following first treatment with a
persisting focus of disease in the right neck and (c) ablation of all iodine-avid tissue.
She has now been disease-free for a decade.
Lung metastases are generally treatable with RAI, especially if they are larger
than 1-2 mm in size. Microscopic metastases may not be treatable because the mean
path length of the beta particle is then much larger than the tumor diameter (Fig.
4). Most of the radiation dose is then deposited in normal lung and may cause
pulmonary fibrosis. Consideration has been given to the use of 125I whose low-energy
Auger electrons have a much shorter path length (less than 72 microns) and would
deposit more energy within the small dimensions of the tumor but this approach is
still investigational.
Bone metastases from papillary carcinoma are very resistant to treatment and
they tend to progress even if they concentrate RAI. In such cases, treatment with
RAI should be followed with external beam radiotherapy to the metastatic site. For
a suspected solitary site, surgical excision might be attempted, but most cases are
associated with other sites of blood-borne metastases. Bone metastases from follicular
carcinoma may respond well to RAI alone as their uptake tends to be greater (Fig. 5).
The historical model for radiation safety concerning RAI treatment considered
the patient as an iodine point source and modeled the radiation safety protocol
accordingly. In fact, the patient is an attenuated, distributed source, which greatly
reduces the gamma dose to others. Nor did the historical model take into account 14
the concept of the effective dose equivalent (see Chapter 2). These three factors,
taken together, reduce the risk posed by the patient by a factor of about 20 relative
to the model. The historical modeling also did not take account of the ability of
patients and families to follow instructions to further reduce exposures from the
patient and the possibility of individualizing instructions to the patient. It was from
this historical model that a fixed limit of 1.1 GBq (30 mCi) for out-patient therapy
was determined.
The assumptions inherent in this model came to be questioned as a result of
recommendations to reduce public exposures from 5 to 1 mSv/year. Mere
intensification of past practices would have greatly increased the need for
hospitalization for both malignant and benign therapies. The re-evaluation has served
to justify modification of the approach to these treatments. In many countries,
290 Nuclear Medicine
14
including the United States and Canada, there is no longer a rigid requirement to
admit patients for large dose therapies.
Most thyroid cancer patients do not require supportive hospital care. If they are
admitted, it is because they do not have the domestic facilities necessary for self-
sequestration at home, that they have come for treatment from a distance or with
public transportation, or that they are considered to be unlikely to follow instructions.
If these patients were treated as out-patients, they would constitute a source of
radiation exposure to members of the public or to juvenile or pregnant family
members. Thus, some patients are still best treated in hospital. For others, out-
Radionuclide therapy of Thyroid Disorders 291
Figure 5. A 62 year
old woman who
presented with a
pathological fracture
secondary to
metastatic follicular
carcinoma. Her
hemoglobin was 80g/
L and platelets
<100x109/L. With RAI
therapy her disease
regressed and bone
marrow function
normalized but the
bone metastases were
never ablated. She
remained active for
most of the next 14
years, requiring
repeated RAI
treatments and
occasional external
beam treatments
when she developed
non-iodine-avid
metastases.
14
patient treatment is both safe and preferable in terms of availability of the emotional
support from families. The requirement for safe home treatment is that family and
other close associates of the patient do not receive more dosage than is permitted by
the local nuclear safety regulations.
The instructions for at home behaviour will be similar to those given to patients
treated for benign disease except that they will take account of the small to absent
organified component. As a result, iodine is much more rapidly eliminated from the
body. Patients who are totally athyrotic and have no tumor uptake may be nearly
clear of radioactivity by 2-3 days. Those with remnants and a significant uptake will
need to remain isolated for a longer time. The patient can be given a behavioural
292 Nuclear Medicine
14
Figure 6. A 26 year old woman with papillary carcinoma of thyroid. Her tumor
does not secrete Tg. The first post-surgical scan showed bilateral positive nodes in
her neck and the follow-up scan showed clearing after one RAI treatment. She then
developed recurrent neck node disease. After a bilateral neck dissection she had a
negative post-therapy RAI scan. However the FDG PET scan showed multiple nodes
in the upper mediastinum, a finding confirmed by a subsequent CT scan. The image
is a coronal section. The margins of the neck are defined by FDG uptake into the
sternocleidomastoid muscles and the positive nodes are located centrally.
Radionuclide therapy of Thyroid Disorders 293
Figure 7. A woman in her 60s who has pulmonary and neck metastases from a
papillary carcinoma that does not take up iodine as in (a). The tumor avidly
concentrates 111In-pentetreotide (b). Note the intense focus in the anterior neck and
the multiple lung lesions. The neck mass corresponds to palpable disease.
prescription concerning contacts with family and other close associates based on
measurements of dose retention from the diagnostic uptake.
Post-Therapy Imaging and Follow-Up
The sensitivity of radioiodine imaging increases asymptotically toward a maximum
with increasing administered dose up to the therapeutic range. Thus, the best possible
opportunity for accurate staging is by whole body scanning 5-7 days following
administration of a therapy dose. Several scenarios will emerge:
If the scan reveals only a small remnant within the thyroid bed and no
metastases and the Tg is normal, then the prognosis is excellent. These 14
patients may be reassured and scheduled for long term follow up.
If the scan reveals only thyroid bed remnants and the Tg is slightly elevated,
the latter may have originated from the stimulated normal tissue. These
patients may be guardedly reassured but should be rescanned in 6-12
months to ensure that ablation was complete and that the Tg has
normalized.
If the scan reveals metastatic disease, the patient will need a further scan
and possible retreatment in 6-12 months.
If the scan reveals no metastases and the Tg is high, then iodine-negative
metastases must be suspected. These patients need supplementary imaging,
including ultrasound of the neck and CT of the neck and chest. Positron
294 Nuclear Medicine
have an acute worsening in the 6-8 weeks following treatment and treatment with
prednisone during this interval should be given. Ophthalmopathy may first occur
years after the hyperthyroid state has been fully treated.
Can you predict if metastatic disease is likely or unlikely to be
iodine-avid before treatment?
No. We do not designate the disease as non-iodine-avid unless the post-therapy
scan is negative in the presence of other evidence of disease i.e., an elevated Tg or a
positive biopsy.
How do you alter your approach for Hurthle cell, tall cell or
insular cell carcinomas?
The initial ablation approach is identical. The Tg level with the patient off thyroid
replacement is very important. Occasionally Hurthle cell tumors and other poorly-
differentiated papillary cancer variants do take up 131I. Tall cell and insular variants
often take up 131I quite avidly. If the post-therapy scan is negative and disease is
suspected, then a PET FDG scan is obtained. Imaging with 111In-octreotide can also
be helpful in such cases.
What treatments are available for non-iodine avid metastases?
Surgery should be considered. If the lesion cannot be excised then external beam
radiotherapy should be considered. If the lesion is thought to have a high risk of
further recurrence then both modalities may be used. Lesions that bear somatostatin
receptors, and that are not readily treated by more established means, may be treatable
with 111In or 90Y-octreotide. Chemotherapy has not been very successful with
differentiated thyroid cancers but occasionally anthracyclines may achieve a remission.
Additional Reading
1. Atay-Rosenthal S. Controversies on treatment of well-differentiated thyroid
carcinoma and factors influencing prognosis. Nuc Med Ann 1999:303-334.
A comprehensive review of long term outcomes of thyroid cancer treatment of a large
population in Turkey.
2. Fatourechi V, Hay ID. Treating the patient with differentiated thyroid cancer with
thyroglobulin-positive iodine-131 diagnostic scan-negative metastases: Including
comments on the role of serum thyroglobulin monitoring in tumor surveillance.
Sem Nucl Med 2000; 30:107-114. 14
The title says it all.
3. Freitas JE. Therapeutic options in the management of toxic and non-toxic goiter.
Sem Nucl Med 2000; 30:88-97.
This article reviews therapeutic options in a more extensive fashion than space allowed
in this Chapter.
4. Freitas JE. Therapy of differentiated thyroid cancer. Nucl Med Ann 1998:83-108.
A comprehensive recent review of thyroid cancer from the perspective of the nuclear
medicine specialist.
5. Garcia M, Bvaskin HJ, Feld S et al. AACE Clinical Practice Guidelines for Evaluation
and Treatment of Hyperthyroidism and Hypothyroidism. 1996. http://
www.aace.com/clin/guides/thyroid_guide.html.
Another formally developed guideline.
6. Ladenson PW. Recombinant thyrotropin versus thyroid hormone withdrawl in
296 Nuclear Medicine
evaluating patients with thyroid carcinoma. Sem Nucl Med 2000; 30:98-106.
Although it deals only with the currently approved diagnostic uses of rh-TSH, this
article will support the reader who wishes to know more about this exciting new drug.
7. Shapiro B, Rufini V, Jarwan A et al. Artifacts, anatomic and physiological variants
and unrelated diseases that might cause false positive whole-body I-131 scans in
patients with thyroid cancer. Sem Nucl Med 2000; 30:115-132.
This is an excellent reference for the resident who must become familiar with pitfalls
and traps in scan reading and thyroid cancer care.
8. Singer P A, Cooper DS, Levy EG et al. Treatment guidelines for patients with
hyperthyroidism and hypothyroidism. JAMA 1995;273:8098-8112.
One of several formally sanctioned clinical guidelines on treatment of thyroid dysfunction.
9. Singer PA, Cooper DS, Daniels GH et al. Treatment guidelines for patients with
thyroid nodules and well-differentiated thyroid cancer. Arch Int Med 1996;
156:2165-2172.
Formal clinical guidelines on the treatment of thyroid cancer.
10. Schlumberger MJ. Papillary and follicular thyroid carcinoma. N Engl J Med 1998;
338(5):297-306.
A compact review of the state of knowledge concerning thyroid cancer.
11. Utiger RD. Follow-up of patients with thyroid carcinoma. New Engl J Med 1997;
337:928-930.
A brief discussion of standards of care for follow-up of thyroid cancer.
12. Braverman LE, Utiger RD, eds. The Thyroid. Eighth ed. Philadelphia: Lippincott
Williams & Wilkins, 2000.
A comprehensive, heavily referenced work concerning thyroid diseases.
14
CHAPTER 1
CHAPTER 15
Tumor Imaging
A.J.B. McEwan
Introduction
The role of nuclear medicine in the management of patients with cancer has
changed significantly over the past 10 years. Improved anatomical visualization
through the introduction of multi-slice CT and MRI sequences for chest and
abdominal imaging has returned the role of radionuclide imaging to the functional
assessment of primary and metastatic cancer.
The ability to target a biochemical, metabolic or pathologic process with a specific
radiopharmaceutical provides nuclear medicine with a unique ability to contribute
to patient management by the functional assessment of tumor status or the
demonstration of functional abnormalities at sites too small to be identified as
abnormal with anatomical imaging techniques.
An understanding of the choice of radiopharmaceuticals is key to the appropriate
use of nuclear medicine in patient management, while the ability to define a metabolic
target against which a radiopharmaceutical can be developed offers a unique
perspective on tumor imaging. Techniques such as positron emission tomography,
peptide receptor imaging and other methods of assessment of functional status such
as hypoxia imaging also allow the use of nuclear medicine as predictive assays of
treatment response, confirmation of tumor presence, definition of tumor type and
whole body assessment of metastatic spread.
The goals of cancer imaging with nuclear medicine techniques are slowly being
redefined (Table 1). Most of the indications listed as under investigation are still
considered developmental and have not progressed beyond Phase I or Phase II trials.
However, convincing data now exist for at least the first three indications in this
group (predictive assay of treatment response, assessment of treatment efficacy, and
assessment of multi-drug resistance) and for specific tumors these may be considered
routine indications.
Mechanisms of Radiopharmaceutocal Uptake
The radiopharmaceuticals most commonly used in imaging in oncology are
reviewed below and are summarized in Table 2. Most of these radiopharmaceuticals
rely on pathophysiological or biochemical differences between neoplastic and normal
cells to provide a target to background ratio which is suitable for imaging.
The development of cancer is characterized by the development in the cell of key
characteristics such as the loss of local contact inhibition, the ability to divide without
control, transmissibility of these changes, angioneogenesis and the ability to
metastasize. The resulting abnormal mass will invade surrounding tissues and cells
may be shed to migrate via lymphatic or vascular channels to form distant metastases.
Normal Distribution
Following intravenous injection, activity is seen in the brain and the heart; cardiac
activity can be decreased by fasting. Excretion is by the renal route; kidneys are
faintly visualized and the bladder may be a confounding element in pelvic imaging.
The liver is also faintly seen throughout the period of the study. Muscle activity,
including talking around the time of injection, will result in accumulation within
the corresponding muscle group. Figure 1 shows the normal biodistribution of FDG.
300 Nuclear Medicine
Note the visualization of the brain, heart and bladder; the kidneys, bone marrow
and GI tract are faintly seen. Patient preparation is key to successful FDG imaging.
15 False positive uptake may be seen in sites of inflammation and granuloma.
Imaging Parameters
Patients should fast for 4-6 hours prior to FDG imaging. Lorazepam (Ativan)
or diazepam (Valium) are commonly administered to reduce muscle activity.
Following intravenous administration, patients should lie quietly for 1-2 hours prior
to imaging. Imaging is performed on a PET scanner as a series of 6-8 tomographic
images which are reconstructed to display a whole body image as well as transverse,
coronal and sagittal slices. Lesion resolution down to 0.5 cm has been reported.
Tumor Imaging 301
67
Ga-Gallium citrate
Mechanism of Uptake
This radiopharmacuetical is a radiometal that binds in vivo to iron binding
proteins such as transferrin, lactoferrin and ferritin. The mechanism of uptake into
the cell is associated with the intravascular gallium-transferrin complex binding to
transferrin receptors on the tumor cell surface; this binding is followed by endocytosis
and binding to intracellular proteins such as lactoferrin. Transferrin receptor and
lactoferrin concentrations may be increased in malignancy. Secondary mechanisms
of uptake include increased diffusion from the intravascular into the interstitial space.
Normal Distribution
Following intravenous injection, normal uptake is seen in salivary and lacrimal
glands, liver, spleen, skeleton, bladder and large bowel. Uptake will also be seen in
the lactating breast and in sites of inflammation such as infection, granulomata and
post surgical sites.
Imaging Parameters
Patients are usually imaged 48-72 hours after injection, by which time optimal
visualization of neoplastic processes will be seen. SPECT imaging will enhance
visualization of small lesions. Delayed imaging at 5 or 6 days may occasionally be
needed to allow for gut clearance and improved visualization of the abdomen and pelvis.
123
I or 131I-metaiodobenzylguanidine (mIBG)
Mechanism of Uptake
mIBG is a guanethidine derivative and functional analogue of norepinephrine.
It enters the cell by the amine uptake 1, sodium dependent pump, following the
same pathway as noradrenaline. Once within the cell, it accumulates in the
intracellular storage granules found in cells of the sympathetic nervous system. The
patient must be screened to ensure that no medications are present that could affect
the uptake mechanism. This includes sympathomimetics (found in some over-the-
counter decongestants and inhaled bronchodilators), tricyclic and atypical
antidepressants, antipsychotics, antihypertensives (reserpine, calcium channel
blockers, labetolol) and cocaine. Excretion is primarily by the renal route. Uptake
has been reported in benign and malignant pheochromocytoma, carcinoid,
neuroblastoma and medullary thyroid cancer.
