Академический Документы
Профессиональный Документы
Культура Документы
Clyde A. Helms, MD
Professor of Radiology and Orthopaedic Surgery
Department of Radiology
Duke University School of Medicine
Durham, North Carolina
1600 John F. Kennedy Blvd.
Ste 1800
Philadelphia, PA 19103-2899
No part of this publication may be reproduced or transmitted in any form or by any means, electronic or
mechanical, including photocopying, recording, or any information storage and retrieval system, without
permission in writing from the publisher. Details on how to seek permission, further information about
the Publisher’s permissions policies and our arrangements with organizations such as the Copyright
Clearance Center and the Copyright Licensing Agency, can be found at our website: www.elsevier.com/
permissions.
This book and the individual contributions contained in it are protected under copyright by the Publisher
(other than as may be noted herein).
Notices
Knowledge and best practice in this field are constantly changing. As new research and experience
broaden our understanding, changes in research methods, professional practices, or medical treatment
may become necessary.
Practitioners and researchers must always rely on their own experience and knowledge in evaluating
and using any information, methods, compounds, or experiments described herein. In using such
information or methods they should be mindful of their own safety and the safety of others, including
parties for whom they have a professional responsibility.
With respect to any drug or pharmaceutical products identified, readers are advised to check the
most current information provided (i) on procedures featured or (ii) by the manufacturer of each
product to be administered, to verify the recommended dose or formula, the method and duration
of administration, and contraindications. It is the responsibility of practitioners, relying on their own
experience and knowledge of their patients, to make diagnoses, to determine dosages and the best
treatment for each individual patient, and to take all appropriate safety precautions.
To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors,
assume any liability for any injury and/or damage to persons or property as a matter of products
liability, negligence or otherwise, or from any use or operation of any methods, products, instruc-
tions, or ideas contained in the material herein.
Printed in China
v
Foreword to the Second Edition
Clyde Helms, as a radiologist in training during increasingly large number of young radiologists
the mid-1970s, was, in many ways, the ideal radi- (his residents and former residents) will attest that,
ology resident. Able, perceptive, informed, and indeed, they do work. The teaching may be un-
responsible, he progressed through the various orthodox, but the learning is real and substantial.
stages of residency training in superb fashion. On This volume is also unorthodox. Several excel-
the other hand, he was different. Whereas the lent, superbly researched and crafted treatises on
traditional “best resident” always had at hand ex- skeletal radiology are available to the radiology
haustive lists of differential diagnoses, Dr. Helms resident and practicing radiologist. This volume
quietly ignored the trivial, the esoteric, and the is not intended as an exhaustive compendium of
information that was not likely to serve him in his skeletal radiology. Rather it is, as indicated by the
work as a radiologist in the “real world,” some- title, an exposition of the basics of skeletal radiol-
times to the discomfiture of the radiology faculty ogy. In keeping with his personalized, unusual
of the University of California, San Francisco approach to teaching, he begins with a discussion
(of which I was then a junior member). Whereas of radiologic examinations that should not be per-
the traditional “best resident” was suitably in formed. The remainder of the book deals with
awe of the faculty (many of whom were truly skeletal conditions that radiologists are likely to
awesome), Dr. Helms fearlessly challenged what encounter any day of the week. The reader who
he perceived as unsupportable dogma. Not one to wishes to become familiar with Scheie syndrome
sit still for pretension, he poked gentle (and or trichorhinophalangeal dysplasia type II must
sometimes not so gentle) fun at the faculty. look elsewhere.
Occasional pranks were perpetrated, sometimes Rather than the usual, formal language found
at the expense of members of the faculty. No one in other radiology texts, the reader will encounter
was immune, no matter how lofty his status. the vernacular used by all radiologists when they
Dr. Helms, as the Sinatra song goes, did it his discuss their work with other radiologists. The
way. Irreverent, witty, occasionally outrageous, text is much like Dr. Helms himself—witty,
and a superb radiologist, he completed his irreverent, unpretentious, and fast-paced. The
residency and went on to fulfill his military reader will find the book refreshing, eminently
commitment. readable, and highly informative.
He returned to UCSF 3 years later as a faculty
member in the skeletal radiology section. That he The ideal condition
had not changed was immediately apparent. Would be, I admit, that men should be
Faculty meetings are disrupted by his irreverent right by instinct;
remarks, frequently hilarious. Now a mentor, his But since we are all likely to go astray,
teaching reflects the same realistic, nontraditional
The reasonable thing is to learn from
approach he used as a resident. He emphasizes not
the exotic or esoteric but the practical, the infor- those who can teach.
mation that is critical in the day-to-day practice of —SOPHOCLES
radiology. Incorporated into his teaching method
are mnemonics, one or two of which might not Clyde Helms can teach!
be recommended for family viewing. On the
other hand, if they work, why not use them? An Hideyo Minagi, M.D.
vi
Preface
The fourth edition of this book, much like the it was irreverent and superficial, but now,
third, shows little substantial change from prior 25 years later, those criticisms wouldn’t bother
editions (how many ways can you say or illus- me. In fact, this book is irreverent and superfi-
trate that the damn bone is broken?); however, it cial! It’s meant to be. It’s not meant to be
has refinements and subtle corrections that make disrespectful to a profession that is worthy
it a more accurate reflection of my experience as of the highest respect – it’s merely a style of
a bone radiologist for the past nearly 35 years. teaching that makes learning easier for many of
The chapters on magnetic resonance imaging us. Once the basics of a subject are mastered,
(MRI) have been considerably updated to reflect one can delve more deeply into a topic. Other
the ongoing improvements in that area. books have been written to address that aspect
In keeping with the previous editions and of bone radiology. Use this book to get started
my basic personality, I have kept the style of and to have some fun while you learn. This
writing very informal and casual. When the book is purposely short and succinct, which is
first edition was written I was concerned that hard to accomplish with such a weighty topic.
my peers and mentors would have judged me To paraphrase Mark Twain, “I would have
to be too irreverent and superficial in dealing written a shorter book if I had more time.”
with a very serious and weighty subject. To my Enjoy it.
surprise the first edition was well received and
became quite popular. I’m sure many did think Clyde A. Helms
vii
CHAPTER 1
Unnecessary Examinations
Before beginning to learn how to interpret There is virtually no finding on a skull series that
pathologic skeletal films, it is important to briefly will alter the next step in the patient’s workup.
consider unnecessary skeletal radiographic ex- Presence or absence of a fracture should not in-
aminations. Dr. Ferris Hall from Boston first fluence whether the patient receives a computed
brought to my attention the idea that just be- tomography (CT) scan or a magnetic resonance
cause we could x-ray something didn’t mean that imaging (MRI) examination. A CT or MRI scan
we should. His article titled “Overutilization of is obtained for other reasons: continued uncon-
Radiologic Examinations” in the August 1976 sciousness or focal neurologic signs. The plain
issue of Radiology1 details many examples of over- films only delay the eventual diagnosis, and in a
use and misuse of radiologic examinations. This patient with a subdural or an epidural hema-
article, even though it is more than 35 years old, toma, that delay could be fatal.3 The mortality
and a similar one by Dr. Herbert Abrams in the from intracranial bleeding is significantly in-
New England Journal of Medicine2 should be man- creased as the time to surgical decompression is
datory reading for every intern before he or she increased; therefore any delay caused by obtain-
begins to order examinations. ing unnecessary examinations (skull films) is
There are many reasons why it is undesir- potentially harmful. There are no findings on a
able to have unnecessary radiologic examina- plain skull series to indicate (or not indicate)
tions: excess cost, excess radiation, waste of the subdural or epidural hematoma (Figure 1-1).
patient’s time, waste of the technician’s and Fewer than 10% of patients with fractures have
radiologist’s time, and false hopes and expecta- subdural or epidural hematomas, and up to 60%
tions based on the outcome of the examination. of patients with subdural or epidural bleeding
In addition, and not least of all, they indicate a have no fractures.4 Therefore, why order the
breakdown in the logical thought pattern con- examinations? Medicolegal reasons? On the
cerning the patient’s workup. contrary! It is well documented that delays in
Many examinations are ordered because of diagnosis in this setting can be fatal, so ordering
so-called medicolegal considerations. It is be- unnecessary examinations might in fact be ask-
lieved that if a certain finding (e.g., a broken rib) ing for a lawsuit. The American College of Ra-
is not documented, the physician could be sued. diology has published appropriateness criteria
In fact, few, if any, examples of medicolegal for when to order particular examinations and
“covering yourself” types of examinations are has endorsed CT scans of the head as the initial
valid. With the move toward greater consumer study of choice in trauma.5
awareness, lawsuits in the future are more likely Despite much documentation in the radiology
to result from unnecessary radiation exposure and emergency department literature that show
because of needless examinations rather than skull films’ lack of utility in trauma, they still are
from too few examinations. commonly routinely ordered in many emergency
departments throughout North America. A sur-
vey performed in 1991 by Hackney and published
EXAMPLES OF UNNECESSARY in Radiology6 reported that more than 50% of the
EXAMINATIONS hospitals in the study “often or always” obtained
skull films for trauma. Every hospital had CT
available. What are they thinking? Obviously
Skull Series they are not thinking about what a skull film will
Except for a depressed skull fracture or the pres- show them that might affect their treatment, be-
ence of intracranial metallic fragments, there is cause it won’t change a thing whether it is positive
no reason to order a skull series for trauma. This or negative.
was once one of the most abused examinations in
radiology, costing millions of dollars per year
unnecessarily. Although the number of unneces-
Sinus Series
sary skull films has decreased, they remain a It is true that an opaque sinus or an air-fluid level
costly burden in many emergency departments. can be seen in a sinus series when sinusitis is
Copyright © 2014 Elsevier, Inc. All rights reserved. 1
2 1 Unnecessary Examinations
FIGURE 1-1 ■ Skull fracture. A thin radiolucent line char- FIGURE 1-2 ■ Waters’ view of the sinuses. This film is
acteristic of a skull fracture is noted (arrow) extending obtained with the patient’s head tilted slightly up-
obliquely across the temporal bone. A fracture in this ward (as if he or she were drinking water—apologies
area is often associated with an epidural hematoma to Dr. Waters). It is an excellent film to obtain when
because the middle meningeal artery lies here. This the maxillary sinuses need to be seen. When done
finding by itself, however, has little or no significance in an upright position, air-fluid levels can be seen
and must be correlated with clinical findings. (arrow).
present. However, the patient with these find- CT scan to search for additional fractures might
ings is often asymptomatic, and just as often, the be in order, but not a nasal series.
sinus series is interpreted as normal in another
patient who has typical clinical findings of sinus-
itis. Both of these patients are treated based on
Rib Series
their clinical, not radiographic, presentation, Fractured ribs are commonly seen in any radio-
which is appropriate. Therefore the information logic practice. The significance of the finding
from the sinus series is ignored. If that is the of a fractured rib or ribs is not well appreciated
way you practice—and many recommend that as by most physicians. If the truth be known, the
being proper—don’t order the sinus series: treat finding of a rib fracture after trauma has almost
the patient. Reserve the sinus series for the no clinical significance and does not alter treat-
patient who doesn’t respond to treatment or has ment. One must rule out a pneumothorax and
an unusual presentation. Also, if it is only sinus- even a lung contusion, both of which are uncom-
itis you are concerned with, most times a simple mon and are best done on chest films, not a rib
upright Waters’ view (Figure 1-2) to examine series. In older patients with chest wall pain
the maxillary and frontal sinuses, rather than a and rib fractures from undetermined causes, it
full sinus series, will suffice, saving money and is extremely difficult and often impossible to dif-
decreasing the patient’s exposure to radiation.7 ferentiate a pathologic rib fracture through a
metastatic focus from a posttraumatic rib frac-
ture. Hence, obtaining x-ray films in a patient
Nasal Bone with focal rib pain to find a fracture serves little
A nasal series is often requested to see if a patient purpose other than to find a cause for the pain.
has suffered a broken nose after trauma to the Most rib series can be eliminated without chang-
face. What if the nasal bone is fractured? It ing the way the patient is treated.
won’t be casted. It won’t be reduced. In other
words, no treatment will be given regardless of
what the x-ray shows. Therefore don’t order the
Coccyx
films in the first place. Occasionally a nasal bone Although not a common x-ray examination, we
is badly enough displaced to warrant interven- have occasional requests to x-ray the coccyx
tion, but even then an acute, posttraumatic x-ray to rule out a fracture. As with the nasal bone
study adds nothing for the patient except ex- and ribs, a fracture in this location will not be
pense and radiation exposure. A facial series or a casted or reduced. Also, this examination has
Copyright © 2014 Elsevier, Inc. All rights reserved.
1 Unnecessary Examinations 3
significantly more gonadal radiation dose than renal osteodystrophy to look for Looser’s frac-
a rib or nasal series. Because no treatment is tures, brown tumors, and subperiosteal bone
predicated on the x-ray results, don’t order the resorption. Most institutions have replaced the
x-ray study for routine trauma to the coccyx. bone survey with hand films, which is preferable
in regard to patient expense and radiation dose.
Subperiosteal bone resorption is seen earliest
Lumbar Spine and easiest on the middle phalanges, radial sides
Plain films of the lumbar spine are probably the (Figure 1-3), and is virtually pathognomonic for
most abused examinations in radiology. They hyperparathyroidism. Looser’s fractures are rare
give the highest gonadal radiation dose of any and not treated. Brown tumors are uncommon
plain film examination, and in most cases they and also are not treated. Therefore, if no treat-
offer no diagnostic information that will be acted ment is based on the x-ray findings, the survey
on by the physician. A significant number of only satisfies curiosity and is not worth the pa-
lumbar spine films are done in persons younger tient’s money or radiation exposure.
than age 40 who have acute onset of back pain
after lifting or straining. There is virtually no
plain film x-ray finding in this patient subgroup
Metastatic Bone Survey
that can be responsible for the acute problem or Little useful information is obtained from the
that can be treated with intervention. Even the majority of metastatic bone surveys. Occult le-
severest spondylolisthesis cannot unequivocally sions that are not found on radionuclide bone
be said to be the origin of the symptoms. Disc scans are seldom encountered. Radionuclide
herniation cannot be identified. Tumors and in- scans are more effective at picking up most
fections are not clinical considerations in this metastatic lesions and could be substituted for
setting. Treatment invariably consists of rest, bone surveys with less cost and better diagnostic
nonsteroidal antiinflammatory drugs (NSAIDs), yield.11 Many investigators believe that search-
overall relaxation of the muscle groups, and then ing for bone metastases is not warranted in every
flexion and extension exercises to strengthen the patient with a primary tumor unless finding
muscles. Radiographs have nothing to offer un- metastatic disease (mets) will obviate surgery or
less the pain is very atypical or the clinical picture otherwise change the patient’s therapy. Radionu-
is clouded by other considerations (e.g., intrave- clide bone scans with x-rays of questionable or
nous drug use, in which case infection must be clinically suspicious areas makes more sense than
ruled out). a complete metastatic bone survey. An exception
The gonadal radiation dose from a lumbar
spine film is the same as that from a daily chest
x-ray for 6,8 16,9 or 98 years,10 depending on
which study you choose to believe. These studies
were based on a three-view lumbosacral spine
series and do not include the oblique views rou-
tinely obtained in many practices. Subtle osseous
changes found on oblique views are thought by
many orthopedists to be insignificant in most
cases anyway.
When should a lumbosacral spine series be
ordered? In cases of severe trauma, possible pri-
mary or metastatic tumor, and possible infec-
tion. Acute low back pain with radicular signs is
no indication for a spine series. An MRI scan
will show disc herniation and would be the pre-
ferred examination over plain films if clinically
warranted. Generally a lumbar spine MRI ex-
amination is performed only after a failed course
of conservative therapy if disc disease is clinically
suspected.
FIGURE 1-3 ■ Phalanges with subperiosteal bone resorp-
Metabolic Bone Survey tion. Subperiosteal bone resorption is seen as a subtle
irregularity or interruption of the cortex. It is best seen
Many institutions routinely order metabolic bone on the radial aspect of the middle phalanges (arrows)
surveys in patients with hyperparathyroidism or and is pathognomonic for hyperparathyroidism.
Ankle Series
The most common cause for presentation to
emergency departments in North America is an
ankle sprain, with more than 30,000 ankle sprains
reported each day.12 Ligamentous injuries can
easily be clinically differentiated from significant
fractures. One study reported that a 50% reduc-
tion of ankle films resulted in no fractures being
missed when the radiology resident simply exam-
ined the patients.13 Another study revealed that if
the patient were able to walk three steps immedi-
ately after the injury or during the examination
in the emergency room there was almost zero
chance of a fracture.14 This study was one of sev-
eral to use what has been called the Ottawa ankle
rules for when to obtain ankle x-ray films. They
are so named after the hometown of the first
authors to implement them and are in wide- FIGURE 1-4 ■ Ankle after trauma. Calcific densities
spread use today in most emergency departments around the ankle that can be mistaken for avulsions
are often seen (arrow). When rounded and smoothly
in North America. Small bony avulsions receive corticated, as in this example, they are either acces-
the same treatment as ligament tears and are sory ossicles or old avulsions. An acute avulsion is
often difficult to differentiate from accessory best diagnosed clinically by noting point tenderness
ossicles (Figure 1-4). Therefore in most cases the at ligament insertion sites. Because a ligament can
x-ray film is not a factor in determining the avulse with or without a fragment of bone being
attached, the x-ray finding will not influence the
patient’s treatment and should be skipped. patient’s treatment.
Lumbar Myelograms
One of the most painful radiologic examina- (Figure 1-6). We can hope that the myelogram
tions extant is the lumbar myelogram, in which will go the way of the pneumoencephalogram
a spinal needle is placed into the subarachnoid and the epidural venogram.
space of the lumbar spine and contrast material So far as choosing between a CT and an MRI
is injected (Figure 1-5). Although this is done scan of the lumbar spine for disc disease and spi-
for tumors, it is most commonly performed in nal stenosis, an MRI examination will give much
the workup of lumbar disc disease. Many studies more diagnostic information and is considered
show that a CT or MRI scan of the lumbar the state-of-the-art imaging examination for the
spine is more accurate than myelography in di- spine, although, as the next section shows, an
agnosing disc disease and emphasize that a CT MRI scan of the lumbar spine is one of the most
or MRI scan should be the study of choice. overused imaging tests in the country.
Many surgeons, however, still request myelo-
grams in addition to the CT or MRI study when
only the CT or MRI need be performed. In ad-
MRI Lumbar Spine
dition to being painful, the myelogram pro- Much has been written in the lay press in the
duces side effects in some people that can be past few years concerning the overuse of medical
pronounced and debilitating; the myelogram testing at a time when our economy is reeling
occasionally necessitates overnight hospitaliza- from skyrocketing medical costs. One of the
tion; the radiation dose from the myelogram is imaging tests mentioned near the top of every
higher overall than with CT; and perhaps most list is an MRI study of the lumbar spine. Multi-
important, the myelogram is not as accurate and ple studies have shown that as many as one third
does not give as complete a picture of additional of asymptomatic individuals older than age 50
back structures as the CT or MRI examination will have one or more focal disc protrusions.15
Copyright © 2014 Elsevier, Inc. All rights reserved.
1 Unnecessary Examinations 5
CT scan unless there is malalignment. A diag- 3. Seelig JM, Becker DP, Miller JD, et al: Traumatic acute
nosis of ligamentous disruption requires that a subdural hematoma. N Engl J Med 304:1511–1518,
1981.
flexion-extension plain film or an MRI scan be 4. Masters JS, McClean PM, Arcarese JS, et al: Skull x-ray
done. Some say obtaining a CT scan on every examinations after head trauma. N Engl J Med 316:84–91,
patient in whom only a plain film is currently 1987.
obtained would inundate most scanners with 5. American College of Radiology: ACR Appropriateness
Criteria Head Trauma, 2001. Available at http: //www.
unnecessary examinations, because the major- acr.org.
ity of plain films of the C-spine aren’t really 6. Hackney DB: Skull radiography in the evaluation of
needed. Hello! If they aren’t really needed acute head trauma: a survey of current practice. Radiology
then don’t get them! 181:711–714, 1991.
7. Williams JJ, Roberts L, Distell B, Simel D: Diagnosing
sinusitis by x-ray: comparing a single Waters view to
4-view paranasal sinus radiographs. J Gen Intern Med
TECHNICAL CONSIDERATIONS 7:481–485, 1992.
8. Webster EW, Merrill OE: Radiation hazards: II. Mea-
Avoiding unnecessary examinations constitutes surements of gonadal dose in radiologic examinations.