15
Normal Distribution
Following intravenous injection, uptake of mIBG is seen in salivary glands, heart,
lungs (early), liver and bladder. Late splanchnic activity is seen in about 20% of
patients and with an 123I label, the normal adrenals may be visualized, particularly
with SPECT imaging.
Imaging Parameters
Imaging may be performed with mIBG labeled with either 123I or 131I. 123I-
mIBG is the imaging agent of choice with far superior imaging characteristics.
Planar and SPECT images are routinely performed 24 and often 48 hours after
302 Nuclear Medicine
Normal Distribution
Normally, 111In-pentetreotide accumulates in liver, spleen, kidneys and bladder.
The gallbladder may be visualized at both imaging time points (4 and 24 hours) and
should not be confused with hepatic metastases. At 24 hours, bowel activity is almost
always seen. The normal pituitary is seen in 20-40% of patients.
Activated lymphocytes also show binding of octreotide and therefore false positive
uptake may be seen in sites of granulomata, acute inflammation, recent surgical
scars, radiation fields and arthropathies.
Imaging Parameters
15 Patients are usually imaged 4 and 24 hours after the injection of the
radiopharmaceutical. SPECT imaging, performed optimally at 4-6 hours, is essential
to the accurate interpretation of these images, particularly in the evaluation of residual
disease and in the examination of the neck and mediastinum in patients with
medullary thyroid cancer.
99m Tc-Sestamibi and 99mTc-Tetrofosmin
Mechanism of Uptake
Like thallium, these two radiopharmaceuticals were introduced as myocardial
perfusion agents. Both agents have relatively high lipophilicity with rapid transit
Tumor Imaging 303
across the cell membrane. Binding within the cell occurs to mitochondrial proteins,
a process that appears to be energy dependent. Transport out of the cell is mediated
by P-glycoprotein.
Mutations of the p53 suppressor gene and p21-ras oncogenes are associated with
increased expression of the multi-drug resistance gene (MDR-1). Amplification of
this gene increases expression of P-glycoprotein which leads to increased outflow of
99mTc-sestamibi, 99mTc-tetrofosmin and many chemotherapeutic agents. This finding
has led to the postulate that absent uptake of either radiopharmaceutical by a tumor
may predict failure of chemotherapy.
Normal Distribution
Following intravenous injection uptake is seen rapidly in salivary glands, thyroid,
heart, liver, bowel, kidneys and bladder. Washout over 1-2 hours is seen in normal
tissue. Delayed washout from tumor provides a tumor to background ratio that can
be seen as focal uptake in the tumor at 2-3 hours post injection. Skeletal muscle is
commonly seen and exercise increases this uptake.
Imaging Parameters
Because of gut excretion, 99mTc-sestamibi and 99mTc-tetrofosmin are most effective
when imaging tumors in the brain and above the diaphragm, and have been reported
as effective in imaging thyroid, lung and breast cancer. Following intravenous
injection, planar images are obtained at 15 and 120-180 minutes. SPECT imaging
will improve the diagnostic yield for small lesions.
99m
Tc-Sulfur colloid
Mechanism of Uptake
Although colloids are not accumulated by tumors, they are useful for sentinel
node mapping which is an increasingly important part of the modern management
of patients with melanoma and breast cancer. Lymphatic spread is a common route
of dissemination of these tumorsif there is metastatic involvement of the regional
nodes then the drainage basin also needs appropriate treatment. The assumption
that lymphatic drainage is an orderly and sequential flow from the primary site has
led to the concept of the sentinel nodethe first nodal group that receives drainage
from the tumor site. This drainage route is not predicted accurately from standard
anatomical models and trunk lesions often have more than one drainage route. If
this nodal group can be defined and biopsied then the presence or absence of
metastatic spread can be confirmed and appropriate treatment instituted. 15
99m
Tc-sulfur colloid injected intradermally or peri-tumorally will be removed
from the injection site by lymphatic channels and will be retained first in the sentinel
lymph node. This can then be identified by imaging or by probe to guide the surgeon
to the biopsy site.
Normal Distribution
Following peritumor intradermal injection, the lymphatic channels may be visu-
alized and the draining nodes will typically be seen. In most cases this occurs rapidly
after injection so imaging should commence as soon after injection as is feasible.
Node visualization is slower following a subcutaneous injection. The 6 hour half life
304 Nuclear Medicine
of 99mTc allows identification of the sentinel node with a surgical probe. The tech-
nique is well described in the surgical literature.
Imaging Protocol
99mTc-sulfur colloid is injected in a small volume (0.25 ml) in 4 quadrants around
the tumor. The patient is then placed under a gamma camera and sequential images
obtained until the sentinel node is visualized.
Optionally, the site is then marked with indelible dye on the skin (some centres
report the successful use of a radio-opaque needle for node localization). The patient
can then be transferred to the surgical suite for probe guided open biopsy.
201
Tl-Thallium chloride
Mechanism of Uptake
Thallium is a physiological analogue of potassium; uptake into the cell is by the
ATP-ase dependent Na+/K+ pump. Initially 201Tl was introduced as a myocardial
perfusion agent and it is believed that uptake by tumor cells is associated with the
increased perfusion seen in angioneogenesis; there is also evidence of delayed washout.
Thallium uptake requires the presence of a vascular supply and the presence of viable cells.
Normal Distribution
Following intravenous injection, uptake is seen rapidly in salivary glands, thyroid,
heart, liver, bowel, skeletal muscle, kidneys and bladder. Washout over 1-2 hours is
seen in normal tissue; delayed washout from tumor provides a tumor to background
ratio that can be perceived on delayed images.
Imaging Parameters
Thallium is most effective when imaging tumors in the brain and above the
diaphragm, such as thyroid cancer. Following intravenous injection, planar images
are obtained at 15 and 60 minutes; SPECT imaging at 60 minutes may improve the
diagnostic yield for small lesions.
Monoclonal Antibodies
Mechanism of Uptake
One of the differentiating features of cancer cells compared to normal cells is the
overexpression of cell surface antigens or the expression of cancer associated antigens.
Monoclonal antibodies (MAbs) may be produced against these antigens, radiolabelled
15 with technetium-99m, indium-111 or iodine-123, and used to image patients with
cancer. For example, 111In-satumomab pendetide (OncoScint) is a MAb directed
against TAG72, a tumor associated glycoprotein. After intravenous injection the
MAb binds to tissues overexpressing the antigen and clears from normal tissue to
provide a high target to background ratio.
In addition to complete antibodies, MAb fragments such as 111In-capromab
pendetide (Prostascint), directed against prostate specific antigen, and 99mTc-
arcitumomab (CEAscan), directed against carcinoembryonic antigen have entered
clinical use. These fragments are characterized by rapid clearance from the background
and more rapid penetration into tumor masses.
Tumor Imaging 305
Normal Distribution
Following intravenous injection, the whole body image initially shows a blood
pool distribution. Thereafter clearance is by the urinary and GI routes; bladder,
liver, and GI tract will always be seen. In MAbs labelled with radiometals, kidney
and splenic visualization is normal.
Imaging Parameters
For MAb fragments immediate and delayed images at 4-6 and 24 hours are
usually obtained to compare blood pool activity with late tumor binding. Some
authors recommend obtaining a blood pool image and using subtraction techniques
to improve diagnostic yield. Intact MAbs clear much more slowly and are usually
labeled with 111In or 131I with imaging up to 7 days.
Images are not easy to interpret and a learning curve is required to appreciate the
significance of the subtle scintigraphic findings. MAb imaging has been used to
define the functional nature of a mass and to evaluate residual disease.
Contributions of Nuclear Medicine to Cancer Imaging
The clinical indications for nuclear imaging with the radiopharmaceuticals
discussed in the preceding section are summarized in Table 3 and are discussed in
more detail below.
Brain tumors
MRI remains the primary diagnostic imaging modality for the diagnosis of patients
with primary and metastatic brain tumors, both at presentation and at time of
recurrence. However, the differentiation of scar tissue or radiation necrosis from
viable tumor and perhaps the in vivo assessment of grade of malignancy may be
most appropriately achieved using molecular imaging techniques. Thallium-201 may
be used for SPECT imaging and 18F-FDG for PET imaging. Radiopharmaceutical
accumulation confirms the presence of viable tumor (Fig. 2). Sensitivity for both
techniques is high, and although numbers of patients are small this is an important
indication for nuclear medicine imaging. The presence of abnormal
radiopharmaceutical uptake at a morphologically abnormal site may be confirmed
by the use of fusion imaging whereby the functional and anatomical images are
overlaid.
Some data show that survival in patients with high grade gliomas is inversely
proportional to FDG uptake; this finding may lead to the development of the
technique as a predictive assay of treatment response.
Abnormal 111In-pentetreotide uptake has been reported in meningioma, pituitary
15
adenoma and astrocytoma. Although sensitivity is moderate to high for these lesions,
there is little indication for routine use. Some authors, however, do suggest a possible
role for in vivo differential diagnosis in highly selected patients.
Breast cancer
Two techniques have been described that contribute to the staging of patients
with breast cancer and to their re-evaluation at time of recurrence.
Axillary lymph node status is one of the most important criteria in assessing
prognosis of patients at presentation. Patients with histological evidence of tumor in
axillary nodes have a significantly lower overall survival and disease free interval;
306 Nuclear Medicine
those with more than 3 nodes involved do less well than those with 1-3 nodes in-
volved. The current standard management of patients with breast cancer includes
axillary dissectionto acquire prognostic information, to effectively treat the axilla
if disease is present, to possibly improve long term survival and to treat locally ad-
vanced disease. However, axillary dissection is associated with significant morbidity
in up to 30% of patients. Symptoms include pain, lymphedema and loss of range of
15 movement. Internal mammary node involvement is less common and more diffi-
cult to diagnose than axillary involvement but may be more frequent when the
primary tumor lies in the inner quadrants. The prognostic implications of internal
mammary node involvement are comparable to those of axillary involvement. Most
radionuclide imaging techniques have been developed with the hope that they can
contribute to improved staging and assessment of internal mammary and axillary
node status.
Lymphoscintigraphy and Sentinel Node Imaging
The goal of sentinel node imaging is to identify the node(s) closest to the primary
tumor site. After peritumor injection of 99mTc- sulfur colloid, imaging of the anterior
Tumor Imaging 307
Figure 2. MRI and 18F-FDG transaxial images in a patient with recurrent glioblastoma
multiforme. The MRI shows an extensive area of abnormal signal in the right
temporal lobe (arrow) in a patient with a suspected recurrent glioblastoma. The
18F-FDG image shows recurrent tumor only anteriorly (arrow). The lower part of
the abnormality seen on the MRI scan is therefore scar (arrowheads). (Courtesy of
ADAC Labs and Dr. Carreras, Clinical PET Institute, Madrid, Spain.)
15
chest and axilla will usually show focal uptake in the sentinel nodes of the axilla and
internal mammary chain within 30 minutes (Fig. 3). The site of the sentinel node
can then be marked on the skin and subsequently identified at dissection using a
gamma probe. In this way a limited number of axillary lymph nodes are removed,
sparing the patient a complete axillary lymph node dissection.
In one series, lymphatic nodal mapping had an accuracy rate of about 40% for
the radioactive technique; 2 false negatives were identified at surgery and at axillary
lymph node dissection, 4 patients had false positive images associated with
histologically negative nodes. However, in other series the success rate in identifying
308 Nuclear Medicine
Figure 3. Sentinel node imaging in a patient with breast cancer. The peritumoral
injection site of 99mTc-sulfur colloid is seen as an area of intense activity. Sentinel
nodes are seen in the internal mammary chain (arrow) and axilla (arrowhead).
(Courtesy of Dr. R. Kloiber, Foothills Hospital, Calgary, Canada.)
sentinel nodes has been as high as 90% (up to 100% in one series of 34 patients).
Experience appears to be important to successful imaging.
The technique is now becoming standard of care in many breast centers allowing
a decrease in the axillary node dissection rate; the effectiveness of this technique in
improving survival or disease free interval is not established.
FDG Imaging
Although the role of FDG imaging in breast cancer is still being defined,
indications will probably include detection of recurrence and monitoring treatment
response. It may also have a role in the initial staging. Although sensitivities as high
as 80-100% have been reported in the visualization of primary breast cancer, the
effectiveness of morphological techniques, scintimammography and fine needle
aspiration mean that PET is less likely to have a major role in this setting.
Staging
For staging, PET imaging offers the ability to detect involvement of both the
axillary and internal mammary chains. While axillary node dissectionperhaps in
conjunction with sentinel node imagingremains the standard for staging the axilla,
there are no effective tests for evaluating the internal mammary chains. Sensitivities
15 of 85-100% have been reported in most large series for detecting nodal metastases
(Fig. 4); specificity is somewhat lower, typically 65-90%. Although some authors
have suggested that a negative PET scan of the axilla can be used to exclude axillary
lymph node dissection, microscopic disease and lesions less than 7-10 mm will be
missed. Distant metastases in the small percentage of patients who present with
Stage IV disease will be accurately demonstrated.
Recurrence
Several studies have shown PET to be more accurate than CT or MRI in detecting
local recurrence, locoregional lymph node metastases and distant metastases; a high
sensitivity in detecting osteolytic metastases has also been claimed. In a patient with
Tumor Imaging 309
Figure 4. Coronal sections from an 18F-FDG scan in a patient with metastatic breast
cancer. Abnormal accumulation of 18F-FDG in a lymph node metastasis in the right
axilla is indicated by the arrowhead. (Myocardial activity is seen in the mediastinum.)
(Courtesy of ADAC Labs and Dr. J.-F. Gaillard, HIA Val de Grace, Paris, France.)
suspected recurrence it is probable that PET imaging will become the primary in-
vestigation in assessing degree of spread.
Monitoring Response
It is well known that a significant number of patients will fail first line
chemotherapy. Data are now available from several centres suggesting that if FDG
uptake by the tumor fails to decrease significantly after the first 2 cycles of
chemotherapy, the patient is likely to be a non-responder. Absent FDG uptake after 15
2 cycles appears to predict a good response. If this is confirmed in additional clinical
trials, then monitoring of response will probably become a routine measurement in
PET imaging.
Scintimammography
Recent years have seen a significant increase in disease free survival in patients
with breast cancer, in part due to earlier diagnosis. A key tool in early diagnosis is the
widespread use of screening mammography. Up to 8% of screening mammograms
show abnormalities considered to be suspicious for malignancy. Patients are then
usually referred for additional investigations, including diagnostic mammography,
310 Nuclear Medicine
ultrasound and MRI. Even if these additional studies show probable cancer, biopsy
confirms the diagnosis in only about 30% of patients. Therefore, while mammography
has a high negative predictive value, it has low specificity, even in conjunction with
additional conventional imaging. The specificity of mammography, ultrasound and
MRI is particularly poor in patients with dense breasts, previous breast conserving
surgery and implants.
It has been postulated that scintimammography with 99mTc-sestamibi has a role
in the management of this latter population of patients (Fig. 5). In several large
series, high sensitivity (80-92%) and high specificity (80-100%) have been
demonstrated; negative predictive values of up to 96% have been reported. However,
these data are for lesions >1 cm in diameter and diagnostic accuracy is lower for smaller
lesions.
Scintimammography is not a screening tool, but as a supplementary examination
15 to mammography and ultrasound it may have a role in evaluating specific populations
of patients as defined above and in evaluating inconclusive morphological images.