N Engl J Med 257:811– 819, 1957.
only one way to decrease unnecessary radiation 9. Antoku S, Russell WJ: Dose to the active bone marrow,
exposure in the general population. Another way gonads, and skin from roentgenography and fluoroscopy.
to significantly diminish exposure is to collimate Radiology 101:669–678, 1957.
the x-ray beam tighter. One study reported that 10. Andron GM, Crooks HE: Gonad radiation dose from
if collimation were limited just to the size of the diagnostic procedures. Br J Radiol 1957;30:295–297.
11. Mall JC, Bekerman C, Hoffer PB, et al: A unified radio-
film, radiation dose could be reduced by one logical approach to the detection of skeletal metastases.
third.18 Exposure could be further reduced by Radiology 118:323–329, 1976.
having proper filtration, fast screen-film combi- 12. Cheung Y, Rosenberg ZS: MR imaging of ligamentous
nation, and adequate gonadal shielding. Digital abnormalities of the ankle and foot. Magn Reson Imag-
ing Clin N Am 9:507–531, 2001.
radiography, which is gaining widespread use, 13. Auletta A, Conway W, Hayes C, et al: Indications for
will further help decrease radiation dose. Cer- radiography in patients with acute ankle injuries: role of
tainly having properly trained technicians and the physical examination. Am J Roentgenol 157:789–791,
properly functioning equipment will diminish 1991.
the number of retakes. These should be high- 14. Stiell I, Greenberg G, McKnight R, et al: Decision rules
for the use of radiography in acute ankle injuries. JAMA
priority goals for all radiologists to make our 269:1127–1132, 1993.
specialty more cost-effective and to provide bet- 15. Jarvik JJ, Hollingworth W, Hoagerty P, et al: The longi-
ter service to both the referring clinician and the tudinal assessment of imaging and disability of the back
patient. It should be part of every radiologist’s (LAIDBACK) study: baseline data. Spine 26:1158–1166,
2001.
responsibility to help educate and guide the un- 16. Mirvis S, Diaconis J, Chirico P, et al: Protocol-driven
knowing clinician in obtaining the appropriate radiologic evaluation of suspected cervical spine injury:
imaging examinations while eliminating those efficacy study. Radiology 170:831–834, 1989.
that are unnecessary. 17. Hoffman JR, Mower WR, Wolfson AB, et al: Validity of
a set of clinical criteria to rule out injury to the cervical
REFERENCES spine in patients with blunt trauma. N Engl J Med 343:
94–99, 2000.
1. Hall F: Overutilization of radiological examinations. 18. Morgan RH: Hearings before the Committee on Com-
Radiology 120:443–448, 1976. merce, Science and Transportation. U.S. Senate, an
2. Abrams HL: The “overutilization” of x-rays: sounding oversight of radiation health and safety. 95th Congress,
board. N Engl J Med 300:1213–1216, 1979. 1st session, June 1977, pp. 241–266.
uses mnemonics.” It serves as a nice starting characteristically. I have tried to pick out findings
point for discussing possibilities that appear as that come as close to always as I can, realizing
benign lytic lesions in bone. That mnemonic that I will often be only 95% accurate. That’s
has been in general use for many years, but I good enough for me. If it is not good enough for
have never heard a claim as to who first coined you, you will have to get your own differential
it. The first mention of it that I saw in print was criteria or discriminators. Try these and see if
in 1972 in a radiology article by Gold, Ross, and they work for you. If they don’t, modify them as
Margulis.1 By itself it is merely a long list— necessary. Whatever you do, develop reasons for
about 14 entities—and needs to be coupled with including things in your differential. Have con-
other criteria so that it can be shortened into a crete criteria of some kind for including or ex-
manageable form for each particular case. For cluding each entity.
instance, if the lesion is epiphyseal, only three I will give a brief description of each entity,
to five entities need to be mentioned, depending because complete descriptions are readily avail-
on how accurate you care to be. If multiple able in any skeletal radiology text. What I will
lesions are present, only half a dozen entities dwell on, however, are the points that are unique
need to be discussed. Ways of narrowing the dif- for each entity, thereby enabling differentiation
ferential are discussed later in this chapter. from the others. Table 2-1 is a synopsis of these
The next step after learning the names of all discriminators.
the lesions is getting some idea of what each one
looks like. This is where experience becomes a
factor. For the medical student or first-year FIBROUS DYSPLASIA
resident it is difficult to go beyond saying that
they all look lytic or bubbly and benign. How- It is unfortunate that this differential starts
ever, the fourth-year resident should have no with fibrous dysplasia because fibrous dysplasia
trouble differentiating between a unicameral can look like almost anything. It can be wild
bone cyst and a giant cell tumor (GCT) because looking, a discrete lucency, patchy, sclerotic,
he or she has seen examples of each many times expansile, multiple, and a host of other descrip-
before and knows what each looks like. The tions. It is therefore difficult to look at a bubbly
fourth-year resident may have a difficult time lytic lesion and unequivocally say it is or is not
verbalizing the differences but should be able to fibrous dysplasia. When assessing such a lesion,
tell them apart. radiology residents usually say, “I suppose it
A novice can quickly gain experience by could be fibrous dysplasia, but I’m not sure.”
looking at the examples of each of these lesions The resident is feeling insecure and becomes
in a major skeletal radiology text. In fact, I immediately defensive, setting the tone for the
highly recommend that you compare my de- entire differential diagnosis. It would be better
scription and differential points on each lesion if the differential started on a positive note, say,
with multiple examples in other books. Some of with GCT or chondroblastoma, where there
these lesions can be diagnosed radiologically are some hard and definite criteria. This would
only on a pattern identification or Aunt Minnie help the resident set a tone of self-assurance
basis. In other words, there are no hard-and-fast and decisiveness rather than appearing wishy-
criteria to differentiate some of the lesions from washy. Then, when the resident mentions fibrous
the others. dysplasia and uses the same words, “I suppose
After getting a feel for what each lesion it could be fibrous dysplasia, but I’m not sure,”
looks like radiographically and overcoming the it is looked upon as thoughtful deliberation
frustration that builds when you realize that rather than insecurity or ignorance. It is pure
many of them look alike, you should try to gamesmanship, but it makes a difference in how
learn ways to differentiate each lesion from the the resident is perceived.
others. I have developed a number of keys that How do you know whether to include or ex-
I call discriminators that will help you differen- clude fibrous dysplasia if it can look like almost
tiate each lesion. These discriminators are 90% anything? Experience is the best guideline. In
to 95% useful (I will mention when they are other words, look in a few texts and find as many
more or less accurate based on my experience) different examples as you can; get a feeling for
and are by no means meant to be absolutes or what fibrous dysplasia looks like. A few examples
dogma. They are guidelines but have a high are shown in Figures 2-1 to 2-6, but pouring
confidence rate. over another text for 10 to 15 minutes will be
Textbooks rarely tell you that a finding always time well spent.
or never occurs. Authors temper their descrip- Fibrous dysplasia will not have periostitis
tions with virtually always, invariably, usually, or associated with it; therefore, if periostitis is
Copyright © 2014 Elsevier, Inc. All rights reserved.
2 Benign Lytic Lesions 9
present, you can safely exclude fibrous dyspla- using ground-glass appearance as a buzzword for
sia. It would be possible to have a pathologic fibrous dysplasia. Fibrous dysplasia is often
fracture through an area of fibrous dysplasia, purely lytic and becomes hazy or takes on a
which then had periostitis, but I have never ground-glass look as the matrix calcifies. It can
seen this occur. Fibrous dysplasia virtually go on to calcify quite a bit, and then it presents
never undergoes malignant degeneration and as a sclerotic lesion. Also, I often see other lytic
should not be a painful lesion in the long bones lesions that have a distinct ground-glass ap-
unless there is a fracture. pearance; therefore the ground-glass quality
Fibrous dysplasia can be either monostotic can be misleading.
(most commonly) or polyostotic and has a pre- When fibrous dysplasia is in the differential
dilection for the pelvis, proximal femur, ribs, diagnosis for a lesion in the tibia, an adamanti-
and skull. When it is present in the pelvis, it noma should also be mentioned (Figure 2-7).
is invariably present in the ipsilateral proximal An adamantinoma is a malignant tumor that
femur (Figures 2-3 and 2-4). I have seen only radiographically and histologically resembles
one case in which the pelvis was involved with fibrous dysplasia. It occurs almost exclusively
fibrous dysplasia and the proximal femur was in the tibia (for unknown reasons) and is rare.
spared. The proximal femur, however, may be Because it is rare, you may choose not to in-
affected alone, without involvement in the pel- clude it in your memory bank—you won’t miss
vis (Figures 2-5 and 2-6). more than one or two in your life, even if you
The classic description of fibrous dysplasia is are a busy radiologist.
that it has a ground-glass or smoky-appearing Polyostotic fibrous dysplasia occasionally oc-
matrix. This description confuses people rather curs in association with café-au-lait spots on
than helps them, and I do not recommend the skin (dark-pigmented, frecklelike lesions)
Copyright © 2014 Elsevier, Inc. All rights reserved.
10 2 Benign Lytic Lesions
B
FIGURE 2-9 ■ A, Enchondroma. A lesion in the distal
femur is seen with irregular speckled calcification
typical of chondroid matrix. This is virtually pathog-
nomonic of an enchondroma. A chondrosarcoma
could have an identical appearance but should be
painful for consideration in the differential diagnosis.
A bone infarct can be similar in appearance to an
enchondroma. B, Bone infarcts. Bilateral lytic lesions
FIGURE 2-8 ■ Enchondroma. A benign lytic lesion in the in the femurs are noted with a densely calcified ser-
hand is an enchondroma until proved otherwise. This piginous border characteristic of bone infarcts. Com-
is a common presentation of an enchondroma. En- pare these lesions with (A), which does not have a
chondromas in any other part of the body should well-defined serpiginous border. Often the differen-
contain some calcified chondroid matrix before they tiation between bone infarcts as in this example
are included in the differential. However, calcified and an enchondroma (A) is not so clear-cut. (Case
chondroid matrix is unusual in the phalanges. courtesy of Dr. Hideyo Minagi.)
It is difficult, if not impossible, to differenti- though that is technically correct. The problem
ate an enchondroma from a chondrosarcoma. is there are few surgeons who will not biopsy a
Clinical findings (primarily pain) serve as a bet- lesion if the dictation mentions chondrosarcoma
ter indicator than radiographic findings, and as a possibility. I have seen multiple examples of
indeed pain in an apparent enchondroma should radical surgery for benign cartilaginous lesions
warrant surgical investigation. Periostitis should being undertaken because a biopsy was incor-
not be seen in an enchondroma either. Trying rectly interpreted as chondrosarcoma (including
to differentiate an enchondroma from a chon- a forequarter shoulder amputation). The correct
drosarcoma histologically is also difficult, if not dictation for an enchondroma should simply say
impossible at times. Therefore biopsy of an ap- “benign-appearing chondroid lesion with no
parent enchondroma should not be performed aggressive features noted”—no differential diag-
routinely for histologic differentiation.2 Mag- nosis should be given.
netic resonance imaging (MRI) criteria for be- Multiple enchondromas occur on occasion,
nign versus malignant includes lack of a soft tissue and this condition has been termed Ollier’s dis-
mass and no surrounding T2 high-signal edema ease (Figure 2-10, A). It is not hereditary and
in benign enchondromas. does not have an increased rate of malignant
Because an enchondroma can histologically degeneration. Older books say that Ollier’s dis-
mimic a chondrosarcoma, biopsies on painless ease has a high rate of malignant degeneration;
chondroid lesions of all types should not be rou- this is because any chondroid lesion can look
tinely done. Most tumor surgeons prefer to malignant when a biopsy is performed and
watch them with serial imaging (every 6 months needs to be correlated radiographically and
to a year) and close clinical supervision. Radiolo- clinically. The presence of multiple enchondro-
gists should refrain from adding “cannot exclude mas associated with soft tissue hemangiomas is
chondrosarcoma” to a formal report when look- known as Maffucci’s syndrome (Figure 2-10, B).
ing at a benign-appearing chondroid lesion, even This syndrome also is not hereditary; however,
A B
FIGURE 2-10 ■ A, Ollier’s disease. Multiple lytic lesions in the hand—multiple enchondromas—are seen in this pa-
tient. This is known as Ollier’s disease. B, Maffucci’s syndrome. Multiple enchondromas associated with soft tissue
hemangiomas are seen in the hand in this patient. This is Maffucci’s syndrome. Note the multiple rounded calcifi-
cations in the soft tissues, which are phleboliths in the hemangiomas.
NONOSSIFYING FIBROMA
An NOF (also called a fibroxanthoma) is proba-
bly the most common bone lesion encountered
by radiologists. It reportedly occurs in up to
20% of children and spontaneously regresses, so
it is seen only rarely after the age of 30. Fibrous
cortical defect is a common synonym, although
some people divide the two lesions on the basis
of size, with a fibrous cortical defect being
smaller than 2 cm in length and an NOF being
larger than 2 cm. Histologically, these lesions
are identical; therefore it seems illogical to sub-
divide the lesion by its size.
NOFs are benign, asymptomatic lesions that
typically occur in the metaphysis of a long bone
emanating from the cortex (Figures 2-16 to 2-18).
They classically have a thin, sclerotic border that
Discriminators:
1. Patient must be younger than age 30.
2. Periostitis and pain must be absent.
OSTEOBLASTOMA
FIGURE 2-19 ■ Healing nonossifying fibroma. A lesion in Osteoblastomas are rare lesions that could justi-
a young adult, characteristic of a nonossifying fibroma fiably be excluded from this differential without
that is beginning to disappear or “heal.” Because
these lesions are typically not seen in patients older
the fear of missing a diagnosis more than once in
than age 30, it is thought that they ossify and then your lifetime. Why, then, include them? The
blend into the normal bone. mnemonic FEGNOMASHIC would not have
Copyright © 2014 Elsevier, Inc. All rights reserved.
2 Benign Lytic Lesions 19
A
FIGURE 2-22 ■ Large nonossifying fibroma. This large,
expansile, well-defined lytic lesion was noted in an
asymptomatic 16-year-old and a biopsy was performed
because it had grown over a several-year follow-up.
It was found to be a nonossifying fibroma. There is
little justification for performing a biopsy on this
lesion, even though it is larger than most nonossifying
fibromas.
A B
C
FIGURE 2-23 ■ MRI of nonossifying fibroma. An anteroposterior plain film (A) of a knee in a 16-year-old girl shows
a benign-appearing lytic lesion (arrows), which is consistent with a nonossifying fibroma. Sagittal T1-weighted
(B) and T2-weighted (C) images show a cortically based low-signal lesion on both the T1 and the T2 sequences.
This is characteristic for a nonossifying fibroma.
cause or association with other lesions. Second- FIGURE 2-31 ■ Solitary bone cyst. A centrally placed
ary ABCs have been described to occur with lesion in the proximal femur that is classic for a soli-
GCTs, osteosarcomas, and almost any other le- tary bone cyst.
sion you can name; however, in my experience
this does not happen very often. As to occurring
after trauma, I don’t understand why they would
be age-limited if trauma were the cause. Also, solitary bone cyst can be excluded if it is not. It
malignant tumors were once thought to occur is one of the few lesions that does not occur most
after trauma because of the frequent antecedent commonly around the knees. Two thirds to three
history of trauma with malignant bone tumors. fourths of these lesions occur in the proximal
This is not seriously considered today and is humerus and proximal femur. By itself this fact
thought to be coincidental. I suspect that ABCs isn’t that helpful, or one third to one fourth of
and trauma are also coincidental, but this is mere the lesions would be missed.
speculation. Patients with solitary bone cysts are usually
ABCs typically present because of pain. They asymptomatic unless the bone fractured—a
occasionally can occur in the epiphyses, but common occurrence (Figure 2-32). Even when
there is no location in which they should be pathologic fractures occur, they rarely form
given more weight in the differential. As with periostitis. They usually occur in young pa-
osteoblastoma, they often occur in the posterior tients, and it is unusual to see one in a patient
elements of the spine. older than age 30. Although long bones are
most commonly involved, solitary bone cysts
Discriminators: have been described in almost every bone in the
1. Lesion must be expansile. body. They begin at the physeal plate in long
2. Patient must be younger than age 30. bones and grow into the shaft of the bone;
therefore they are not epiphyseal lesions. They
can, however, extend up into an epiphysis after
SOLITARY BONE CYST the plate closes, but this is unusual. A fairly
common location is in the calcaneus, where
Solitary bone cysts are also called simple bone they have a characteristic triangular appearance
cysts or unicameral bone cysts. They are not (Figure 2-33).
necessarily unicameral (one compartment), how-
ever. This is the only lesion in FEGNOMASHIC Discriminators:
that is always central in location (Figure 2-31). 1. Cyst must be central.
Many of the other lesions may be central, but a 2. Patient must be younger than age 30.
Copyright © 2014 Elsevier, Inc. All rights reserved.
24 2 Benign Lytic Lesions
HYPERPARATHYROIDISM
(BROWN TUMORS)
Brown tumors of hyperparathyroidism (HPT)
can have almost any appearance, from a purely
lytic lesion to a sclerotic process (Figure 2-34).
Generally, when the patient’s HPT is treated,
the brown tumor undergoes sclerosis and even-
tually disappears. If a brown tumor is going to be
considered in the differential, additional radio-
graphic findings of HPT should be seen. Sub-
periosteal bone resorption is pathognomonic for
HPT and should be searched for in the phalan-
ges (particularly the radial aspect of the middle
phalanges), distal clavicles (resorption), medial
aspect of the proximal tibias, and sacroiliac joints
(it has the appearance of bilateral sacroiliac joint
erosive disease). Osteoporosis or osteosclerosis
might suggest that renal osteodystrophy with
secondary HPT is present, but subperiosteal
resorption must be present; otherwise, brown multiple lytic lesions are encountered and the
tumor can be excluded from the differential. patient is asymptomatic.
Most authorities believe that brown tumors
occur most commonly in primary HPT; how-
ever, because we see so many more patients with INFECTION
secondary HPT, more brown tumors are seen in
patients with secondary rather than primary Unfortunately there is no reliable way to radio-
HPT. In fact, we rarely see brown tumors today, graphically exclude a focus of osteomyelitis (Figure
likely because of the more aggressive treatment 2-36). It has a protean radiographic appearance
of renal disease, which has resulted in fewer and can occur at any location and in a patient of
cases of hyperparathyroidism. any age. It may or may not be expansile, have a
sclerotic or nonsclerotic border, or have periostitis
Discriminator: Must be other evidence of HPT. associated with it. Soft tissue findings such as oblit-
eration of adjacent fat planes are notoriously unre-
liable and even misleading, because tumors and
HEMANGIOMAS EG can do the same thing.
When osteomyelitis occurs near a joint, if
Multiple hemangiomas, also known as cystic the articular surface is abutted, invariably the
angiomatosis or cystic lymphangiomatosis, al- adjacent joint will be involved and show either
though uncommon, are seen more frequently cartilage loss, an effusion, or both (Figure 2-37).
than brown tumors and should probably replace This finding is not particularly helpful, because
them as the “H” in FEGNOMASHIC. Cystic any lesion can cause an effusion, but it’s the best
angiomatosis usually is an incidental finding of I can come up with.
multiple lytic lesions throughout the skeleton
(Figure 2-35). Tissue abnormalities may also be
present, which can be painful. Although the
hemangiomas are incidental, some believe they
are in a similar category to Gorham’s disease
(massive osteolysis or disappearing bone dis-
ease), but without the destructive potential. Cys-
tic angiomatosis should be considered when
B
FIGURE 2-37 ■ Osteomyelitis with pyoarthritis. A plain film of the shoulder (A) in this child with shoulder pain
shows a well-defined lytic lesion. Coronal T2-weighted MRI scan (B) shows high signal throughout the lesion,
which extends through the cortex (arrow) into the joint. This is a typical appearance for osteomyelitis with exten-
sion into the joint.
When a sclerotic margin is present, infection lymphoma, and fibrosarcoma. Most cases of
usually causes the sclerotic margin to be thick osteomyelitis, however, do not have sequestra,
and ill defined or fuzzy on its outermost portion, so this feature is also of limited use. Therefore
but these findings are by no means always pres- infection will be in almost every differential
ent and are of only limited usefulness. diagnosis of a lytic lesion, which is all right,
If a bony sequestrum is present, osteomyeli- because it is one of the most common lesions
tis should be strongly considered. As previ- encountered.
ously mentioned, the only lesions described
that demonstrate sequestra are infection, EG, Discriminators: None.
Copyright © 2014 Elsevier, Inc. All rights reserved.