The technique is likely to become routine for the evaluation of the dense breast and
the assessment of recurrence in patients who have had previous surgery. The
advantages of functional imaging become clear in these settings.
Gastrointestinal Tract
CT and ultrasound are the primary investigations for staging and assessing
recurrence of these tumors. Nuclear medicine techniques may contribute to
management of some cases of esophageal and colorectal cancers.
Tumor Imaging 311
Figure 6. Coronal PET images of a patient with a large metastatic colorectal lesion
in the left lung (arrow) and a local recurrence in the pelvis (arrowhead). (Courtesy
of ADAC Labs and Dr. J.-F. Gaillard, HIA Val de Grace, Paris, France.)
Esophageal Cancer
The incidence of esophageal cancer in the United States is 3.5/100,000, although
it is considerably higher in parts of China and is more common in France, Singapore,
Iceland and Switzerland. Initial presenting symptoms include dysphagia and weight
loss. Two year survival is only 20%.
Early data have suggested that FDG imaging may be an important tool in staging
this population of patients. FDG images were more accurate than CT in
demonstrating spread to mediastinal lymph nodes and were considered to have altered
staging and management in approximately one-third of the patients. As surgical and
radiotherapeutic morbidity are high, accurate staging of these patients will not only
improve management but should also contribute to improved quality of life.
Colorectal Cancer
Colorectal cancer affects approximately 1 in 20 people in the United States and
in most developed countries. If diagnosed early, it is curable by surgery. Spread is
initially through the muscularis mucosa and into the submucosa. With penetration
of the bowel wall, it spreads by local invasion; nodal metastases occur in approximately 15
20% of patients in whom the cancer is localized to the bowel wall. Nodal spread is
usually through the lymphatic network along the major vessels. The liver is the
commonest site of distant metastasis. Treatment of advanced disease is by
chemotherapy with surgery only if required for control of symptoms.
FDG has no role in the initial staging of this population. However, in patients
with locally advanced disease, FDG has been shown to be an effective management
tool for identifying distant metastases (Fig. 6). The most important role of FDG
imaging is in discriminating between residual tumor and scar in patients with prior
therapy and in identifying the presence of local recurrence (Fig. 7). For both
indications FDG has demonstrated sensitivities of the order of 90-95% compared
312 Nuclear Medicine
with CT sensitivity of 65-75%; FDG specificity is also high. FDG imaging alters
the management of colorectal cancer in about 30% of patients. Some data suggest
that FDG imaging might monitor response to therapy. A fall in FDG uptake after 1
cycle of chemotherapy appears to predict a good response whilst no change or an
increase in uptake suggests treatment failure and progression.
MAbs have also been reported as being effective in evaluating patients with
colorectal cancer. Mabs such as arcitumomab (CEAscan) and satumomab pendetide
(OncoScint) may either be used in conjunction with an intraoperative gamma
probe or by conventional imaging techniques. Arcitumomab (CEAscan) has been
demonstrated to increase the clinical accuracy of CT when used in conjunction
with CT: it improved the correct prediction of resectability by 40% and the correct
prediction of unresectability by 100% when compared with CT alone. Using the
15 same MAb, radioimmunoguided surgery has been shown to increase the number of
tumor sites in a locoregional distribution; although long term follow-up has not
been reported, the study argues for the inclusion of radioimmunoguided surgery in
the management of patients with locoregional extension.
MAb imaging is technically more difficult than many nuclear medicine procedures
and there is a steep learning curve to image interpretation; routine imaging with
MAbs should only be performed in departments with experience in the technique
and training courses are available for the use of radioimmunoguided surgery. If
available, 18F-FDG imaging is probably a more sensitive and routinely interpretable
investigation.
Tumor Imaging 313
Figure 8. 18F-FDG scan in a patient with a primary lung cancer involving the right
apex (arrow). The mediastinum is normal and the patient is a surgical candidate.
(Courtesy of ADAC Labs and Dr. J.-F. Gaillard, HIA Val de Grace, Paris, France.)
Lung cancer
Lung cancer is now the most common single cause of cancer related death in
both sexes, having recently overtaken breast cancer in women. In the United States
there are in excess of 160,000 new cases annually and survival at 5 years is approximately
15%.
The most important contributions of nuclear medicine to the management of
patients with lung cancer are in the evaluation of solitary pulmonary nodules and in
staging of confirmed disease using PET imaging with 18F-FDG; both indications
can often be performed with a single examination.
Most series have shown the sensitivity of PET to be of the order of 95% in
diagnosing cancer in a solitary pulmonary nodule; quantitative assessment with the
standardized uptake value (SUV) > 2.5 has a specificity greater than 90%. (The
SUV is an index of lesion uptake and is calculated as decay-corrected lesion activity
divided by the injected activity/body weight).
Once the presence of non-small cell lung cancer (NSCLC) has been confirmed,
an evaluation of the mediastinum for locoregional metastatic involvement is required
to plan therapy. CT has a sensitivity of 60-70% and specificity of 65-75% in defining
mediastinal involvement, and mediastinoscopy an accuracy of 89%. Metastatic
involvement has been shown to be present in 13% of nodes <1 cm on CT and in
only 36% in nodes >2 cm, so current staging techniques have significant limitations.
FDG imaging can improve patient management. Sensitivity and specificity in 15
detecting mediastinal disease has been reported as 90-95% and up to 95%,
respectively. The commonest cause of false positive uptake is granulomatous disease.
Prospective comparisons of conventional staging (CT, ultrasonography, bone
scanning, and, when indicated, needle biopsies) with FDG PET have shown that
PET improves the rate of detection of local and distant metastases in patients with
NSCLC. In jurisdictions where PET is readily available, it has become the
investigation of choice (Fig. 8). In addition to evaluating mediastinal involvement
the ability of PET to provide whole body images will define that group of patients
who present with Stage IV disease (Fig. 9).
Using these criteria for staging patients with NSCLC, several groups have shown
changes in management based on the PET images; one recent study has shown
314 Nuclear Medicine
management changes in 67% of patients based on the PET scan. In addition there
are good data suggesting that the appropriate use of PET imaging in patients with
NSCLC can produce significant savings to the healthcare system by avoiding
unnecessaryand unhelpfulinterventions.
FDG imaging should be the standard of care in the management of patients
with NSCLC at presentation and probably for assessment of possible relapse.
Lymphoma
Lymphomas make up 8% of all malignancies. Hodgkins disease (HD) is
characterized by the presence of Reed-Sternberg cells. Four histologic subtypes have
been described: nodular sclerosing (40-60%), lymphocyte predominant (2-10%),
mixed cellularity (20-40%) and lymphocyte depleted (2-15%). Spread is
characteristically sequential from nodal group to nodal group and intrathoracic disease
is common. Non-Hodgkins lymphoma (NHL) is a more heterogeneous group of
tumors, defined as high, intermediate or low grade by the Modified Working
Classification. High grade tumors are aggressive with a short median survival if
untreated. Low grade tumors are more indolent with a longer median survival.
Staging for both HD and NHL uses the Ann Arbor system. The presence of
systemic symptoms (fever, night sweats, weight loss > 10% body weight) or pruritus
are considered poor prognostic indicators. Treatment, according to stage and
15 classification, is by chemotherapy and/or radiation therapy. Peripheral stem cell or
autologous bone marrow transplantation have proven effective salvage therapy options
and antibody therapies are now completing Phase III trials. To date one agent, 99Y-
ibritumomab tiuxetan (Zevalin) has received FDA approval for treatment of low-
grade CD20+ `-cell NHL.
Lymphoma is, therefore, a treatable cancer for which accurate staging,
demonstration of therapeutic effectiveness and early detection of recurrence are key
elements of successful management strategies. Imaging with gallium has
conventionally been the nuclear medicine investigation used in the management of
this patient population; however, the Centers for Medicaid Services (CMS, formerly
Tumor Imaging 315
Figure 10. Anterior and posterior views of a gallium scan in a patient with extensive
lymphoma in the right supraclavicular fossa extending into the mediastinum. Note
normal distribution in liver, bone marrow, large bowel and bladder.
HCFA) in the United States has recently approved the use of FDG in settings where 15
gallium would previously have been used. 111In-pentetreotide has also been shown
to have uptake in a percentage of lymphomas, but the role of the radiopharmaceutical
in this patient population has not yet been defined.
Staging
Currently, the two most important investigations in the management of
lymphoma are CT scanning above and below the diaphragm and biopsy for tumor
characterization. Gallium imaging has a limited rolein the identification of lymph
nodes or other abnormalities felt to be equivocal on CT and, less usefully, as a whole
body screening tool in the assessment of patients with Stages III and IV disease.
316 Nuclear Medicine
Figure 11A. Anterior and posterior gallium scans in a patient with non-Hodgkins
lymphoma (NHL) in the left supraclavicular fossa. Although asymmetry is seen
between the 2 supraclavicular fossae, the disease extent is not well perceived.
15
While planar imaging may be diagnostic (Fig. 10), in other cases SPECT is a neces-
sary tool in determining the presence and extent of disease (Fig. 11).
If the radiopharmaceutical is to be used to monitor treatment response or detect
relapse, it is also important to have a pre-treatment gallium scan to confirm that the
tumor is gallium-avid. About 95% of patients with HD and between 70-80% of
patients with NHL have positive uptake. Gallium imaging can be used to exclude
distant metastases (Fig. 12). Recent chemotherapy and in particular G-CSF can
cause significant alterations in gallium biodistribution (Fig. 13).
18F-FDG has higher sensitivity and specificity than gallium imaging and if PET
is available should be the investigation of choice (Fig. 14). PET is more sensitive
Tumor Imaging 317
Figure 11B. Coronal SPECT slices. Note how the disease extent can be readily seen
(arrow), in contrast to the planar images. Extension into the mediastinum is now
obvious (arrowhead).
15
than CT scanning in evaluating tumor burden at presentation, although anatomical
localization remains important for treatment planning. The data supporting the
routine use of FDG at presentation are more persuasive than they are for gallium
and most centres which have PET available as a routine tool will use FDG imaging
as part of their staging protocol.
Figure 12. Anterior and posterior gallium scans in a 10 year old boy with Burkitts
lymphoma affecting the right proximal femur (arrows). No additional lesions are
seen.
15
sites is fundamental to the definition of responsemost series suggest that 50-60%
of those patients do not have viable disease. Both gallium and FDG have high
sensitivity and specificity in defining the presence of active tumor and are the
investigations of choice.
Repeating the gallium imaging after 2 cycles of chemotherapy may predict
treatment failure if gallium uptake persists. Positive gallium uptake at the end of
treatment also appears to predict reduced survival and reduced disease free interval.
The same parameters for predicting treatment failure can probably be applied to
PET. In children a confounding feature is uptake within a hyperplastic thymus; this
Tumor Imaging 319
Figure 13. Gallium scan after G-CSF. There is altered biodistribution of the
radiopharmaceutical. Almost all activity is seen in the skeleton, with non-
visualization of the liver. (Same patient as Figure 12.)
15
may be differentiated by tomographic imaging where a characteristic H pattern of
uptake is seen on coronal sections.
Figure 14. Anterior 18F-FDG scan of a patient with extensive mediastinal lymphoma
(arrows). Metastases to pelvic lymph nodes are also present (arrowheads). (Courtesy
of ADAC Labs and Dr. A. Alavi, University of Pennsylvania, Philadelphia, USA).
body screening in high risk patients or those with palpable nodes. Standard of care
in this patient population should include imaging with either gallium or FDG.
As PET imaging becomes routinely available, FDG will almost certainly replace
gallium as the investigation of choice. If the preliminary data supporting its role as a
predictive assay of response are confirmed, it will be routinely performed prior to
treatment, after 1 or 2 cycles of therapy, and during follow-up.
Melanoma
Melanoma is increasing in incidence in most of the developed world, and in
countries such as Australia the incidence is doubling every ten years. At presentation
the most important prognostic indicator is tumor thickness as described by the
Breslow classification.
15 Five year survival in patients with Stage IIB (tumor thickness * 4.00mm) or
worse is poorchances of curative therapy depend entirely on the accuracy of staging
and the completeness of radical dissection. Locoregional recurrence is high and
restaging at recurrence carries the same requirements as at presentation. Interferon
alpha-2a (IFN alpha-2a) is efficacious as adjuvant therapy in stage IIB and III
melanoma, but is toxic and expensive. Accurate staging is therefore required before
embarking on a course of therapy.
Lymphoscintigraphy
Lymphoscintigraphy with sentinel node mapping has revolutionized the staging
of patients with melanoma, providing the surgeon with an accurate localization of
sentinel nodes. Orderly and sequential drainage to the sentinel node and then to the
Tumor Imaging 321
A B
FIgure 15. A) Sentinel node imaging showing 1 mintue sequential frames after
intradermal injection of 99mTc-sulfur colloid. The sentinel node is seen as a small
focus of activity at about 2 oclock superior to the injection site. B) Summed view
of the sequential images. The injections site has been shielded and there is improved
visualization of the sentinel node (arrow).
regional nodal basin is typical of melanoma, with less than 1% of patients present-
ing with skip lesions. The histological evaluation of the sentinel node, therefore, is
an important prognostic indicator and one which directs appropriate wide field
dissection. Five year survival in patients with tumor negative sentinel lymph nodes
is 90% following selective dissection and 93% following complete lymph node
dissection. Surgical intervention therefore can, with confidence, be limited in patients
with negative sentinel lymph nodes.
Imaging after peritumoral intradermal injection usually shows the lymphatic
channels within 20 minutes of administration and the sentinel node(s) should be
seen by 30 minutes (Fig. 15). Whilst one nodal basin is usually seen for lesions on
the limbs, up to three geographically disparate nodal basins may be demonstrated
after injection around a melanoma site on the trunk or head and neck region and up
to 40% of patients will show unexpected drainage patterns. A skin marker is placed
over the node(s) and the patient moved to the operating room for sentinel node
excision. A hand held gamma probe is an important adjunct to the localization of
the marked node at surgery.
In this population of patients, sentinel node imaging has been shown to be effective
in identifying the sentinel nodes in 90-95% of patients in most series and should
now be regarded as standard of care in the management of this rapidly expanding
15
patient group.
FDG Imaging
Most authors believe that PET imaging with FDG is the most effective way of
identifying locoregional and distant metastases. PET imaging methods show a
sensitivity and specificity of over 90% in patients with medium to high pre-test
probabilities. Tumor deposits as small as 5 mm have consistently been identified,
and sensitivity is significantly higher than for either CT or MRI. However, because
of high background activity in the liver and brain, small lesions in these sites will be
more effectively diagnosed with morphological imaging techniques. False positive
322 Nuclear Medicine
studies will occur in patients with infection and recent surgery; these factors should
be considered when requesting the examination.
At present FDG imaging in patients with melanoma is most effectively used in
staging patients with a high likelihood of distant metastases (i.e., Stage IIB or III)
and in evaluating patients at high risk of recurrence.
Gallium Imaging
15 Gallium shows high sensitivity for melanoma (80-85% with high dose
administrations), but specificity has been lower (typically 60-75%) and imaging of
the lungs and brain is ineffective.
However, if PET is not available, then there will be a small group of patients in
whom CT or MRI is equivocal and biopsy is difficult and in whom gallium imaging
with SPECT may make a contribution to management.