2 Benign Lytic Lesions 27
CHONDROBLASTOMA
Chondroblastomas are among the easiest lesions
to deal with because they occur only in the
epiphyses (a handful of cases have been reported
in the metaphyses—they’re rare), and they occur
almost exclusively in patients younger than
age 30 (Figure 2-38). What could be easier?
Anywhere from 40% to 60% demonstrate calci-
fication, so absence of calcification is not helpful.
Presence of calcification is helpful as long as you
can be sure that it is not detritus or sequestra
from infection or EG (both of which can occur
in the epiphyses). Because I can never be certain
about the calcification, I don’t worry about it.
The differential diagnosis of a lytic lesion in FIGURE 2-39 ■ Chondroblastoma. A lytic lesion in the
the epiphysis of a patient younger than age 30 is calcaneus that abuts the superior portion of the
short and simple: (1) infection (most common), calcaneus has a fairly short differential diagnosis.
(2) chondroblastoma, and (3) GCT (it has its own The differential is essentially the same as for an
diagnostic criteria, so it can usually be definitely epiphysis. On biopsy this lesion was found to be a
chondroblastoma.
ruled out or in). This is an old, classic differential
and probably takes care of 98% of epiphyseal le-
sions. If you want to be 99% certain, you must
add two more lesions—ABC (be certain it is ex- that there are three entities in the epiphyseal
pansile) and EG—and add mets and myeloma if differential and then pick from the longer list of
the patient is older than age 40. I no longer rec- five entities, often leaving out the main things
ommend using the longer differential because it while including EG and ABC. Most of residents
seems to cause confusion. Residents remember have too much to commit to memory already
without trying to stuff in a few rare presentations
that will not be seen very often.
A caveat on epiphyseal lesions is to always
consider the possibility of a subchondral cyst or
geode, which has been described in anything that
can result in degenerative joint disease (DJD). Be
certain no joint pathology that might cause DJD
is present, or an unnecessary biopsy of a geode
might be performed based on your differential of
an epiphyseal lesion (see Chapter 4).
The carpal bones, the tarsal bones, and the
patella have a tendency to behave like epiphyses
as far as their differential diagnosis of lesions is
concerned. Therefore a lytic lesion in these
areas has a similar differential as an epiphyseal
lesion (Figure 2-39); although in my experience
this differential diagnosis in the epiphyses is
close to 99% inclusive, in these other locations
it is only about 50% inclusive—still, a good
differential to begin with.
Discriminators:
1. Patient must be younger than age 30.
2. Lesion must be epiphyseal.
BOX 2-3 “Automatics” age 30. Make sure you give a thoughtful pause
after inspecting the film—as if you really con-
Younger Than Age 30: sidered the pros and cons of mentioning them.
Eosinophilic granuloma This is gamesmanship, but it adds to your cred-
Infection ibility. In patients older than age 40, mets and
Older Than Age 40: infection are automatics. No consideration for
Mets the lesion’s appearance or location needs to be
Infection made when using the automatics—simply find a
lesion.
DIFFERENTIAL DIAGNOSIS OF A
struts or compartments in the lesion, and size of SCLEROTIC LESION
the lesion.
There are a few lesions that I call automatics Many lytic lesions spontaneously regress and are
(Box 2-3). That is, they need to be included not usually seen in patients older than age 30.
automatically in almost every case. In patients When these lesions regress, they often fill in
younger than 30, EG and infection must be with new bone and have a sclerotic or blastic
included in every differential of a lytic bone le- appearance. Therefore, when a sclerotic focus
sion. In fact, they should be mentioned in all is identified in a 20- to 40-year-old patient, es-
cases except for trauma or arthritis because pecially if it is an asymptomatic, incidental
they can have virtually any appearance—lytic, finding, the following lesions should be consid-
sclerotic, mixed, benign, aggressive, and so on. ered: NOF (Figure 2-41), EG, solitary bone
Hence they can mimic almost any bone lesion. cyst, ABC, and chondroblastoma. Several other
I recommend mentioning EG and infection lesions that can also appear sclerotic should be
for every bone lesion in patients younger than included: fibrous dysplasia, osteoid osteoma,
A B
FIGURE 2-41 ■ A, Healing nonossifying fibroma. This faintly sclerotic lesion in the proximal tibia of a 30-year-old
asymptomatic patient is characteristic for a healing or disappearing nonossifying fibroma. Prior films showed a
typical lytic nonossifying fibroma. B, Healing nonossifying fibroma. This densely sclerotic lesion in the posterior
proximal tibia in a young asymptomatic patient was thought to represent an osteoid osteoma or osteomyelitis.
Even though the patient was asymptomatic, a biopsy was performed. It revealed a nonossifying fibroma that had
ossified. In a patient older than age 40, metastatic disease would need to be considered.
FIGURE 2-42 ■ Giant bone island. A large sclerotic lesion is seen in the right iliac bone, which could easily be mis-
taken for a focus of metastatic disease in a patient older than age 40. This is a giant bone island and has the
typical pattern of irregular “feathered” margins with the trabecula blending into the normal bone. The axis of the
lesion is in the long axis of the bone or in the direction of the primary weight-bearing trabecula.
infection, brown tumor (healing), and perhaps from that misdiagnosis—it is still benign and in
a giant bone island (Figure 2-42). As these le- most cases an incidental finding.
sions continue to resolve, they often will be In any patient older than age 40, the num-
seen to have varying amounts of normal fatty ber 1 possibility for a sclerotic lesion should
marrow (Figure 2-43). They should not be mis- be metastatic disease (now that I’ve passed
takenly called intraosseous lipomas just because 40, I’m seriously considering moving the age
they contain fat, although no harm would come to 50).
A B
C
FIGURE 2-43 ■ Resolving nonossifying fibroma. An anteroposterior plain film of a knee (A) in a 28-year-old man
shows a sclerotic lesion in the metaphysis of the tibia. Coronal T1-weighted (B) and T2-weighted (C) images of
the knee reveal a lesion that on the T1-weighted image has mixed high and low signal, consistent with fatty tissue
in some areas, which is mixed high and low signal on the T2-weighted image. This is typical for a resolving non-
ossifying fibroma.
Malignant Tumors
Radiology residents have difficulty dealing with think of a malignant lesion when using the
malignant bone tumors, and the difficulty gets Gestalt approach, but the lesion must also have
worse in the years after residency. This is simply other criteria for a malignant process, such as a
because malignant bone tumors, thankfully, are wide zone of transition.
not very common. Nevertheless, every radiolo-
gist will encounter one or two a year in most
practices and should be able to recognize them
Periostitis
and provide a good differential diagnosis. Periosteal reaction occurs in a nonspecific man-
First, how do you recognize a malignant ner whenever the periosteum is irritated, whether
tumor and differentiate it from a benign process? it is irritated by a malignant tumor, a benign
This can be difficult and often even impossible. tumor, infection, or trauma. Callus formation in
Recognizing that it is aggressive is usually easy, a fracture is actually just periosteal reaction of
but saying that it is malignant is another matter the most benign type. Periosteal reaction occurs
altogether. Processes such as infection and eosin- in two types: benign (Figure 3-3, A and B) and
ophilic granuloma can mimic malignant tumors aggressive (Figure 3–3, C and D). The difference
and are, of course, benign. They will often be between the two is based more on the timing
included in the differential diagnosis of an aggres- of the growth of the irritation than on whether
sive lesion along with malignant tumors. the process causing the periostitis is malignant
or benign. For example, a slow-growing benign
DIFFERENTIATION OF MALIGNANT
FROM BENIGN
What radiologic criteria are useful for deter-
mining malignant versus benign? Standard text-
books and the literature give four aspects of a
lesion to be examined: (1) cortical destruction,
(2) periostitis, (3) orientation or axis of the
lesion, and (4) zone of transition.
Let me expound on each of these criteria
and show why only the last one—the zone of
transition—is the most reliable, with an accu-
racy rate of 90%.
Cortical Destruction
Often cortical bone is replaced by part of the
noncalcified matrix (fibrous matrix or chon-
droid matrix) of benign fibro-osseous lesions
and cartilaginous lesions. This can give the false
impression of cortical destruction on plain films
(Figure 3-1) or computed tomography (CT).
Also, benign processes such as infection and FIGURE 3-1 ■ Apparent cortical destruction. This be-
nign chondroblastoma has noncalcified chondroid tis-
eosinophilic granuloma can cause extensive cor- sue replacing cortical bone in the proximal femur,
tical destruction and mimic a malignant tumor. which gives the lesion a destructive appearance. This
Aneurysmal bone cysts are known to cause such is an example of cortical replacement rather than of
thinning of the cortex as to make it radiograph- cortical destruction, which can be very confusing if one
uses cortical destruction as an aggressive or malignant
ically undetectable (Figure 3-2). For these rea- key in differential diagnosis. Note that the zone of tran-
sons I find cortical destruction to occasionally sition is narrow, as one would expect in a benign
be misleading. Cortical destruction makes one lesion such as this.
tumor will cause thick, wavy, uniform, or dense periostitis will solidify and appear benign. There-
periostitis because it is a low-grade chronic irrita- fore, when periostitis is seen, the radiologist
tion that gives the periosteum time to lay down should try to characterize it into either a benign
thick new bone and remodel into more normal (thick, dense, wavy) type or an aggressive (lamel-
cortex. A malignant tumor causes a periosteal lated, amorphous, sunburst) type.
reaction that is high grade and more acute; Unfortunately, judging the lesion by its peri-
hence, the periosteum does not have time to ostitis can be misleading. First, it takes consid-
consolidate. It appears lamellated (onion skinned) erable experience to accurately characterize
or amorphous, or it may even appear sunburst. If periostitis because many times the reaction is
the irritation stops or diminishes, the aggressive not clearly benign or aggressive. Second, many
A B
C
FIGURE 3-2 ■ A, Aneurysmal bone cyst. This benign lesion has thinned the cortex to such a degree as to make it
imperceptible on a plain radiograph. As in Figure 3-1, this can be misconstrued as cortical destruction, giving the
false impression of a malignant or very aggressive lesion. B, Giant cell tumor. This lesion in the proximal femur
has expanded and thinned the cortex to such a degree that it is imperceptible on the plain film. C, On a CT scan
through this region a very thin cortex that was greatly expanded was identified. This giant cell tumor originated
from the greater trochanter, which is an epiphyseal equivalent. Note that it has a well-defined but nonsclerotic
zone of transition, as all giant cell tumors in long bones do.