111
In-Pentetreotide
Several groups have now reported sensitivities of up to 75% in patients with
melanoma. No attempt has yet been made to test the contribution to patient
management, though some preliminary data suggest that uptake may be a predictor
Tumor Imaging 323
15
Figure 17.Anterior and posterior 123I-mIBG images in a patient with very advanced
metastatic carcinoid from an ileocecal primary seen in the right lower quadrant
(arrow). Note extensive metastatic involvement of the liver and bones (arrowheads).
Neuroendocrine tumors
Carcinoid Tumors
Carcinoid tumors arise from enterochromaffin cells and are the most common
neuroendocrine tumors of the GI tract. The commonest sites of primary presentation
are the appendix and small bowel. Tumors arise, rarely, in the bronchus, thymus and
pancreas. Primary treatment is surgical; spread to the liver is not uncommon and in
a minority of patients causes the carcinoid syndrome, characterized by flushing,
diarrhea and asthma.
The role of nuclear medicine imaging is to confirm the neuroendocrine nature
of masses visualized on CT or ultrasound, to assess disease extent and to plan for
therapy. 123I-mIBG is highly accurate (80% in most series) in detecting metastases
from mid and hind gut carcinoid. Primary lesions will occasionally be seen (Fig.
17). The overall accuracy of 111In-pentetreotide (Fig. 18) is somewhat higher than
123
I-mIBG, probably in the order of 80-85%. Many patients appear to have 2 clones
of cells, those which take up mIBG and those which have somatostatin receptors
and bind octreotide. Both radiopharmaceuticals may be used to determine metastatic
burden, although 111In-pentetreotide should be the agent of choice if it is available.
Symptoms of carcinoid syndrome may be effectively controlled with subcutaneous
octreotide (Sandostatin). In the absence of 111In-pentetreotide uptake by hepatic
metastases, it is improbable that this will be an effective treatment. We perform
111In-pentetreotide scanning in all patients with metastatic carcinoid prior to
commencing octreotide therapy and exclude those patients who show no uptake.
Figure 18B. Coronal SPECT images showing the ability of SPECT to better delineate
the distribution of tumor within the mediastinum. The arrowhead indicates an apical
pulmonary metastasis.
(which have relatively low sensitivity) in patients with a rising calcitonin or CEA, as
a post operative evaluation or in patients with symptoms.
Neuroblastoma
Neuroblastoma is the commonest solid tumor of childhood and the fourth
commonest pediatric neoplasm. Its incidence is approximately 10 per million. Despite
aggressive combination chemotherapy regimes, the outlook remains bleak for those
presenting with Stage IV disease. Approximately 50% of younger children and 75%
of older children present at this advanced stage.
Tumor Imaging 327
Reprinted with permission from Oxford Medical Publications: Thyroid Cancer and
Endocrine Tumors. In: The Management of Advanced Cancer. Eds: N. Cherny, et
al, In Press.
328 Nuclear Medicine
Figure 19. Anterior and posterior 111In-pentetreotide images of a patient with a non-
functioning neuroendocrine tumor with an unknown primary. Metastases are seen
in the right supraclavicular fossa and the mediastinum (arrows).
15
123
I-mIBG is the investigation of choice for initial staging, restaging after treat-
ment, assessment of post surgical residual disease (Fig. 21), monitoring treatment
response and for the early detection of recurrence on routine follow-up. The sensitivity
and specificity of the technique in these settings are in excess of 94% and 95%,
respectively. It may be the only indicator of bone marrow involvement where a
pattern of diffuse uptake in long bones is characteristically seen (Fig. 22). mIBG
uptake in an abdominal mass in a small child is pathognomonic of neuroblastoma
and can often speed diagnosis and progress to treatment.
Tumor Imaging 329
131
I-mIBG may be used to evaluate these patients but clinical accuracy is reduced
because of technical factors, and small or subtle lesions will be missed; it is not 15
recommended if 123I-mIBG is available.
Pheochromocytoma
Pheochromocytoma is a rare tumor of the sympathetic nervous system which
produces characteristic symptoms of headache, sweating, diaphoresis and palpitations
caused by catecholamine overproduction. Usually they present as unilateral intra-
adrenal tumors, but bilateral tumors, extra-adrenal or pediatric presentations and
metastatic spread occur in about 10% of patients. The diagnostic screening tool is
measurement of urinary catecholamines or metanephrines while CT or ultrasound
will define the presence of a mass.
330 Nuclear Medicine
Figure 21. 123I-mIBG scan of a child with residual abdominal neuroblastoma seen
as activity distributed on both sides of the midline (arrow).
15
Figure 22. 123I-mIBG scan in a five year old child with extensive skeletal metastatic
involvement with neuroblastoma. There is diffuse abnormal activity throughout
the skeleton resulting in an 123I-mIBG scan which looks like a bone scan.
possible technical difficulties such as prior neck surgery; (iii) patients in whom sur-
gery is high risk; or (iv) equivocal ultrasound or CT (sensitivities of 50-75%).
Although thallium imaging with 99mTc-pertechnetate thyroid subtraction was
performed in the past, most centers now use 99mTc-sestamibi, either with 99mTc-
pertechnetate or alone. 99mTc-sestamibi has high sensitivity and specificity (85-95%),
332 Nuclear Medicine
Figure 23. Anterior and posterior 123I-mIBG scan in a patient with a left adrenal
pheochromocytoma (arrowheads).
is technically easy to perform, and has become the investigation of choice. Images
are obtained immediately after 99mTc-sestamibi injection and at about 2-3 hours by
which time washout from normal thyroid tissue will have occurred and activity will
be retained in the parathyroid adenoma (Fig. 25). 99mTc-pertechnetate or 123I imaging
may be useful in defining the functional anatomy of the thyroid gland for comparison
15 with 99mTc-sestamibi abnormalities.
Prostate cancer
The most frequent nuclear medicine examination performed in patients with
prostate cancer is the bone scan which is discussed elsewhere in this handbook.
More recently, there has been interest in the use of a radiolabelled monoclonal
antibody, capromab pendetide (Prostascint), in evaluating patients with: (i) a new
diagnosis of prostate cancer, (ii) those with a high risk of recurrence, or (iii) a rising
PSA with negative bone scan and CT. Several groups have reported a management
benefit when antibody imaging is used as an adjunct to CT, particularly in the first
2 indications. There is a steep learning curve to the interpretation of these images,
Tumor Imaging 333
15
Figure 24. Anterior and posterior I-mIBG images in a patient with advanced
123
but in experienced hands the technique can improve patient management (sensitiv-
ity and specificity of 50-75%).
Radioisotope Therapy
The ability of nuclear medicine to design radiolabelled molecules that will target
specific cancers for diagnostic use raises the intriguing possibility that the same
334 Nuclear Medicine
Tissue (mm)
131
I-Sodium iodide 8d ` 3 Thyroid cancer
131
I-mIBG 8d ` 3 Neuro-
endocrine
tumors
32
P-Phosphate 14.3 d ` 8 Polycythemia
rubra vera
89
Sr-Strontium chloride 50.5 d ` 6.7 Pain palliation
153
Sm-EDTMP 2d ` 3.4 Pain palliation
90
Y-Ibritumomab 2.6 d ` 11 Non-Hodgkins
lymphoma
111
In/90Y-Octreotide 2.6 d b 11 Neuro-
(experimental) endocrine
tumors
Pain palliation
In patients with metastatic cancers of the prostate, lung and breast, bone me-
tastases are seen in over 85%. In most of these, pain will become a dominant man-
agement problem. This pain is often severe and progressive, requiring escalating
doses of narcotic analgesics to achieve palliation. Associated symptoms such as de-
pression, anxiety and fear can also erode patients quality of life. It has been esti-
mated that up to 50% of patients with cancer metastatic to bone have inadequate
control of their symptoms. Table 6 outlines strategies that may be used to achieve
symptom control.
In patients with multiple sites of pain, recurrent pain in a radiation field or with
progressive symptoms, palliation may be achieved by the administration of bone
seeking radiopharmaceuticals, of which strontium-89 (Metastron) is the most
commonly used (Table 5). Strontium-89 is a pure ` emitter with a half life of 50.5
days. Biologically strontium is metabolized by the same pathway as calcium; it is
incorporated into the inorganic matrix and the degree of uptake is proportional to
osteoblastic activity. There is increased retention of strontium-89 in metastases which
show increased uptake of activity on a bone scantypically there is a ratio of about 15
10:1 between metastatic lesions and normal bone.
Prerequisites for the use of strontium-89 have been extensively discussed in the
literature and are summarized in Table 7. Appropriate patient selection is important
to the effectiveness of the treatment. Contraindications to the use of this treatment
for pain palliation have been established (Table 8).
Most efficacy data have been published for patients with prostate and breast
cancer metastatic to bone, although the therapy has been reported as effective in
most primary cancers which have metastases with positive bone scans. Overall
response rate is about 75-80% and approximately one-third of these patients become
pain free. Responses typically take 7-10 days to develop and are sustained, on average,
336 Nuclear Medicine
Reprinted with permission from Mary Ann Liebert, Inc.: Cancer Biotherapy &
Radiopharmaceuticals, 1998; 13(6):413-426
Modified with permission from Mary Ann Liebert, Inc.: Cancer Biotherapy &
Radiopharmaceuticals, 1998; 13(6):413-426
from 3-6 months. Retreatment is possible and is usually effective if the initial treat-
ment was successful.
Toxicity is limited to thrombocytopenia which should not be severe if the pa-
15 tient is appropriately selected. In patients with advanced prostate cancer dissemi-
nated intravascular coagulation (DIC) should be excluded.
Frequently Asked Questions
What nuclear medicine approach can be taken to diagnose
adrenocortical tumors?
The radiopharmaceutical most commonly used is norcholesterol labeled with
131
I (NP-59). When NP-59 is injected intravenously it is incorporated into low-
density lipoprotein (LDL). Cellular uptake is mediated by LDL receptors. Cholesterol
is the normal substrate for adrenal steroid hormone synthesis and the radiolabelled
analogues are esterified, stored in the intracellular lipid pool, but not further
Tumor Imaging 337
Modified with permission from W.B. Saunders Company: Semin Nucl Med, 1997;
27(2):165-182.
metabolized. It therefore provides a model for imaging either increased cellular mass,
increased receptor mediated transport or intracellular storage. This uptake and
retention within the cell is controlled by action of the hypothalmic-pituitary-adre-
nal axis (Cushings disease) or of the renin-angiotesin-aldosterone axis (hyperaldos-
teronism).
By 2-3 days the normal adrenals may be faintly and symmetrically visualized.
Uptake may be quantified by standard nuclear medicine techniques and normal
ranges have been published. Abnormal uptake is seen as intense and progressive
uptake within the adrenal which quantitatively lies outside the established normal
range. Anterior and posterior images of the abdomen are obtained 5 and 7 days after
intravenous injection of the radiopharmaceutical. Images may be acquired after
dexamethasone suppression which suppresses uptake in the normal adrenal but not
in an autonomously functioning gland or adenoma. In this way the detection of
mineralocorticoid and androgen secreting tumors is enhanced.
The demonstration of adrenal masses by abdominal CT is a not uncommon
finding in routine imaging studies. While these might require follow-up for patient
management decisions, radionuclide imaging is most clearly indicated when there is
evidence of adrenocortical hypersecretion based on clinical diagnosis and appropriate
hormonal measurements. Whilst several groups have reported on adrenocortical
scintigraphy in this group of patients with incidentalomas, there is no evidence
15
that there is a routine clinical role for the technique.
The three clinical syndromes for which adrenocortical scintigraphy may be used
are Cushings syndrome, hyperaldosteronism and hyperandrogenism. The imaging
techniques and interpretation for these latter 2 syndromes are identical to that used
in Cushings syndrome but require dexamethasone suppression.
The results of imaging are reported both on the basis of qualitative and quantitative
uptake; the abnormal patterns of uptake seen are shown in Table 9. Figure 26 shows
bilateral adrenal hyperplasia in a patient with Cushings syndrome. Whilst this
technique is not commonly performed, it can provide invaluable data to the surgeon
in planning his surgical approach.
338 Nuclear Medicine
15
Additional Reading
1. Aktolun, Tauxe WN, eds. Nuclear Oncology. Berlin: C. Springer-Verlag, 1999.
Major textbook of the role of nuclear medicine in oncology. Provides most of the key
references required for further reading and expands the scope of discussion on individual
areas.
2. Berman CG, Choi J, Hersh MR, Clark RA. Melanoma lymphoscintigraphy and
lymphatic mapping. Semin Nucl Med 2000; 30:49-55.
Background reading of the theory of the technique, clinical overview of the state of trials
and current indications for use.
3. Hoefnagel CA. Radionuclide cancer therapy. Ann Nucl Med 1998; 12(2):61-70.
General overview of radioisotope therapy with background on most radiopharmaceuticals
in use or in trials.
4. Maisey MM, Wahl RL, Barrington SF. Atlas of Clinical Positron emission
tomography. Arnold Publishing Group, 1999.
The definitive textbook on clinical applications of PET.
5. McEwan AJ. Use of radionuclides for the palliation of bone metastases. Semin
Radiat Oncol 2000; 10:103-114.
Overview of indications for palliative treatment of bone metastases with
radiopharmaceuticals. Review of the agents available, indications and contraindications
and departmental requirements for use.
6. Pauwels EKJ, McCready VR, Stoot JHMB, van Deurzen DFP. The mechanism of
accumulation of tumour-localising radiopharmaceuticals. Eur J Nucl Med 1998;
25:277-305.
The definitive paper of radiopharmaceuticals used in nuclear oncology. Extensive reference
list and background information on all current radiopharmaceuticals and many currently
being evaluated.
7. Phelps ME. PET: The merging of biology and imaging into molecular imaging. J
Nucl Med 2000; 41:661-681.
Past, present and future of PET and molecular imaging with particular reference to
cancer.
8. Pieterman RM, van Putten JW, Meuzelaar JJ, Mooyaart EL, Vaalburg W, Koeter
GH et al. Preoperative staging of non-small-cell lung cancer with positron-emission
tomography. N Engl J Med 2000; 343(4):254-261.
A recent prospective comparison of conventional staging with FDG PET showing that
PET improves the rate of detection of local and distant metastases in patients with non-
small-cell lung cancer.
15
CHAPTER 16
Neuropsychiatric Disorders
Jean-Paul Soucy, Denis Lacroix and Catherine Kissel
Introduction
Nuclear medicine has a lengthy and distinguished reputation in the fields of
neurology and psychiatry as a powerful research and clinical tool. Noninvasive in
vivo demonstration of brain lesions was initially performed more than forty years
ago with agents which accumulated nonspecifically. Since then we have witnessed
an important evolution in the techniques used for the diagnosis of neuropsychiatric
disorders. The traditional brain scan, based on agents sensitive to blood-brain
barrier (BBB) disruption (Fig. 1), has been replaced by anatomically superior
approaches such as x-ray transmission computed tomography (CT) and magnetic
resonance imaging (MRI) for the study of most focal lesions.
More recently, the demonstration by nuclear medicine techniques of disturbances
in regional perfusion patterns, energy consumption and neurotransmission in the
brains of psychiatric patients has made a major contribution to the recognition that
mental diseases (including drug addictions) are organically-based metabolic
alterations, in every way comparable to diseases found in other systems. Presently,
the most frequently performed clinical nuclear medicine test in neuropsychiatric
disorders is SPECT evaluation of cerebral perfusion. These perfusion maps provide
a snapshot of brain activity and are altered in a variety of neurological and psychiatric
conditions even when anatomic imaging is normal.
Until recently in vivo study of neurotransmission was only possible with positron
emission tomography (PET), but new SPECT agents have made this much more
widely available. PET itself is undergoing a major increase in availability through
technological developments that are making it a far less complicated and costly
technique. Neurology and psychiatry can only benefit from this transformation.
Regional Cerebral Perfusion
Non-invasive measurements of regional cerebral blood flow (rCBF) were ini-
tially performed with freely diffusible tracers such as xenon-133 (imaged in planar
or SPECT mode) or 15O-water and 11C-butanol (imaged with PET). Unfortunately,
those techniques had limited availability due to complexity and the high cost of the
required equipment. The development of iodine-123-labelled compounds (123I-IMP,
123I -HIPDM) which have a blood-flow related distribution in the brain afforded an
easy means of evaluating rCBF, not through absolute quantification but through the
description of its relative distribution. Subsequent development in the field of
radiochemistry led to the synthesis of the currently used radiopharmaceuticals, 99mTc-
HMPAO (Hexa Methyl Propylene Amine Oxime, Ceretec) and 99mTc-ECD
Figure 1. Disrupted blood-brain barrier (BBB). Two metastases (left posterior frontal,
A, and right temporal, B) from a lung carcinoma: the tumors are devoid of normal
astrocytes and therefore of an intact blood-brain barrier. This allows passage of
99m
Tc-glucoheptonate, normally excluded from the cerebral parenchyma, into the 16
lesions.
16
Figure 2. Normal brain perfusion. (A) Transaxial, sagittal and coronal slices of a
normal rCBF SPECT study with 99mTc-ECD. Notice preferential perfusion to cortical
and sub-cortical grey matter. (B) Some of the structures which should be routinely
depicted on a state of the art rCBF SPECT study are A: cerebellar hemisphere. B:
temporal lobe. C: brain stem. D: vermis. E: hippocampal region. F: lateral temporal
cortex. G: occipital cortex (calcarine region). H: head of the caudate. I: putamen. J:
thalamus. K: frontal lobe. L: parietal lobe.
Neuropsychiatric Disorders 343
of the molecules in the blood before they reach the brain. The more blood that flows
through a given brain region, the more tracer it receives. Both agents are highly
lipophilic and easily cross the BBB. Once in the brain, they undergo conversion to
polar compounds that cannot diffuse back into the blood. The precise mechanisms
involved are still disputed: in the case of 99mTc-HMPAO a significant part of the
transformation is through reaction of the native molecule with reduced glutathione
in astrocytes, whereas for 99mTc-ECD most of the transformation is through the
action of esterases located both extra- and intra-cellularly.
Four principles must be kept in mind when interpreting SPECT studies:
After first pass uptake, tracer trapped within brain tissue remains essentially
stable over time. One can therefore consider that the SPECT images
produced are a snapshot of the distribution of rCBF at the time of
injection of the tracer.
Although these tracers are highly lipophilic, they are diffusion limited at
high flow rates. Overall, high flow regions accumulate more radioactivity
than low flow ones, but the relationship is not linear: as flow increases,
the relative accumulation decreases. Techniques that correct for this effect
are either cumbersome or inexact. Practically speaking, underestimation
of high blood flow is of limited clinical importance as precise determination
of rCBF is not generally required.
Since 99mTc-HMPAO and 99mTc-ECD do not undergo the same chemical
transformations, it is not surprising that their exact distributions differ
slightly. This is most important for the diagnosis of sub-acute cerebral
infarctions. During the luxury perfusion phase these lesions can show
significant uptake of 99mTc-HMPAO, sometimes to levels that overestimate
actual perfusion, perhaps due to locally increased glutathione
concentrations. In contrast, 99mTc-ECD uptake is decreased in these cases,
probably because the maintenance of the esterase activity which allows
trapping of the tracer is energy dependent.
SPECT rCBF studies represent a technically simple method for assessing
complex neurophysiological events. Whereas it might be simple to
interpret the very focal cortical activation associated with a strong sensory
stimulation (strobe light in the eyes, vibrating ball in the hand) or a
convulsive episode, evaluation of the more elaborate operations involved
in the performance of tasks of greater complexity, cognitive function or
emotion is more daunting. The emerging view of brain function is that
the completion of even simple tasks relies on the activation of large
distributed networks which interact extensively, are shared by a variety of 16
different tasks, and can be efficiently replaced by secondary centres. In
that context, highly specific associations between a given clinical state
and a brain perfusion pattern are the exception rather than the norm.
Decreased rCBF arises through two mechanisms:
Decreased vascular supply: Occlusive vascular disease directly interferes
with nutritive blood flow. SPECT imaging can help to determine the
hemodynamic effect of a vascular lesion.
344 Nuclear Medicine
cerebellum, basal ganglia and thalamus (white matter shows little uptake due to
relatively low levels of blood flow).
More complex approaches for quantification are used in research and in some
clinical settings. These methods often require spatial registration, resizing and warp-
ing of the images in an effort to superimpose different scans from the same patient
(either of the same or different modalities obtained at different times) or from different
patients (fused in order to compare groups of subjects). Most analytical tools were
developed for PET or fMRI studies which have spatial resolution and temporal
sampling capacities quite different from SPECT. Although the same techniques are
applicable to SPECT, they have not been optimised for it and therefore have to be
utilised with a good understanding of their characteristics and limitations.
Clinical Applications of SPECT rCBF Imaging
To date nuclear neuroimaging has been used more for research than for clinical
problem-solving. The clinical usefulness of SPECT rCBF imaging was carefully
reviewed by an expert panel of the American Academy of Neurology. Based upon
quality and strength of evidence they rated various clinical applications (Table 1).
The indications designated as established are few and have not been expanded
since the publication of this report in 1996. Although many other applications are
promising and may eventually be accepted, for the moment caution is recommended
before using rCBF SPECT studies on a systematic clinical basis in these conditions.
Many of these conditions have not been evaluated in sufficient numbers and in
adequately controlled prospective studies to allow firm, evidence-based conclusions
to be drawn on the general usefulness of this application. This should not be construed
to mean that it is useless to perform such studies on patients affected with these
conditions, since many may benefit from the information provided. In some cases,
the perfusion study will be useful in simply confirming an observable organic basis
to symptoms and signs otherwise not linked to recognisable abnormalities on
conventional imaging (CT or MRI). One such application is in the field of head
trauma, where the very nature of some brain lesions (shearing at the microscopic
level) can produce persistent memory and concentration problems despite normal
CT and MRI examinations. Such lesions are expected to alter perfusion, and the
demonstration of cerebral perfusion abnormalities helps to support claims for
compensation or disability payments. An even more direct application is in the
evaluation of vascular spasm following subarachnoid haemorrhage (SAH) where the
cause of a deterioration in the neurological status of a patient with a known SAH is
an alteration in blood flow (Fig 3). The results from SPECT rCBF studies seem to
correlate very well with those of angiography but SPECT is a much less invasive
technique for recognizing this serious complication and evaluating response to therapy.
Other neurological conditions (such as AIDS dementia complex and viral
16
encephalitis) have been reported to show altered patterns of brain perfusion, but the
available evidence remains insufficient at this time to gauge the overall performance
of SPECT perfusion studies in such cases.
The same remarks can be made for psychiatric illnesses, which are also associated
with a variety of abnormal brain perfusion patterns. The observation of these patterns
has been useful in helping to understand the pathophysiology of these afflictions,
but to date they do not allow the use of SPECT rCBF studies as clinical tools for
diagnosis or follow-up. For instance, the reported observations of activation in the
346 Nuclear Medicine
Figure 3. Vasospasm post subarachnoid hemorrhage (SAH). Brain SPECT ECD scans
were performed 2 and 4 days post-SAH. Only the transaxial images are shown.
The initial study shows decreased left frontal perfusion secondary to
intraparenchymal bleeding (arrow). The delayed study performed after the patient
showed severe clinical deterioration reveals perfusion limited to the cerebellum
and the lower brain stem (arrowheads), indicating bilateral cerebral hemispheric
infarction with no possibility of regaining normal cerebral function.
(CT or MRI) relies on the detection of an anatomical lesion which may or may not
be the actual ictal focus. A recent meta-analysis has confirmed the value of ictal
SPECT (and PET) rCBF studies: focally increased perfusion has a sensitivity of
97% and a specificity of at least 98.5% for temporal lobe epilepsy localisation (Fig.
4). Although the most useful part of a SPECT evaluation is the ictal study, an interictal
one (obtained if possible with EEG monitoring to exclude subclinical ictal activity
at the time of injection) is warranted in order to facilitate interpretation of the ictal
scan. Unlike the ictal scan, the one obtained interictally demonstrates decreased
perfusion of the epileptogenic focus. Simple visual inspection of the two studies can
be supplemented with statistical analysis that looks for regions of significantly
increased perfusion in the ictal phase. The interictal study alone, however, cannot be
relied upon to localise epileptic foci (sensitivity only 50% for temporal lobe epilepsy
with false localisation rate of 10%).
For ictal studies, injection must be performed immediately after seizure onset, as
rapid modification in the distribution of brain activation and perfusion can occur.
For extra-temporal epilepsy, generalisation can be extremely rapid. Frontal lobe
epilepsy requires an injection delay of no more than 10 seconds after clinical initiation
of the seizure before epileptic activity spreads to other parts of the brain. In temporal
lobe epilepsy the optimal interval for injection is during the first 30 seconds, although
acceptable studies can be obtained for up to 60 seconds. If a first try is negative then
repeat examinations must be obtained: the importance of precisely determining which
area of the brain to remove can hardly be overemphasized. Its non-invasiveness, 16
relatively low-cost, and remarkable performance amply justify the use of rCBF SPECT
in the preoperative assessment of epilepsy.
Figure 4. Epilepsy. Interictal and ictal transaxial, coronal and sagittal slices showing
right temporal focus of decreased interictal perfusion and ictal hyperfusion. These
images are 3D co-registered, allowing for precise subtraction (Ictal-Interictal) and
superimposition over an MRI atlas of voxels with a subtraction value of more than
2 SD over the mean.
Alzheimers disease, diffuse Lewy body dementia, frontotemporal dementia and vas-
cular dementiaaccount for 90% of all cases.
SPECT rCBF imaging can be a valuable adjunct to clinical evaluation and con-
ventional neuroimaging (CT and MRI) in patients with dementia. Certain scan
patterns increase the likelihood of specific sub-types of dementia (Fig. 5), information
that is valuable in establishing prognosis. This may also help in targeting drug
therapies, a potentially useful application now that several medications have been
approved for the treatment of Alzheimers disease with additional agents undergoing
clinical trials. However, SPECT interpretation must be integrated with all other
information on the patient: on its own it can neither be used to prove nor disprove
the presence of dementia, nor to establish a specific diagnosis of a sub-type of
dementia, as patients with no clinical signs of dementia can have perfusion studies
with a pattern suggestive of Alzheimers disease (bilaterally decreased parietotemporal
perfusion), and conversely some patients with documented dementia can retain a
normal perfusion pattern. Special reconstruction techniques along the hippocampal
16 long axis can reveal severe hypoperfusion of these structures. Some investigators
believe that SPECT can help to differentiate nondementing illnesses (e.g., late-onset
depression) from organic dementia, though others have questioned this.
Table summarizes data from multiple studies, usually small case series.
350 Nuclear Medicine
stroke. Since reduced blood flow is the primary event, SPECT rCBF studies become
abnormal immediately and are only limited by the size and location of the affected area.
The clinical usefulness of SPECT rCBF to demonstrate early interruption in
blood flow to the brain is still unclear as no obvious impact in terms of prognostic
information or patient management has emerged. This situation may change if it
can be shown to help guide patient selection for therapies such as thrombolysis.
Successful lysis of blood clots obstructing cerebral blood vessels can dramatically
improve post-stroke functional status if it is achieved in the first few hours. This
benefit is partially offset by a significant risk of intracranial hemorrhage. This
complication might be predicted by SPECT rCBF studies: the presence of large
areas of profoundly decreased perfusion reveals tissue that has been submitted to
severe ischemia, and in which vascular cells may have been irreversibly damaged.
Reflow into such vessels is prone to intraparenchymal bleeding with a high rate of
complications related to acutely increased intracranial pressure. Inclusion of SPECT
rCBF studies in stroke protocols to evaluate suitability for thrombolysis will depend
on clinical confirmation that it reliably predicts which patients are at increased risk
of bleeding (Fig. 6). In theory, it should be easy to detect the large, severe decrease in
tracer uptake that would portend a poor outcome.
Energy Metabolism and Neurotransmission Studies
Although still largely limited to research applications, direct measurement of
energy metabolism and of specific binding of radioligands to a variety of molecular
targets in the brain are emerging as potential clinical applications in neuropsychiatric
nuclear medicine. As more agents aimed at specific molecular targets become available,
nuclear medicine will be able to characterise brain metabolism and neurotransmission
with ever greater precision, thereby shedding light not only on the mechanisms
underlying the diseases studied but also contributing to the development of new
therapies and allowing disease diagnosis through molecular characterisation.
Energy Metabolism Studies
Glucose Consumption
Cerebral metabolism is mainly assessed with the most frequently used
radiopharmaceutical in PET, 18F- fluorodeoxyglucose (FDG). This glucose analogue
is transferred across the BBB and taken up by cellular glucose transporters (largely
on the cell membranes of astrocytes). Once inside the cell, 18F-FDG is phosphorylated
to 18F-FDG-6-PO4 but cannot be further metabolised by the enzymes involved in
glycolysis. As there is also very little phosphatase activity in the brain, the 18F-FDG-
6-PO4 compound is trapped where it was initially taken up. Mathematical modeling
16 of the combined imaging and arterial blood measurements allows calculation of the
regional cerebral metabolic rate for glucose (rCMRGlu).
As with blood flow, 18F-FDG uptake most closely parallels the distribution of
glutamatergic transmission. 18F-FDG PET studies also benefit from the much greater
spatial resolution and quantitative capacities of PET as compared to SPECT. On
the other hand, 18F-FDG PET studies have low temporal resolution, since
accumulation of the tracer in brain tissue must proceed for up to 45 minutes before
acquisition can begin: the resulting data reflect the integrated activity of the brain
over that period of time. This is not a problem when studying a disease or a
Neuropsychiatric Disorders 351
16
Figure 5. Alzheimers disease. (A) SPECT brain perfusion in a patient presenting
with dementia and moderate diffuse atrophy on CT scan. This single transaxial
slice shows marked hypoperfusion bilaterally at the level of the parieto-occipital
junctions (arrows). (B) 18F-FDG PET study (transaxial images) showing marked
symmetrical glucose hypometabolism in the posterior parieto-temporal cortex
(arrowheads). (Case provided by Dr. D. Worsely.) These scan patterns are typical of
Alzheimers disease.
352 Nuclear Medicine
Figure 6. Thrombolysis assessment. Transaxial slice from patient with acute non-
hemorrhagic infarction and a limited perfusion defect (arrows). Thrombolysis with
tPA should be considered within 3 hours of symptom onset when the CT scan is
normal. Large or multiple lesions are probably contraindications to thrombolysis
due to an unacceptably high risk of hemorrhagic transformation. (Case provided
by Dr. W.D. Leslie.)
phenomenon which is stable over such a time frame, but is obviously a drawback for
activation studies where it is often difficult to perform a given task for such a long time.