A B
C D
FIGURE 3-3 ■ Benign periostitis. A, An osteoid osteoma in the midshaft of the tibia has caused thick, wavy, dense
periostitis, which is classic for benign type of periostitis. Malignant lesions are incapable of forming this type of
periostitis and should not be considered in the differential. This type of periostitis is basically indistinguishable
from callus formation in a fracture. B, Thick, wavy periostitis (arrows) along the ilium in a child with a permeative
lesion in the pelvis is characteristic for infection or eosinophilic granuloma. Ewing’s sarcoma was initially consid-
ered in the differential; however, the benign periostitis would make a malignant lesion very unlikely. Biopsy
showed this lesion to be eosinophilic granuloma. C, Aggressive periostitis. Amorphous, sunburst periostitis with
a Codman’s triangle (arrows) in a patient with a mixed lytic-sclerotic lesion of the humerus, which on biopsy was
shown to be Ewing’s sarcoma. Although this type of periostitis is characteristic for a malignant lesion, it could
also be seen with benign processes such as eosinophilic granuloma or infection. D, Lamellated, or onion-skinned,
periostitis is characteristic of an aggressive process, such as in this patient with Ewing’s sarcoma of the femur.
Again, this aggressive type of periostitis could conceivably occur in a benign process such as infection or eo-
sinophilic granuloma.
Copyright © 2014 Elsevier, Inc. All rights reserved.
3 Malignant Tumors 35
Zone of Transition
Without question the zone of transition is the
most reliable indicator in determining benign
versus malignant lesions. Unfortunately it also
has some drawbacks, on which I will elaborate.
The zone of transition is the border between
the lesion and the normal bone. It is said to be
narrow if it is so well defined that it can be
drawn with a fine-point pen (Figure 3-4). If it
is imperceptible and cannot be clearly drawn, it
is said to be wide (Figure 3-5). Obviously all
shades of gray lie in between, but most lesions
can be characterized as having either a narrow or
a wide zone of transition. If the lesion has a scle-
rotic border, it, by definition, has a narrow zone
of transition.
If a lesion has a narrow zone of transition, it
is a benign process. The exceptions to that are
rare, and I’m willing to miss them. If a lesion has
a wide zone of transition, it is aggressive. Notice
that I said aggressive and not malignant. As with
aggressive periostitis, many benign lesions can
have a wide zone of transition. A few of the same
processes that can cause aggressive periostitis, FIGURE 3-5 ■ Wide zone of transition. A lytic perme-
and thereby mimic a malignant tumor, can have ative process is seen in the midshaft of the femur in
a wide zone of transition (i.e., infection and eo- this patient. On biopsy it was found to be a malignant
fibrous histiocytoma. The zone of transition in this
sinophilic granuloma). They are aggressive in their lesion is said to be wide, because it cannot be easily
radiographic appearance because they are fast- drawn with a fine-point pen. A permeative lesion such
acting, aggressive lesions. The zone of transition as this, by definition, has a wide zone of transition.
B
FIGURE 3-7 ■ Zone of transition on MRI. A, A T2-weighted
MRI scan of the knee in this child with knee pain
shows a well-defined lesion (arrow) with a low signal
border, suggesting a narrow zone of transition with a
sclerotic border. Because the knee pain was de-
scribed as probably related to the menisci, this lesion
was believed to be an incidental, benign process,
such as a nonossifying fibroma. However, the plain
film (B), obtained later, shows a barely discernible
lytic lesion with a wide zone of transition (arrow).
This lesion was, in fact, painful, and the differential
diagnosis, based solely on the plain film, would in-
clude osteosarcoma—the eventual histologic diagno-
sis. The zone of transition can be used only on plain
FIGURE 3-6 ■ Permeative pattern. A permeative pattern films; it is invalid on MRI scans.
is defined as multiple, small, irregular holes in bone
and indicates an aggressive process. Ewing’s sarcoma
typically has a permeative pattern; however, infection
and eosinophilic granuloma, as in this example, can
also have a permeative pattern. This is a fine-detailed
film of the same case shown in Figure 3–3, B.
B C
FIGURE 3-11 ■ Parosteal osteogenic sarcoma. A, A lateral film of the distal femur shows a large calcific mass with
an ill-defined, fluffy, calcific periphery growing off the posterior femur. The location and appearance are character-
istic for a parosteal osteogenic sarcoma. B, A CT scan through this lesion shows cortical and medullary involve-
ment, which indicates a more sinister lesion, with treatment and prognosis similar to that of a central osteogenic
sarcoma. Without intramedullary involvement, a parosteal sarcoma has a favorable prognosis. C, A proton-density
MRI scan in another patient with a parosteal osteosarcoma shows how the vessels (arrow) can be easily identified
in relation to the tumor. In this example the vessels are displaced posteriorly by the tumor.
It often wraps around the diaphysis without appear somewhat aggressive. Unfortunately it
breaking through the cortex. It occurs in an can appear malignant histologically, so biopsy
older age-group than the central osteosarcomas can lead to disastrous consequences. Amputations
and is not as aggressive. Treatment used to con- for benign cortical desmoids being mistaken for
sist of merely shaving the tumor off the bone it malignancies have occurred. (See Chapter 4 for
originated from; however, recurrence rates were further points on cortical desmoids.) Another
so high that now wide-bloc excisions are per- lesion that can be mistaken for a parosteal osteo-
formed. Once a parosteal osteosarcoma violates sarcoma is an area of myositis ossificans (Figure
the cortex of the adjacent bone, some consider 3-13). Like cortical desmoids, areas of myositis
it to be as aggressive as a central osteosarcoma ossificans can be histologically mistaken for
and treat it in a similar fashion (i.e., amputation malignancies with disastrous consequences. There-
or radical excision). Therefore the radiologist fore differentiation is vital. Fortunately differ-
needs to evaluate the lesion for invasion of the entiation between parosteal osteosarcoma and
adjacent cortex to help determine treatment and myositis ossificans is fairly easily done radio-
prognosis. This is best done with CT or MRI graphically. (See Chapter 4 for differential points
examination (Figure 3-11, B and C). between parosteal osteosarcoma and myositis
A common location for parosteal osteosarco- ossificans.)
mas to arise is from the posterior femur, near the Another type of osteosarcoma that residents
knee. A lesion that can mimic an early parosteal often mention in their differential of a parosteal
osteosarcoma in this location is a so-called corti- osteosarcoma is a periosteal osteosarcoma. It
cal desmoid (Figure 3-12). A cortical desmoid is should not get such attention. First, it is ex-
an avulsion injury that is totally benign but can tremely rare, with fewer than 50 such lesions
FIGURE 3-12 ■ Cortical desmoid. Irregular periostitis FIGURE 3-13 ■ Myositis ossificans. This lesion is often
(arrows) off the medial supracondylar ridge of the confused with parosteal osteogenic sarcoma, because
distal femur is pathognomonic for a small avulsion of at first glance it has a close resemblance. However,
the adductor muscles and is called a cortical desmoid. myositis ossificans, as in this example, will demon-
Biopsy of this lesion can easily result in a mistaken strate calcification that is densest in the periphery and
diagnosis of a sarcomatous lesion and should there- well defined, whereas parosteal osteogenic sarcoma
fore be avoided. A cortical desmoid should not be has calcification that is most concentrated centrally
confused with an early parosteal osteogenic sarcoma and ill defined in the periphery. Myositis ossificans is
on radiographs, because it will then lead to an unnec- another lesion that can be confused with a malignant
essary biopsy, perhaps with dire consequences. An lesion on biopsy and therefore should be radiographi-
incidental finding is a nonossifying fibroma seen just cally rather than surgically diagnosed.
proximal to the cortical desmoid.
Ewing’s Sarcoma
The classic Ewing’s sarcoma is a permeative
(multiple small holes) lesion in the diaphysis of FIGURE 3-15 ■ Ewing’s sarcoma. A mixed lytic-sclerotic
lesion in the femur of a child with periostitis that is
a long bone in a child (see Figure 3-3, D). amorphous and sunburst, such as in this example, is
However, only about 40% of these tumors oc- characteristic of a Ewing’s sarcoma. An osteosarcoma
cur in the diaphysis, with the remainder being could have a similar appearance.
Desmoid
A desmoid tumor (not to be confused with a cor-
tical desmoid; see Chapter 4) is a half-grade fibro-
sarcoma. It has also been called a desmoplastic
fibroma or aggressive fibromatosis. These lesions,
FIGURE 3-19 ■ Chondrosarcoma. An amorphous, irregu-
like fibrosarcoma, are lytic but are usually fairly
lar calcification is seen in a lesion arising in the ischium well defined because of their slow growth. They
(arrow). This is fairly typical for a chondrosarcoma. often have benign periostitis present that has
Copyright © 2014 Elsevier, Inc. All rights reserved.
44 3 Malignant Tumors
FIGURE 3-22 ■ MFH. A large, fairly well-defined destructive process of the entire right iliac wing is noted. On
biopsy this was shown to be an MFH. An MFH can be very slow growing and will occasionally have a narrow
zone of transition, such as this.
A B
FIGURE 3-23 ■ Desmoid. A large soft tissue mass (arrows) is seen in the right buttock, which is low signal on both
the T1-weighted (A) and the T2-weighted (B) images. This is typical for a soft tissue desmoid tumor.
A B
FIGURE 3-25 ■ Desmoid tumor. A and B, A multilocular, heavily septated, destructive lytic lesion of the distal femur
is noted, which is fairly characteristic for a desmoid tumor. The thick septa and narrow zone of transition
are characteristic of a benign process, whereas the Codman’s triangle and the large amount of bony destruction
indicate an aggressive process.
Myeloma
Like metastases, myeloma should be considered
only in patients older than age 40, although some
radiologists use age 35 for the lower limits of
myeloma. Myeloma typically has a diffuse per-
meative appearance that can mimic a Ewing’s
sarcoma or a primary lymphoma of bone (Figure
3-29). It often involves the calvarium (Figure
3-30). Because of the age criteria, Ewing’s sar-
coma and myeloma are not in the same differen-
tial, however. Rarely, myeloma can present with
multiple sclerotic foci that resemble diffuse mets.
Myeloma is one of the only lesions that is not
characteristically “hot” on a radionuclide bone
FIGURE 3-26 ■ Primary lymphoma of bone. A diffuse
permeative pattern throughout the humerus in this
scan; therefore radiologic “bone surveys” are
35-year-old patient is characteristic of primary lym- performed instead of radionuclide bone scans
phoma of bone. when evidence of myeloma is found clinically.