Oxygen Consumption
Cerebral energy metabolism can also be assessed through measurements of mo-
lecular oxygen (O 2) consumption. Oxygen (as 15O-O 2) is administered by
inhalation and measurement of its accumulation is used to determine the regional
cerebral metabolic rate for oxygen (rCMRO2). This can aid in determining whether
decreased rCBF results from a primary vascular problem (obstruction) or decreased
metabolic demands (parenchymal dysfunction). By simultaneously measuring rCBF
and rCMRO2 it is possible to obtain an oxygen extraction fraction (OEF). A normal
OEF signals a match between needs and supply, and implies parenchymal
hypoactivity; an increase indicates abnormally decreased perfusion with neural tissue
attempting to extract more oxygen from a decreased supply. One day this approach
may help select patients for a cerebral revascularisation procedure.
Clinical Applications
Excellent results have been obtained in studying epilepsy and the dementias (where
results are even better than with SPECT perfusion studies) with 18F-FDG PET
imaging. It also can help guide biopsies, establish prognosis, evaluate recurrence of
brain neoplasms, and differentiate residual masses from other conditions such as
radiation necrosis (see section on Intracranial Mass Lesions). The high levels of cor-
tical glycolysis normally observed can mask detection of low-grade brain tumors.
Studies of Neurotransmission
Nuclear medicine is still the only means for studying neurotransmission directly
and non-invasively in vivo. A broad array of SPECT and PET tracers are available,
including neuroreceptor ligands (generally antagonists), neurotransmitter metabolic
precursors, ligands of plasma membrane and synaptic vesicle transporters, substrates
of neurotransmitter catabolising enzymes and components of intracellular
transduction chains (Fig. 7). PET has long dominated this landscape, but over the
past ten years progress in the field of radiopharmacy has contributed many SPECT
tracers.
Mathematical modeling of dynamic PET or SPECT acquisitions permits
estimation of physiologically relevant parameters such as neuroreceptor densities for
a large variety of chemically defined transmission systems, synthetic and catabolic
enzymatic activities within those systems, numbers of neuronal terminals of a given
nature, neurotransmitter concentrations in the synaptic cleft and even second
messenger generation. Data acquired in these domains has revolutionised our
understanding of neurologic and psychiatric diseases.
Clinical Applications
While very few clinical applications are presently established for neurotransmitter
studies, their potential for growth is probably the greatest in all of neuropsychiatric
nuclear medicine. One suggested use is in the monitoring of therapy for psychoses.
Classical neuroleptics, such as the phenothiazines, block the D2-subtype of the
dopamine receptors. These receptors exist in high concentration in the striatum and
are intimately involved in regulation of motor activity. Excessive blockade (beyond
80-85%) of striatal D2 dopamine receptors has been associated with an increased
risk of developing long term side effects such as tardive dyskinesia. The level of D2
blockade can be measured with both PET and SPECT ligands, and patients found
to have excessive blockade may benefit from a lower dose of medication or from an
atypical neuroleptic with minimal affinity for D2 receptors.
Predicting response to dopaminergic therapy in patients with a parkinsonian
syndrome might also be possible. The striatal complement of D2 dopamine receptors
is normal (or even increased through receptor up-regulation) in Parkinsons disease 16
since the lesion resides in the nigro-striatal fibres (Fig. 8). In diseases such as progressive
supranuclear palsy or syndromes producing primary striatal degeneration the D2
dopamine receptor concentration is reduced. The latter are not responsive to
dopaminergic therapy since the defect is on the receiving end of the dopaminergic
transmission process. This could be quite useful since even in specialised centres
Parkinsons disease and parkinsonian syndromes are confused in up to 20% of patients.
354 Nuclear Medicine
16 Physiology
The ventriculo-subarachnoid system consists of two fluid compartments: the
intracranial component occupies 140 ml (about 25 ml of which is intraventricular)
and the spine is bathed with an additional 75 ml. Daily production of CSF reaches
500 ml, 90% of which is produced by the choroid plexi. The remainder comes from
exudation through intraparenchymal vessels, reaching the subarachnoid space by
diffusion along the spaces of Virchow-Robbins, and from extracellular fluid reaching
the ventricular system across the highly permeable ependyma. Production is largely
Neuropsychiatric Disorders 355
through arachnoidal invaginations into the lumen of the sinus (called Paccionian
granulations). A small portion of the CSF is reabsorbed across the arachnoid into
blood vessels along the neuraxis. Progression of the CSF along these different paths
is ensured by a pressure gradient between the production and resorption sites, vascular
pulsations, and intracranial and intraspinal pressure waves of multifactorial origin.
Nuclear Cisternography
The circulation of the CSF can be studied with nuclear cisternography in which
an inert radiotracer (usually 99mTc- or 111In-DTPA) is injected into the lumbar
subarachnoid space. Sequential scintigrams of the spine and head are obtained for
24-72 hours. A normal study shows rapid progression of radioactivity from the
injection site towards the head, with intracranial activity noted anywhere between 1
and 3 hours post-administration. Activity progressively permeates all of the
intracranial subarachnoid space without significant penetration into the ventricular
system (though transient but minimal ventricular penetration is not unusual). By
24 hours, most of the activity should have been transferred to the superior sagittal
sinus and excreted through the kidneys with most remaining intracranial activity
superior to the cerebral convexities (Fig. 9).
Clinical Applications
Hydrocephalus
Since an increase in CSF pressure does not block its formation, CSF can
accumulate intracranially with symptoms that vary depending on the speed at which
the build-up occurs. For instance, malformations, hemorrhages or tumors in the
posterior fossa can obstruct the egress of CSF from the ventricular system, giving
rise to a condition known as non-communicating hydrocephalus (i.e., no
communication between the ventricular system and subarachnoid space). More
commonly, scinticisternography is performed for suspicion of communicating
hydrocephalus. After such events as meningitis, subarachnoid hemorrhage or
intracranial surgery, the subarachnoid space may become obstructed, impeding the
normal flow of CSF. This situation can give rise to normal pressure hydrocephalus
(NPH). The term NPH is a misnomer: by the time the disease comes to clinical
attention spinal opening pressure may be normal because compensatory mechanisms
have come into play to reduce intracranial pressure. Before this stage is reached,
CSF pressure is probably periodically increased, resulting in ventricular dilatation.
The classic clinical triad of NPH is dementia, gait disturbance and incontinence.
The dementia is potentially reversible, and in a small number of carefully selected
individuals there is a dramatic response to CSF shunting.
16 The major compensatory mechanism to NPH is a marked increase in
transependymal passage of CSF into brain parenchyma. This is accompanied by an
inverted flow of CSF from the subarachnoid space into the ventricles. This
phenomenon is well depicted by scinticisternography, where marked ventricular
activity will accumulate and persist for a sustained period of time. This pattern is
considered confirmatory of the clinical diagnosis of NPH. Most neurologists do not
recommend performing this test on a routine basis, reserving it for atypical cases
that do not manifest the classic clinical presentation. It has been suggested that
reflux of radioactivity into the ventricles that lasts for more than 24 hours indicates
Neuropsychiatric Disorders 357
that compensatory reabsorption is not well established and that deterioration is likely
(Fig. 10). This may indicate a better response to ventriculo-peritoneal shunting than
when only transient reflux is observed, since the latter suggests a stabilised condition
with established, nonreversible damage to the periventricular brain tissue. In fact,
no test (including scinticisternography) is particularly successful at predicting the 16
response to surgery, emphasizing the importance of recognizing the typical clini-
cal presentation.
CSF Leaks
The second major indication for scinticisternography is in the identification of a
CSF leak. Such leaks can occur after skull fractures, destructive infections or
neoplasms, surgical interventions, and radiation therapy. This can lead to repeated
358 Nuclear Medicine
Figure 11. Brain mass lesion with 99mTc-sestamibi uptake. This 39 year old HIV
positive man was admitted to hospital with dyspnea and fever in association with
Staphylococcus aureus bacteremia and then developed new-onset seizures. A CT
scan (left) showed an enhancing mass lesion in the periventricular white matter of
the left parietal-occipital lobe (arrow) measuring approximately 3 cm in diameter
with peritumoral edema. The radiologic fetatures were most consistent with CNS
lymphoma, with toxoplasmosis considered a less likely alternative. SPECT sestamibi
brain scanning (right) confirmed increased uptake in the lesion (arrow) further
supporting a diagnosis of lymphoma which is almost uniformly sestamibi-avid while
toxoplasmosis is usually negative. (Case provided by Dr. W.D. Leslie.)
Figure 12. Brain mass lesion without 99mTc-sestamibi uptake. This 28 year old man
with HIV (CD4 count only 6) presented to hospital with a two week history of
gradually progressive left-sided weakness (arm worse than leg) and left homonymous
hemianopsia. The patient had declined anti-retroviral therapy, with the result that
he had numerous infectious complications from HIV. A CT scan showed decreased
attenuation in the right temporal lobe white matter without contrast enhancement,
hemorrhage or significant mass effect. An MRI scan (left) showed extensive white
matter changes in the right temporal lobe extendng into the right internal and external
capsule, the right frontal lobe, the right midbrain and the right cerebellar peduncle.
There was no evidence of gadolinium enhancement or mass effect. Differential
diagnosis was felt to be most likely progressive multi-focal leukoencepholopathy
or non-enhancing lymphoma. Brain SPECT (right) was performed with 99mTc-
sestamibi and failed to show any uptake (arrow). This is strong evidence against
CNS lymphoma which is typically strongly avid for flow tracers such as sestamibi
and thallium. PML is more variable in its appearance, since it can show absent or
increased uptake, but is strongly supported by the absence of uptake in this case.
The patients clinical condition deteriorated and he died with a tentative diagnosis
of PML without a post-mortem examination. (Case provided by Dr. W.D. Leslie.)
Tumor Grading
PET and SPECT imaging with 18F-FDG, thallium-201, 99mTc-sestamibi or 99mTc-
tetrofosmin also seem to offer prognostic information on the behaviour of brain
tumours Those with the highest uptake tend to have a higher grade and more
16 aggressive behavior. Early imaging after the initiation of therapy has been reported
to be a good predictor of the final therapeutic response, and PET/SPECT is superior
to CT or MRI for this purpose.
Somatostatin Receptors
Finally, 111In- or 99mTc-pentetreotide, somatostatin analogues used essentially for
the diagnosis of neuroendocrine tumours, can be taken up by the somatostatin
receptors found on the cells of gliomas, glioblastomas, meningiomas and possibly
schwannomas. These agent cannot cross the intact BBB and will not detect lesions
Neuropsychiatric Disorders 361
with somatostatin receptors that do not disturb the BBB, such as low-grade gliomas.
Tumours which are outside of the BBB, such as meningiomas, consistently
concentrate the tracer.
Conclusions
The full scope of nuclear medicines role in the clinical evaluation of brain disorders
is still evolving. Some applications have emerged as clearly useful, but others must
be used in a prudent manner until there is additional documentation of their efficacy.
Research in psychiatry, neurology and nuclear imaging should help us to characterise
diseases affecting the human brain at their most fundamental, molecular level. Nuclear
medicine imaging techniques are uniquely suited to the in vivo, non invasive
measurement of such parameters. Neuropsychiatric nuclear medicine is probably at
the threshold of an explosion in its clinical use.
Frequently Asked Questions (FAQs)
What is a shuntogram?
A shuntogram is a nuclear medicine procedure used in patients who have
previously had a ventriculo-peritoneal or venticulo-atrial shunt inserted. The tubing
may become obstructed at different levels (intracranial segment, extracranial
subcutaneous tubing, reservoir, reservoir exit valve, or subcutaneous tubing to the 16
peritoneal cavity or left atrium). It is simple to inject radioactive tracer directly into
the reservoir and then follow its migration. Depending upon shunt design, the
reservoir may optionally be pumped to facilitate CSF movement. The pattern of
CSF flow can be used to deduce the site of obstruction, if any (Fig. 13).
How is a brain death study done and what is its accuracy?
The development and evolution of the concept of brain death has been necessary
due to our technologic advances in medical care and organ transplantation. The
362 Nuclear Medicine
Figure 14. Brain death. This young woman sustained anoxic brain injury while
recovering from emeregency surgical replacement of a mitral valve prosthesis
infected with Staphylococcus aureus. Cardiorespiratory arrest led to 18 minutes of
anoxia before circulation could be restored and she remained in a comatose state.
Neurological assessment of brain stem function strongly suggested brain death, but
a confirmatory test was requested before discontinuation of support. The patient
was injected with 99mTc-ECD in the intensive care unit and scanned in Nuclear
Medicine shortly thereafter. The scan demonstrates only scalp and facial flow with
an empty skull. There is absent perfusion of the cerebral cortex and brain stem.
(Case provided by Dr. W.D. Leslie.)
current definition of brain death describes the clinical state of totally absent central
nervous system function in a hemodynamically stable, normothermic, nonintoxicated
patient, which is followed inevitably by cardiovascular collapse. Diagnosis is based
on coma, absent brain stem reflexes, and apnea (despite a documented pCO2 of >
60 mm Hg), with selective use of confirmatory testing. Under normal conditions,
the presence of an electrocerebrally inactive EEG is a valid indicator of brain death.
However, in some situations (such as in the presence of high doses of sedative/hypnotic
medications), the EEG can be unreliable. Complete cessation of cerebral perfusion
as demonstrated by brain scintigraphy, transcranial Doppler sonography, or cerebral
panangiography is also evidence of brain death. Contrast angiography is least desirable
since it is invasive and the contrast exposure can threaten subsequent organ harvesting.
Numerous studies confirm the reliability of brain scintigraphy in the diagnosis of
brain death. The great majority of patients (98.5% in one large series) judged to be
brain dead by other criteria show absent brain blood flow (empty skull) on
conventional radionuclide angiography or rCBF imaging (Fig. 14). Even if an
16 initial examination shows some preserved blood flow, a repeat study 72 hours later is
usually diagnostic. There are no reported cases of neurological recovery following a
definite scintigraphic diagnosis of brain death.
Can rCBF studies be used in brain activation studies?
Activation studies involve having the patient use the function being studied while
changes in rCBF are measured as a marker of the structures being activated or
inhibited. The design of such protocols is a highly specialised field calling for close
collaboration between imaging specialists, statisticians, and clinical scientists from
Neuropsychiatric Disorders 363
Figure 15. Activation study. Baseline 99mTc-ECD SPECT scans were obtained at rest
with eyes closed. The patient was then exposed to bright, colored, rapidly moving
geometric shapes for 10 seconds prior to and for 3 minutes after a second injection
of 99mTc-ECD. Statistical maps were obtained after spatially coregistering the studies,
subtracting the baseline from the activation study and superimposing activated
voxels (subtraction value more than 2 SD over the mean) on an MRI atlas. Note
intense activation of calcarine visual cortex regions.
neurology, psychiatry and psychology. Presently, fMRI, by virtue of its high tempo-
ral and spatial resolutions, is considered by many to be the tool of choice for such
studies, and in large measure has displaced PET for this purpose. However, SPECT
with either 99mTc-HMPAO or 99mTc-ECD, has an advantage over both of these
methods: as these radiopharmaceuticals capture the distribution of cerebral blood
flow at the time of injection and allow for delayed imaging, the patient can undergo
complex activation protocols using instrumentation that either cannot be brought
close to the powerful magnetic field of a fMRI scanner or that cannot be physically
accommodated by PET or fMRI systems. In SPECT, the subject can be prepared
with a simple intravenous line and then submitted to the activation protocols during 16
which she/he is injected with the radiopharmaceutical at the moment of maximal
stimulation of the structures involved in the task (Fig. 15). Later, after completion
of other measurements, the patient can be brought to the imaging laboratory for the
acquisition phase of the study. SPECT rCBF is the only non-invasive approach
capable of this type of intervention.
364 Nuclear Medicine
Additional Reading
1. Talbot PR, Lloyd JJ, Snowden JS, Neary D, Testa HJ. A clinical role for 99mTc-
HMPAO SPECT in the investigation of dementia ? J Neurol Neurosurg Psychiatry
1998; 64:306-313.
A well written paper proposing a framework within which rCBF studies could be used
in the evaluation of dementia.
2. Assessment of brain SPECTReport of the Therapeutics and Technology
Assessment Subcommittee of the American Academy of Neurology. Neurology
1996; 46:278-285.
A must read reference despite its publication date: it covers all of the present and
foreseeable indications of rCBF studies.
3. Alexandrov AV, Masdeu JC, Devous MD, Black SE, Grotta JC. Brain single-photon
emission CT with HMPAO and safety of thrombolytic therapy in acute ischemic
stroke. Stroke 1997; 28:1830-1834.
This paper describes what could possibly be one of the most important indications for
rCBF studies in the coming years, and provides a good basis for standardizing
interpretation.
4. Sperling RA, Sandson TA, Johnson KA. Functional imaging in Alzheimers disease.
In: Scinto LFM, Daffner KR, eds. Early Diagnosis of Alzheimers Disease. Totowa:
Humana Press, Inc., 2000.
A very up-to-date review of the field.
5. Juni JE, Waxman AD, Devous MD, Sr., Tikofsky RS, Ichise M, Van Heertum RL
et al. Procedure guideline for brain perfusion SPECT using technetium-99m
radiopharmaceuticals. J Nucl Med 1998; 39:923-926.
The official basic technique. A useful introduction but like every recipe, variations
based upon experience will enrich it.
6. Devous MD, Thisted RA, Morgan GF, Leroy RF, Rowe CC. SPECT brain imaging
in epilepsy: A meta-analysis. J Nucl Med 1998; 39:285-293.
A very strong paper that makes it clear why the previous paper endorsed rCBF studies
for epileptic focus localization.
16
CHAPTER 1
CHAPTER 17
an imaging stretcher or couch. Children are best immobilized with restrainers strapped
around the stretcher top using Velcro straps.
Sedation
If moderate restraint is not successful then sedation is required for a technically
satisfactory study. This is especially true for lengthy procedures such as SPECT or
whole body imaging. Sedation is recommended in overly anxious patients who refuse
to cooperate, very young or hyperactive patients who are unable to remain still, and
retarded patients who lack the mental capacity to follow simple instructions.
Sedation may take several forms. In the correct setting, sedation with intravenous
pentobarbital sodium (Nembutal) is very effective. Using a dose of 5 mg kg-1 (to a
maximum of 100 mg), we administer half the volume rapidly, wait 60 seconds and
then administer a quarter of the dose. This usually puts the child to sleep in about
two to three minutes. If not, the remaining quarter dose is given. The child remains
asleep for about 45-60 minutes. The advantages of this technique are that the child
falls asleep very quickly, the effect is more reliable than with intramuscular injection,
and the child recovers faster. As with all sedatives, cardiac and respiratory status
must be closely monitored by appropriately trained individuals. We currently use an
automated pulse oximeter.
Pentobarbital is contraindicated in neonates less than 2 months of age. This
group lacks adequate levels of the liver enzymes required to metabolize pentobarbital.
In place of pentobarbital we use an elixir of promethazine (Phenergan) and chloral
hydrate administered 30-45 minutes before scanning. While the effect is less
pronounced than with pentobarbital, it is usually adequate and the patient arouses
readily. Midazolam (Versed) orally is useful in some cases.
Diazepam (Valium) and pentobarbital suppositories have all been found to be
inadequate. None produces the deep sleep required to perform nuclear medicine
procedures on the patient.
Injection Techniques
Radiopharmaceutical injection in small children presents several minor difficulties,
all of which are easily overcome by modifying adult techniques. Children may be
completely covered by the head of a large field of view camera. This increases the
childs anxiety and makes injection difficult. This is resolved by placing the child
supine on the gantry or stretcher top and having the camera underneath. In smaller
children, for head, feet or hand imaging, holding the child directly on the camera
collimator is a very good technique.
Finding a suitable vein for intravenous injections is rarely a problem. Although
the antecubital fossa frequently has the largest vein, the elbow is less easily immobilized
than the hand or foot which are the preferred injection sites. With procedures
requiring repeated blood samplings (e.g., GFR determination), the insertion of an
17 intravenous catheter is the best approach.
Radiopharmaceutical Dosage
The amount of radioactivity administered can be readily calculated by referring
to a chart, using body surface area estimated from the patients weight and height.
The percentage of the standard adult dose is then determined according to the patients
Pediatric Nuclear Medicine 367
body surface area. This dose calculation is preferred to using weight alone since it
results in equal tracer concentration per unit area of planar organ imaging.
It is very important to establish a minimum dose for each radiopharmaceutical.
To get an adequate study, especially a dynamic one, there have to be enough photons
detected to adequately assess the patients problem. It is better to accept a slightly
higher delivered radiation dose than to have an uninterpretable study. This is especially
true in children where one wants the imaging time to be as short as possible in order
to avoid motion. The risk of the higher radiation dose is negligible, especially if we
have contributed to an accurate diagnosis.
Image Evaluation
It is mandatory to view all images on the imaging workstation using its windowing
capability. In musculoskeletal imaging, it is very important to see the physes (growth
plates) clearly as well as the diaphyseal long bones which can only be properly
performed with multiple window level settings. Reviewing a dynamic study on the
workstation is valuable. For example, in voiding cystography a whiff of reflux not
apparent on hard copy is easily appreciated on the monitor display.
Clinical Role in the Assessment of Childhood
Musculoskeletal Disorders
In benign bone diseases the two major indications for nuclear imaging are pain
and fever/infection. Plain film examination should be carried out first. Frequently,
CT is done to detect and characterize a lesion. The usual bone scanning technique is
to acquire blood pool images and static or SPECT images of the areas of clinical
interest. The highest resolution system available is used to obtain the best quality
images in small children and infants. Magnification with converging or pinhole
collimation is frequently used.
Osteomyelitis
The child with osteomyelitis may present with joint pain or tenderness, limited
range of motion, soft tissue swelling, erythema, fever or bacteremia. Differentiation
of osteomyelitis from cellulitis or septic arthritis may be difficult clinically, and
unfortunately, the plain film examination is often non-diagnostic. Frank radiographic
bone changes of osteomyelitis are rarely seen due to early diagnosis and treatment.
The combined use of blood pool and bone scintigraphy makes it possible to
differentiate osteomyelitis from cellulitis.
The typical appearance of osteomyelitis is a well-defined focus of increased bone
metabolism associated with an identical area of hyperemia in the blood pool images
(Fig. 1). This is most often located in the metaphysis of a long bone. Occasionally
other bones are involved, especially if a puncture wound has occurred. This appearance
is specific for osteomyelitis and is readily differentiated from the patterns of cellulitis
and septic arthritis. However there is some evidence that the technique has less 17
sensitivity in infants of less than 44 weeks gestational age.
The hyperemia involving the metaphysis of the long bone is usually seen between
24 and 36 hours after the onset of symptoms, whereas bone scintigraphy usually
becomes positive between 36 and 72 hours. If the findings are not typical of
osteomyelitis then gallium imaging is used to determine whether or not infection is
present (Fig. 2).
368 Nuclear Medicine
Figure 1. Acute osteomyelitis. (A) Anterior blood pool of the knees, (B) delayed
anterior bone scan of the knees and (C) delayed medial bone scan of the right knee
following injection with 99mTc-MDP. There is increased blood pool activity and
delayed uptake in the metaphysis of the right femur (arrows). Note the loss of
demarcation of the physis and the metaphysis. The location and appearance of the
abnormality are typical of acute osteomyelitis.
Cellulitis
Cellulitis has a distinctive appearance, which is a diffuse increase in radioactivity
involving both the soft tissues and the bone. This is more readily apparent in the
blood pool images than in the bone images. The appearance is due to a diffuse soft
tissue hyperemia and is readily distinguished from the appearance described above
for osteomyelitis (Fig. 3).
Septic Arthritis
Septic arthritis has a similar appearance to that described for cellulitis but the
hyperemia involves both sides of the joint. Subchondral bone on either side of the
affected joint has marked increased metabolism. These investigations are usually
carried out in conjunction with ultrasound which is very sensitive for detecting the
effusions associated with septic arthritis. The two studies should be interpreted
together.
Legg-Perthes Disease
Avascular necrosis of the femoral head (i.e.., Legg-Perthes disease) is usually
detected radiologically. In those patients where the radiographs are normal or show
mild capsular distention, the use of MRI has become the main means of evaluating
the hip. Bone scintigraphy can be used either as an adjunct to determine degree of
viability or to detect avascular necrosis in the presence of an inconclusive MRI study.
Bone scintigraphy, especially when using pinhole magnification, remains an excellent
17 technique for detecting avascularity of the femoral capital epiphysis (Fig. 4).
Non-Accidental Trauma
Our socio-medico-legal system has developed an enlightened sensitivity to child
abuse. The estimated incidence of reported child abuse has increased from 30/1000
in 1985 to 45/1000 in 1992. The incidence of skeletal injury in these children is
approximately 20% and is more common among those under 1 year of age. Fractures
Pediatric Nuclear Medicine 369
Figure 2. Cellulitis and early osteomyelitis. (A) Blood pool image, (B) delayed bone
scan image and (C) gallium image. There is increased blood pool activity and
increased gallium uptake (arrows) in the right mid-leg. While on the delayed bone
scan image, mild diffusely increased activity is present in the right mid-tibia related
to the adjacent soft tissue inflammation, there is a focal abnormality suggestive of
osteomyelitis in the right proximal tibial metaphysis that also has increased gallium
accumulation (arrowheads). Osteomyelitis was confirmed by aspiration.
are usually multiple, involving the long bones, skull, vertebrae, ribs, and facial bones,
and usually show different stages of healing. In most pediatric hospitals, the approach
is to radiograph the clinically affected body parts. If the suspicion of child abuse
arises then a total body radiographic examination is done. After this a total body
bone scan is performed to locate any unsuspected bony injuries. Metaphyseal-
epiphyseal injury is common. Careful positioning and correct window level settings 17
are important to avoid the pitfall of blooming in the image, which may obscure a
mild abnormality. Accurate interpretation will depend on assessment of the intensity
and shape of the abnormality. Prime sites of investigation are the ribs, costovertebral
junctions, spine, and diaphyses of long bones (Fig. 5). If the bone scan is normal 3
days or more after injury, then the probability of bony injury is very low. Disadvantages
370 Nuclear Medicine
Figure 3. Pneumococcal cellulitis after trauma from a hockey puck. (A) Blood pool
image and (B) delayed bone scan of the lower legs. Increased blood pool activity is
seen (arrow). While there is mild diffusely increased uptake in the left tibia, no
focal area of increased uptake is seen.
of the bone scan include: inability to determine the type, extent, and age of each
injury; poor bone accumulation of the 99mTc-MDP in cases of severe malnutrition;
and inability to differentiate systemic and metabolic disorders associated with trauma
from trauma alone.
If one or more areas are found to be abnormal then high detail, multiple view
radiography is performed to confirm the abnormality and to help date it. With
greater recognition of this problem, more studies are being performed. Skull
radiography is best performed to detect trauma to the calvarium, as the detection of
fractures of the flat bones of the skull by bone scan is very poor.
Clinical Role in Childhood Malignancies
Osteogenic Sarcoma
Osteogenic sarcomas usually occur in the metaphyses of long bones, a common
site being the distal femur or proximal tibia. The common age range is from 10 to
17 years. Occasionally, several sites may be involved at the time of presentation. The
plain film always is abnormal but the appearance can mimic other diseases such as
chronic osteomyelitis.
Untreated osteogenic sarcoma has a typical bone scan appearance. It demonstrates
an uneven intensely increased distribution of tracer in the metaphysis. The blood
17 pool image shows increased blood volume within the tumor. The flow study however
has a markedly increased blood flow when compared to what would be expected
from the blood pool image. This is due to the presence of arteriovenous connections
with direct arteriovenous shunting (Fig. 6).
Pediatric Nuclear Medicine 371
The use of bone scanning in the follow-up of the primary tumor is limited since
the scan remains positive because of ongoing bone remodeling. There is a good
correlation however between the initial uptake of thallium-201 or 99mTc-sestamibi
and the histological response to treatment. A reduction in thallium or sestamibi
uptake after treatment indicates a response to chemotherapy.
Recently, there have been attempts to predict the response to chemotherapy by
17
measuring the washout of 99mTc-sestamibi from the tumour. P-glycoprotein is present
in cell membranes and protects cells from a variety of chemotherapeutic drugs by
facilitating their washout from the cell. Those tumor cells with a high level of P-
glycoprotein will be protected from the accumulation of chemotherapeutic agents.
372 Nuclear Medicine
Figure 5. Non-accidental trauma. (A) Anterior bone scan of the legs with (B) posterior
view of the upper body. The child refused to move the left leg. Intense uptake
related to a fracture of the distal left tibia is readily seen (A, arrow). An unexpected
fractue of the distal left humerus was also detected (B, arrow).
Figure 6. Osteogenic sarcoma. (A) Anterior flow image, (B) anterior blood pool
image and (C) anterior delayed bone scan of the knees. Arrows show site of
involvement in left distal femur. The blood pool and bone scan images both have
decreased activity centrally at the site of the increased flow on the frame from the
radionuclide angiogram. This indicates arteriovenous shunting in the core of the
rapidly growing sarcoma which is typical of osteogenic sarcoma. The blood pool
image does not capture the shunting but does reflect the bone repair in response to
the tumor.
Figure 7. Lymphoma. Anterior (left) and posterior (right) whole body gallium scans
showing abnormal uptake in both hila (arrowheads). There is normal accumulation
in the skeleton, liver and lacrimal glands.
17
neonatal hepatitis there is typically reduced hepatic extraction and delayed clearance
with radiotracer appearing in the bowel within 24 hours. We have found that although
no neonate with biliary atresia will excrete the radiotracer, a significant number of
babies with neonatal hepatitis will also show no excretion. Therefore, while the
Pediatric Nuclear Medicine 375
Figure 8. Neuroblastoma. (A) 99mTc-MDP whole body bone scan and (B) 123I-MIBG
images of the knees. On the bone scan, abnormal uptake is particularly evident in
the long bones where increased metaphyeal uptake (arrow) causes loss of definition
of the physis. While the bone scan reflects the response of normal bone to tumor
infiltration, the 123I-MIBG scan shows tumor uptake in the left distal femur, left tibia
and right femur (arrowheads).