Copyright © 2014 Elsevier, Inc. All rights reserved.
3 Malignant Tumors 47
FIGURE 3-27 ■ Metastatic prostate carcinoma. Diffuse blastic metastases are seen throughout the pelvis and
proximal femurs, with a lytic destructive lesion seen in the right proximal femur. Prostate metastases tend to be
blastic, but as shown here, they can occasionally be lytic.
FIGURE 3-30 ■ Multiple myeloma. A lateral view of the skull shows multiple lytic lesions in the calvarium, which is
a characteristic appearance of multiple myeloma.
A B
FIGURE 3-33 ■ MRI appearance of plasmacytoma. Axial proton-density (A) and T2-weighted (B) images through a
vertebral body with a plasmacytoma show a characteristic, if not pathognomonic, appearance of a mini-brain,
which is seen only with plasmacytoma.
characteristic Aunt Minnie on CT but is not odds and mention the two most common soft
appreciated on plain films—both the acute and tissue tumors, MFH and liposarcoma, as the best
chronic myeloma simply appear osteopenic on candidates, but any cell type can produce a be-
plain films. nign or malignant tumor and mimic any other
Occasionally myeloma will present with a soft tissue tumor. A lipoma can be separated out
lytic bone lesion called a plasmacytoma. This by the appearance of fat, but a liposarcoma may
lesion can mimic any lytic bone lesion, benign or or may not have fat present. Therefore we are
aggressive, in its appearance; it can precede left to give descriptions of size and extent, and
other evidence of myeloma by up to 3 years. It let the pathologist tell us the rest. Most of us are
can be solitary or multiple, although some insist uncomfortable with this approach because our
that to be considered a plasmacytoma, it must be training has been to derive an answer—or at
solitary; when it is multifocal, they argue it is least a listing of probable lesions. This is simply
myeloma. Whatever. It has an appearance on not possible for soft tissue tumors.
MRI scans in the vertebral body that is virtually A few words about soft tissue tumors that may
pathognomonic. It resembles the cadaveric sec- be helpful: as mentioned earlier, liposarcomas do
tion of a brain with which every medical student not have to have fat visible in the tumor. There
is familiar, hence the name mini-brain appear- are at least three subtypes of liposarcomas, two
ance (Figure 3-33). of which have only small amounts of fat present.
Synovial sarcomas, or synoviomas, only rarely, if
ever, originate in a joint. They are often adjacent
Soft Tissue Tumors to joints. There are no malignant tumors that
Most radiology residents feel uneasy when faced routinely need to be considered in the differen-
with the differential diagnosis of a soft tissue tial diagnosis of joint lesions. Synovial osteo-
tumor. They will give elaborate description with chondromatosis is a benign joint lesion that
plenty of pertinent and not-so-pertinent nega- occurs from metaplasia of the synovium and leads
tives, such as “no calcifications are seen,” “no to multiple calcific loose bodies in a joint. This
bony destruction is noted,” and “no obliteration can histologically mimic a chondrosarcoma;
of fat planes is apparent.” Then, when faced with therefore it is best diagnosed radiographically,
finally giving a differential, few can give an au- because it has a pathognomonic radiographic
thoritative list of the possibilities. The reason for appearance (Figure 3-34). Up to 30% of the
this is simple: there is no authoritative, useful time, however, the loose bodies do not calcify,
differential for soft tissue tumors, whether or and the lesion then can mimic pigmented villo-
not there is calcification, bony destruction, fat nodular synovitis (PVNS) on plain films. On
plane involvement, or whatever. You can play the MRI scans, synovial osteochondromatosis can
Copyright © 2014 Elsevier, Inc. All rights reserved.
50 3 Malignant Tumors
A B
FIGURE 3-35 ■ Tumefactive synovial osteochondromatosis. A, A plain film of the shoulder shows a partially calci-
fied mass, which is eroding the medial aspect of the humerus. Coronal proton-density (B) and T2-weighted
(C) images of the shoulder reveal a large mass encircling the humeral head, which was interpreted as a sarcoma.
A biopsy was performed, and the mass was labeled as a chondrosarcoma, which resulted in a forequarter ampu-
tation. The intraarticular nature of the mass was not appreciated until after the radical surgery, when it was
correctly recognized as synovial chondromatosis.
C
FIGURE 3-35, cont’d. For legend see opposite page.
FIGURE 3-36 ■ Pigmented villonodular synovitis (PVNS). FIGURE 3-37 ■ Hemangioma. Multiple irregular lytic
Large erosions in the femoral head and acetabulum lesions, predominantly cortical in nature, are seen in
are characteristic for pigmented villonodular synovitis; the tibia in this patient with a soft tissue mass. Cortical
however, nonossified synovial osteochondromatosis holes such as this occur almost exclusively in radiation
could present like this. and soft tissue hemangioma. Note the phleboliths in
the posterior soft tissues (arrows), which are often
seen in hemangioma and make this an easy diagnosis.
adjacent joint with an identical signal (Figure A hematoma can present as a focal mass and
3-38). Differentiation of a solid mass from a gan- be misdiagnosed on MRI as a tumor. If high
glion is one of the few uses of Gd-DTPA in mus- signal is present on a T1-weighted image, blood
culoskeletal MRI. A solid tumor will diffusely products should be considered. However, keep
enhance with gadolinium, whereas a ganglion will in mind that tumors often have bleeding within,
have rim enhancement only. When a mass that so that alone won’t indicate a hematoma. In
looks like a fluid collection or cyst is seen in a addition, blood may or may not have high T1
location atypical for a cyst or ganglion, intrave- signal. Gadolinium can be misleading in hema-
nous contrast should be given. Two tumors that tomas because some enhancement can occur in
typically mimic fluid on T2-weighted MRI scans long-standing hematomas, making one think it
are synovial sarcoma and neural tumors (neurofi- is a solid tumor. Heightened suspicion for a he-
bromas and schwannomas) (Figure 3-39). matoma should be present if a history of trauma
B C
FIGURE 3-38 ■ Atypical synovial cyst. A, CT scan through the distal femurs in a patient with a soft tissue mass
around the right knee shows a multilocular soft tissue mass adjacent to the distal right femur (arrows). B, An MRI
examination with T1 weighting through the same area as imaged in A shows intermediate-intensity signal in a
multilocular soft tissue mass. C, An MRI examination with T2 weighting shows high-intensity signal in the lesion,
which is characteristic for fluid. This was an atypical synovial cyst arising from the knee joint.
A B
FIGURE 3-39 ■ Schwannoma. Coronal T1-weighted (A) and T2-weighted (B) images of an ankle in a young
woman with a painful mass show a homogeneous mass that seems to emanate from the joint itself, suggesting
an atypical ganglion. Intravenous contrast should have been administered to determine whether this was a
solid or a cystic mass. At surgery this was found to be a solid tumor, a benign schwannoma. It did not extend
into the joint.
A B
FIGURE 3-40 ■ Hematoma. T1-weighted (A) and T2-weighted (B) sagittal images through the shoulder of an athlete
with trauma to his pectoralis major shows a mass with surrounding high T1 signal that is pathognomonic for a
hematoma. On the T2-weighted image (B), it has the appearance of a sarcoma.
A B
C
FIGURE 3-41 ■ Hemorrhagic cyst. A 61-year-old man sought treatment for a mass on his thigh, which was revealed
to be of intermediate signal on an axial T1-weighted image (A) (arrow). A coronal T2-weighted image (B) shows
marked high signal, which is homogeneous. Is this solid or cystic? A coronal T1-weighted image with fat suppres-
sion was performed after administration of gadolinium (C), which shows apparent diffuse enhancement, indicat-
ing this is a solid mass. However, the increased signal cannot reliably be attributed to enhancement from the
gadolinium because fat suppression was applied, which, in some instances, can make a mass appear brighter
simply because of its relatively higher signal compared with that of the muscles. Notice on the T1-weighted image
in A that the mass is higher in signal than the muscles; therefore, when fat suppression is applied, the next high-
est signal structure is the mass. Scaling the image then makes it appear bright, which can be misinterpreted as
secondary to contrast enhancement. This was incorrectly called a solid tumor, and at surgery it was found to be
a hemorrhagic cyst.
compared, it is important to not change any- pictures, but at the expense of an occasional mis-
thing else; otherwise, you won’t know for sure diagnosis. If you must fat-suppress the postcon-
what’s causing the increased signal in the mass. trast images, you must also fat-suppress the
There’s no good reason to fat-suppress the post- precontrast images so that you have changed
contrast images—it simply makes for prettier only one variable.
A B
FIGURE 4-2 ■ Myositis ossificans. A, Hazy calcification is seen adjacent to the humeral shaft, with underlying peri-
osteal reaction noted. It is difficult to ascertain whether the calcification is circumferential. B, A CT scan through
this mass shows that the calcification is unequivocally circumferential, making the diagnosis of myositis ossifi-
cans a certainty.
A B C
FIGURE 4-4 ■ Myositis ossificans. A, A plain film of the humerus in this 30-year-old man shows a calcific mass
adjacent to the diaphysis of the humerus. The calcification is not clearly peripheral in nature, although the central
portion is less well mineralized. B, An axial T2-weighted image through the mass shows only a high-signal mass
without evidence of calcification. C, A CT scan through the mass demonstrates the typical peripheral calcification,
which is virtually pathognomonic for myositis ossificans.
people. Biopsy should be avoided in all cases. findings, such as joint space narrowing, sclerosis,
They are often seen as incidental findings and osteophytes, a diagnosis should be made
on MRI and have a characteristic appearance radiographically (Figures 4-10 and 4-11). How-
(Figure 4-9). ever, on occasion the additional findings are
Trauma can lead to large cystic geodes or sub- subtle and can be missed (Figure 4-12). Geodes
chondral cysts near joints that can be mistaken for can also occur in the setting of calcium pyrophos-
other lytic lesions, and thus a biopsy is performed. phate dihydrate crystal disease (also known as
Although the biopsy specimen is not likely to CPPD or pseudogout), rheumatoid arthritis, and
mimic a malignant process, it is nevertheless avascular necrosis.2
avoidable. Because geodes from degenerative dis- An entity that is often confused with meta-
ease almost always are associated with additional static disease to the spine is discogenic vertebral
Copyright © 2014 Elsevier, Inc. All rights reserved.