17
presence of tracer in the bowel eliminates the possibility of biliary atresia, the con-
verse is not true as neonatal hepatitis, bile duct paucity syndrome and total parenteral
nutrition (TPN) can cause absent excretion of the radiotracer into the bowel. It is
important to premedicate the patient for five days with phenobarbital (2.5 mg kg1
376 Nuclear Medicine
twice daily for five days) before injecting the biliary radiopharmaceutical. This ap-
proach enhances the likelihood of biliary excretion in cases of neonatal hepatitis.
Clinical Role in Rectal Bleeding
The etiology of rectal bleeding in infants and children is often obscure. In a 10
year period from 1952 to 1962, 801 patients were admitted to our institution with
17 a history of rectal bleeding. Sixty-one patients, in whom all examinations were
negative, underwent a laparotomy. A Meckels diverticulum was present in 24, of
which 20 contained ectopic gastric mucosa. Meckels diverticulum is a common
congenital anomaly (0.3-3% in mature individuals); however, a relatively small
number of these patients (approximately 20%) have ectopic gastric mucosa within
their Meckels diverticulum.
Pediatric Nuclear Medicine 377
Figure 10. Biliary atresia. Anterior images following the injection of 99mTc-
mebrofenin. While activity accumulates in the liver, the biliary tree is not visualized
nor is there excretion into the gastrointestinal tract up to 24 hours. Faint activity in
the left abdomen at 4 and 24 hours represents renal accumulation.
Figure 11. Meckels diverticulum. Anterior images at (A) 5 minutes and (B) 30
minutes. 99mTc-pertechnetate is seen in the stomach as well as in the Meckels
diverticulum in the right lower quadrant (arrows).
regional enteritis produces abnormal scans but the distribution of activity is diffuse
or the time sequence is different from that in the stomach.
Our results suggest that the pertechnetate abdominal scan detects ectopic gastric
mucosa with an accuracy of at least 95 percent. This is a significant improvement
over previously published reports, which we attribute to the combination of sequential
early imaging, multiple views at 15 and 30 min, and the use of ranitidine.
Clinical Role in Genitourinary Disorders
Pelviureteric Obstruction
The child with pelvi-ureteric obstruction is usually evaluated pre- and post-
operatively for the degree of relative renal function, absolute renal function and rate
of egress of the radioactive urine from the kidneys. Interestingly, many young children
with antenatal hydronephrosis have increased renal mass on the side of the
hydronephrosis. This has been confirmed in 99mTc-DMSA studies carried out
subsequent to the original diuretic washout study. In cases where an obstructive
component to urine flow is suspected, a furosemide (Lasix) stress test is begun
between 10 and 30 minutes after the injection of 99mTc-DTPA or 99mTc-MAG3.
Intravenous furosemide (1 mg kg-1 to a maximum of 20 mg) is given and images
recorded for 20 min into the computer. The clearance rate is calculated using the
computer. This provides a quantitative method for determining the effect of surgery
and/or post-operative complications on the renal drainage system. Normally the
time to empty one-half of the collecting system activity (T 1/2) is less than 8-10
min. This value may be higher (up to 20 minutes) in grossly dilated pelves or after
surgery. In cases where the T1/2 is above 10 minutes, retention at 20 minutes is
17 calculated (counts at 20 minutes divided by the counts at time 0) (Fig. 12). If the
patient has a T1/2 of 20 minutes then the retention value at 20 minutes is 50%. The
furosemide washout study appears to reflect the state of urinary flow better than the
Whitaker test in the post-operative period.
In children who have had repair of a pelviureteric junction obstruction or
reimplantation of the ureters as an antireflux surgical procedure, there may be a
significant delay of about six months before there is a change in relative renal func-
Pediatric Nuclear Medicine 379
Figure 12. Diuretic renal scan. (A) delayed posterior images, (B) time activity curves
of left renal pelvis and right renal pelvis and (C) ROI around left kidney. 99mTc-
DTPA was injected and an initial set of images acquired (not shown). After a delay
to allow for filling of the left renal pelvis, furosemide was injected. Visually, the
right renal pelvis empties completely and the left renal pelvis partially. The left
kidney T 1/2 is 32 minutes with 67% retention at 20 minutesvalues that are in
the obstructed range.
tion. When evaluated after that time, the kidneys usually function and drain nor-
mally.
Several important points have to be made regarding the assessment of diuretic
washout curves. Firstly, the T1/2 is invalid when performed on normal kidneys that
have already emptied their pelves. Secondly, when there is poor renal function, the
kidneys cannot be evaluated as there may be little, if any, response to the furosemide
stimulation. Lastly, in large hydronephrotic kidneys of greater than 70 ml capacity,
there may be a significant delay in emptying even though there is no mechanical
obstruction.
Urinary tract infection
Radionuclide techniques are the primary imaging modalities in evaluating children
with suspected upper urinary tract infections (UTI). The direct radionuclide
cystogram is the most sensitive method for detecting refluxa condition which is
often a prelude to an upper UTI. Direct radionuclide cystography entails bladder
catheterization and instillation of saline mixed with a non-absorbable tracer such as
99m
Tc-MDP. The detection rate of reflux is as good as or better than that of the
voiding cystourethrogram (VCUG) while the radiation dose is significantly less than
that of the radiological study. In addition, the bladder and ureters are constantly
monitored throughout both filling and emptying, something not feasible using
17
fluoroscopy. Recently, computer analyzed antegrade voiding cystograms have been
studied and appear to be almost as accurate as the retrograde version. In this indirect
technique, a renal scan is performed and once the kidneys have emptied and bladder
filled, continuous imaging is carried out during voiding. This may become the
preferred procedure as it is more physiological. However, due to possible pelvic
380 Nuclear Medicine
Figure 13. Bilateral vesico-ureteric reflux to the level of the renal pelves is clearly
demonstrated.
17
Figure 14. Pyelonephritis. 99mTc-DMSA study with planar imaging in the (A) posterior
and (B) right posterior oblique projections. A defect is seen in the upper pole of the
right kidney (arrows).
Pediatric Nuclear Medicine 381
Figure 15. Renal scar. 99mTc-DMSA scan with (A) planar quantitation, (B) coronal
SPECT and (C) sagittal SPECT images. Scarring is seen in the lower half of the
kidney (arrows).
dilatation and thus retention of radioactive urine (which could mask reflux), the
most widely accepted method is still direct radionuclide cystography (Fig. 13).
If an upper UTI is suspected, the child must be evaluated to determine whether
or not acute pyelonephritis is present. If the child is sick enough to be admitted to
hospital, then a 99mTc-DMSA SPECT study should be performed to determine
whether or not pyelonephritis is present. Ultrasound should also be performed to
see if there is dilatation of the renal pelvis. Finally, a cystogram should be performed
to determine if reflux is present. If the child is suspected of having an upper UTI,
but is not sufficiently ill to be admitted to hospital, the first study should be a
cystogram and if reflux is demonstrated, this should be followed by ultrasound and
DMSA studies. Similarly, if there are repeated UTIs at the time of presentation, this
patient should be investigated in the same manner.
Acute pyelonephritis appears as a non-segmental single or multifocal reduction
in cortical accumulation of the 99mTc-DMSA (Fig. 14). In the acute phase, it is
sometimes difficult to do SPECT imaging and in these cases planar imaging with
oblique views is a satisfactory substitute. Subsequent 99mTc-DMSA scans to monitor
the progress of acute pyelonephritis should probably be done using SPECT imaging.
Our experience is that while SPECT imaging does not increase the sensitivity of
diagnosis in acute pyelonephritis, it is useful in defining the location, size and
multiplicity of lesions. A major concern after an episode of acute pyelonephritis is
the development of scars. In this case, SPECT imaging appears to be the method of
choice. Not only is it easier to define whether scars are present or not, but location
and definition are also enhanced (Fig. 15). In addition, it is much easier to determine
whether an apparent defect toward the upper pole of the left kidney is a true scar or
17
secondary to splenic impression. Cortical mantle thickness is also much more readily
defined with SPECT.
382 Nuclear Medicine
17
APPENDIX
Index
230 Iodine metabolism 263
Fracture 43, 93, 96, 97, 100, 101, 103, Iodine therapy 23, 267, 276, 280, 286
105-119, 122-128, 130, 131, 134, Islet cell tumors 324, 327
136, 145, 149, 156, 158, 336, 357,
368, 370, 373 L
G Leukocytes 29, 130, 131, 134,
136-140, 174, 233-238, 242, 243,
Gamma cameras 10, 11, 13 245, 247-249, 252, 255-258
Gastric motility 199 Lung cancer 156, 229, 254, 302, 312,
Gastroesophageal reflux disease 313
(GERD) 197, 199, 200, 208-210 Lung perfusion 12
Gastrointestinal bleeding 196, 204, Lung ventilation 79, 80, 82
253 Lymphoma 158, 172, 200, 212, 249,
Gastroparesis 199, 200, 202, 203 250, 270, 273, 302, 306, 314, 315,
Glomerular filtration 163, 165, 185 317, 319, 320, 335, 358, 359, 361,
Goitre 264, 266, 267, 270, 273, 275, 372, 374
277, 278, 281, 282 Lymphoscintigraphy 306, 320
H M
Hepatitis 63, 243, 276, 373-376 Medullary thyroid cancer 264, 273,
Hepatocellular adenoma 229 275, 301, 302, 325
Hepatoma 212, 213, 226, 229, 230 Melanoma 270, 303, 306, 320-322
History 1, 45, 53, 57, 84, 108, 110, Metaiodobenzylguanidine (MIBG)
111, 115, 116, 135, 158, 186, 199, 264, 274, 275, 301, 373, 375
203, 204, 208, 219, 252, 254, 264, Myocardial infarction 31, 50-54, 68,
269, 282, 284, 361, 376 74
Hydrocephalus 349, 356, 358 Myocardial viability 38, 54
Hydronephrosis 378, 172, 174, 176,
178, 181, 195 N
Hyperparathyroidism 97, 128, 158,
273, 330 Neonatal hepatitis 373-376
Hypothyroidism 22, 200, 269, 277, Neonatal jaundice 373
278, 282, 286, 288, 295 Neuroblastoma 301, 326, 328, 331,
334, 372, 373, 375, 377
I Neuroendocrine tumors 306, 324
Neurotransmission 340, 344, 350,
Infection 4, 12, 76, 125, 126, 128, 130, 353, 354
134, 136, 137, 156, 159, 172, 206, Nuclear imaging 12, 174, 191, 232,
208-210, 213, 220, 234, 237, 239, 255, 305, 361, 367
243-245, 248-253, 255-258, 268, Nutcracker esophagus 196, 199
301, 322, 357, 367, 379, 382
388 Nuclear Imaging
O Q
Octreotide 264, 275, 287, 295, 302, Quantitative ultrasound (QUS)
306, 324, 335 99-101
Orthopedic prostheses 245
Osteitis deformans (Pagets disease) R
Index
158, 161
Osteoblast 93, 94, 116, 118, 121, 122, Radiation protection 16, 24, 25, 276,
126, 139, 146, 148, 158, 159, 335 279
Osteoclast 93-95, 113, 115, 121, 124, Radioiodine 276, 282, 21
126, 152, 153 Radioiodine 8, 21, 23, 27, 265, 267,
Osteomyelitis 122, 129-131, 136, 139, 269, 272, 276, 279, 280, 282, 286,
140, 158, 237, 252, 254, 257, 258, 293
367-370 Radionuclide imaging 59, 74, 297,
Osteoporosis 96, 100, 105, 108-113, 306, 330, 337
116, 118, 119, 122, 133, 158, 161 Radionuclide production 5
Osteosarcoma 149-151 Radionuclide therapy 20, 27, 276
Radionuclides 5, 8, 13, 19, 27, 97,
P 195, 211, 266, 279, 299
Red blood cell labelling (RBC) 62-64,
Pain palliation 334, 335, 337 69, 204, 211, 212, 227-229, 232
Parathyroid adenoma 330, 332, 335 Reflex sympathetic dystrophy (RSD)
Pediatrics 373 133-135
Pentetreotide 212, 214, 264, 274, 275, Renal failure 163, 172, 174, 179-181,
292, 294, 299, 302, 305, 315, 322, 188, 221, 249, 288
324, 326, 328-330, 335, 360 Renal infarction 174, 191
Peptic ulcer disease 200, 206, 208, Renal physiology 163, 170, 180
327 Renal scarring 165
Pheochromocytoma 273, 275, 301, Renal trauma 166, 195
330, 332-334 Renovascular hypertension 163,
Positron emission tomography (PET) 179-183, 187, 188
5, 10, 13, 14, 38, 54, 57, 243, Risk stratification 50-52, 57
293-295, 297, 299, 300, 305, 306,
308-310, 313, 314, 316-318, S
320-322, 340, 345, 347, 350, 352,
353, 359, 362, 363 Safety 16, 19, 21, 25, 26, 88, 279, 281,
Pregnancy 22, 23, 26-29, 268, 276, 289, 291, 364,
277 Scintimammography 308-310
Preoperative assessment 347 Scleroderma 172, 196, 199, 200
Prostate cancer 332, 336 Sedation 366
Pseudoarthrosis 129, 130 Selective estrogen receptor
Pulmonary aspiration 208, 382 modulators 113, 114
Pulmonary embolism 23, 75, 76, 83, Sentinel lymph nodes 321
84, 86, 88, 92 Sestamibi 25, 29, 35, 37-39, 41, 46,
Pulmonary hypertension 71, 75, 88 50, 51, 54, 63, 151, 263, 264, 274,
Pyelonephritis 165, 167, 172, 174, 294, 299, 302, 303, 310, 311, 331,
249, 256, 381 332, 335, 358, 359, 361, 371, 372
Shin splints 122, 128
Index 389
Single photon emission computer Thyroid 6, 18, 20-22, 97, 111, 115,
tomography (SPECT) 5, 13, 116, 128, 153, 158, 196, 200, 201,
38, 39, 41, 43, 46, 48, 49, 56, 57, 260-273, 275-279, 281-288, 290,
73, 74, 82, 126, 127, 132, 148, 293-295, 301-304, 311, 325, 327,
151, 167, 214, 215, 225-230, 237, 330-332, 334, 335
242, 245, 301-305, 316, 322, 323, Thyrotoxicosis 265, 267-269, 275,
Index
326, 340, 343-350, 352, 353, 358, 276
359, 361-363, 366, 367, 372, 380, Thyroxine 22, 260, 262, 264, 265, 267,
381 275, 286, 294
Somatostatin 200, 212, 214, 264, 275, Trauma 1, 84, 89, 118, 119, 122, 125,
287, 295, 302, 324, 327, 360 126, 128, 131, 133, 143, 146, 156,
Stochastic 20-23, 25, 28, 29 158, 159, 166, 195, 204, 225, 230,
Stress fractures 126 285, 345, 347, 349, 368, 370, 371,
Stress testing 51 373
Synovectomy 137-140 Tubular secretion 163, 165
T U
Tetrofosmin 35, 38, 41, 299, 302, 303, Unstable angina 31, 34, 36, 51, 52, 54
358, 359 Urea breath testing 206
Thallium-201 305, 358, 359, 371 Urinary tract infection (UTI) 379,
Thromboembolic disease 75, 89 381
Thrombus imaging 82
Thyroglobulin 158, 260, 263, 264, V
286, 295
Valvular heart disease 60, 70, 71
Vascular graft infection 237, 243, 252
Ventricular function 54, 55, 60, 64,
68, 70-74
Vesicoureteral reflux 174