58 4 “Don’t Touch” Lesions
A B
C
FIGURE 4-9 ■ Cortical desmoid. A, An anteroposterior film of the knee in a child shows a faint lytic lesion (arrows)
in the medial aspect of the distal femur. Axial T1-weighted (B) and T2-weighted (C) images through the lesion
show a cortically based process (arrows) in the medial supracondylar ridge, which is characteristic for a cortical
desmoid.
of the humerus (Figures 4-15 and 4-16). This after with additional views. A transscapular or an
results from a fracture with hemarthrosis, which axillary view is the key to making the diagnosis of
causes distention of the joint and migration of a pseudodislocation. With either of these views
the humeral head inferiorly. An axial or trans- the humeral head can be seen to be normally
scapular view shows that it is not anteriorly or positioned in relation to the glenoid, although it
posteriorly dislocated (the usual forms of shoul- may appear somewhat inferiorly displaced. If
der dislocation) but merely inferiorly displaced. necessary, the joint can be aspirated to confirm
On an anteroposterior view it can mimic a pos- the presence of a bloody effusion and to show the
terior dislocation. Often attempts are made to normal position of the humeral head with no
“relocate” the humeral head, efforts that are fluid in the joint.5
both fruitless (because it is not dislocated) and Costochondritis, or Tietze’s syndrome, can
painful. A fracture is invariably present and if cause a bulbous swelling of a rib (Figure 4-17) as
not seen on the initial films should be sought a result of periostitis, which can mimic a rib
A B
FIGURE 4-10 ■ Geode. A, A plain film of the hip in this older patient with hip pain shows in the supraacetabular re-
gion a lytic lesion (arrows), which has a benign appearance. Mild osteoarthritis was thought to be present (when
compared with the opposite hip, joint space narrowing and minimal sclerosis were seen); hence this was believed
to be a subchondral cyst or geode. B, Several years later the same hip shows a large lytic lesion (arrows) that still
appears benign. The osteoarthritis has increased in severity. However, because of the growth of the lesion, a
biopsy was performed and it was found to be a geode. A biopsy should have been avoided.
A A
B
B
FIGURE 4-12 ■ Geode. A, A cystic lesion was noted in
the femoral head (arrows) of a young male with a pain-
ful hip. B, A CT scan through this area shows the sub-
articular nature and adjacent sclerosis. The differential
diagnosis of infection, eosinophilic granuloma, and
chondroblastoma was given. A ring of osteophytes
(open arrow heads) was noted in retrospect on the
plain film (A) in the subcapital region, which indicates
degenerative disease of the hip. This is an extremely
unusual presentation in a healthy 20-year-old male;
however, it makes the lytic lesion in the femoral head
almost certainly a subchondral cyst or geode. This was
an active soccer player who had been playing with
pain in his hip for several years after an injury that had
caused the degenerative disease. Unfortunately a biopsy
was performed anyway, and a subchondral cyst or
geode was confirmed. C
FIGURE 4-13 ■ Discogenic vertebral sclerosis. A–C,
These films all show patients with sclerosis on the
inferior portion of the L4 vertebral body associated
with minimal osteophytosis and joint space narrow-
ing at the adjacent disc space. This is the classic
appearance for discogenic vertebral sclerosis, and a
biopsy to rule out metastatic disease should not be
performed.
A
A
B
FIGURE 4-14 ■ Fracture mimicking osteosarcoma. A, This
16-year-old had experienced pain around the knee for
2 weeks before these radiographs were taken. The knee
films showed diffuse sclerosis and extensive periostitis B
about the distal femur, which was believed to be char-
acteristic for an osteogenic sarcoma. The periosteal FIGURE 4-15 ■ Pseudodislocation of the shoulder. A, This
reaction, however, was believed to be much too thick, patient experienced trauma to the shoulder, with resul-
dense, and wavy to represent malignant type of peri- tant pain and immobility, and was thought to have a
ostitis. B, A small offset of the epiphysis can be seen dislocation of the shoulder after the anteroposterior film
(arrow), which indicates an epiphyseal slippage con- was seen. The humeral head is inferiorly placed in rela-
sistent with a Salter epiphyseal fracture. The patient tion to the glenoid; however, this is not the characteris-
had fallen off his bicycle and fractured his femur, yet tic location of an anterior or posterior dislocation. B, The
he continued to be active. The lack of immobility transscapular view shows the humeral head to be situ-
caused exuberant periostitis or callus with a large ated normally over the glenoid without anterior or pos-
amount of reactive sclerosis, all of which mimicked an terior dislocation. These findings are characteristic for a
osteogenic sarcoma. pseudodislocation caused by hemarthrosis, or blood in
the joint, which allows the shoulder to be subluxed
rather than dislocated. Aspiration of the blood will result
in the humeral head returning to its normal position
in relation to the glenoid; however, this is not usually
necessary. When a pseudodislocation is seen, as in this
example, a search for an occult fracture should ensue.
In this case, as seen in A, a fracture (arrow) was initially
missed.
A
FIGURE 4-16 ■ Pseudodislocation of the shoulder.
The humeral head is inferiorly placed in relation to
the glenoid. This is the characteristic location when
a hemarthrosis is present. A minimally displaced
fracture of the neck of the humerus with avulsion of
the greater tuberosity has occurred, causing the
hemarthrosis.
B
FIGURE 4-18 ■ Sacral insufficiency fracture. A, A woman
with a history of breast cancer presented with sacral
pain, and linear low signal adjacent to the sacroiliac
joint was seen (arrow). This was called an insuffi-
ciency fracture by the radiologist; however, a bone
scan was interpreted as possible metastatic disease.
Therefore the patient underwent radiation therapy.
B, She had worsening sacral pain, and an MRI 6 months
later shows bilateral sacral insufficiency fractures.
Radiation is one of the causative factors for insuffi-
ciency fractures and, in this case, caused additional
pain and suffering when the initial imaging examina-
tion was diagnostic.
B
A normal variant that has been described in FIGURE 4-19 ■ Supraacetabular insufficiency fracture.
the patella is a lytic defect in the upper outer A, An elderly woman with a history of breast cancer
presented with right hip pain and had an ill-defined
quadrant called dorsal defect of the patella area of sclerosis in the right supraacetabular region
(Figure 4-20).6 It can mimic a focus of infection, (compared with the opposite side). A biopsy was
osteochondritis dissecans, or a lytic lesion. It is a requested, but the radiologist thought this might
normal developmental anomaly, however, and represent an insufficiency fracture. B, An MRI was
obtained, which revealed a curvilinear fracture line
because of its characteristic location, a biopsy of (arrow), characteristic for a supraacetabular insuffi-
it should not be performed. It is diagnosed based ciency fracture.
on MRI examination and its characteristic loca-
tion, because the signal characteristics are similar
to those of tumor or infection (Figure 4-21). have been repeated (see Figure 4-22, C) after
Another entity often confused with a lytic the initial pathology report stated “normal
pathologic process is a pseudocyst of the hu- bone—no lesion in specimen.” Because of the
merus (Figure 4-22). This is merely an anatomic associated hyperemia from the shoulder disor-
variant caused by the increased cancellous bone der (be it rotator cuff or whatever), a bone scan
in the region of the greater tuberosity of the can show increased radionuclide uptake and
humerus, which gives this region a more lucent thus sway the surgeon to do a biopsy of this
appearance on radiographs. With hyperemia normal variant. It is radiographically character-
and disuse caused by rotator cuff problems or istic in its location and appearance, and a biopsy
any other shoulder disorder, this area of lucency of it should not be done.7 Although other le-
may appear strikingly more lucent and mimic a sions, such as a chondroblastoma (Figure 4-23),
lytic lesion. Biopsies of many of these lesions infection, or even a metastatic focus, could oc-
have been done mistakenly and in several cases cur in a similar location, they do not have quite
Copyright © 2014 Elsevier, Inc. All rights reserved.
4 “Don’t Touch” Lesions 65
A B
FIGURE 4-20 ■ Dorsal defect of the patella. A lytic defect in the upper outer quadrant of the patella (A) was seen in
this patient (arrows), which is characteristic for a normal variant called dorsal defect of the patella. It occurs only
in the upper outer quadrant and should be asymptomatic. It lies adjacent to the articular surface, as shown on the
sunrise view (B).
A B
FIGURE 4-21 ■ MR images of dorsal defect of the patella. (A) T1-weighted and (B) T2-weighted axial MR images
through the patella in a patient with a lytic lesion in the upper outer quadrant of the patella shows low signal on
the T1 image, which is high signal on T2. This is a characteristic appearance for a dorsal defect of the patella
because of its location.
the same appearance as a pseudocyst of the they are all unstable. Radiologists should recog-
humerus. nize that this process is not acute, and as such,
A normal variant of the cervical spine that the patient can be saved from having Crutchfield
may, in fact, be posttraumatic is an os odontoi- tongs or a halo applied and from possible im-
deum.8 It is an unfused dens that may move mediate surgical intervention. Most patients are
anterior to the C2 body with flexion and can seen after trauma occurred, and if no neurologic
mimic a fractured dens (Figures 4-24 and 4-25). deficits are present, these patients can be seen
Many of these lesions require surgical fixation; electively and spared the horrors and morbidity
some surgeons fuse every case, believing that associated with treatment of the acutely fractured
Copyright © 2014 Elsevier, Inc. All rights reserved.
66 4 “Don’t Touch” Lesions
A B
A B
FIGURE 4-24 ■ Os odontoideum. Flexion (A) and extension (B) views show that the anterior arch (a) of the C1 ver-
tebra has moved markedly posterior in relation to the body of C2 in extension. The odontoid or dens is difficult
to see but appears to be separated from the body of C2. Because of the smooth borders of the separated dens
and because of the cortical hypertrophy of the anterior arch of C1, this can safely be called an os odontoideum,
which is a congenital or long-standing posttraumatic abnormality rather than an acute fracture. Obviously these
patients should have no neurologic problems, yet in many instances the lesions are still believed to be unstable
and are surgically fused. Surgery, if indicated, can be done on an elective basis.
A B
FIGURE 4-25 ■ Os odontoideum. Extension (A) and flexion (B) show extreme motion of the anterior arch (a) of C1
as compared with the C2 vertebral body. The dens is difficult to find in this example but is certainly not attached
to the C2 body. Again, the smooth margins where the dens should be attached and the cortical hypertrophy of
the anterior arch of C1 make this a congenital or long-standing process consistent with an os odontoideum rather
than an acute fracture.
cervical spine. The radiologic signs for recogniz- as benign and left alone. These are lesions that
ing an os odontoideum are the smooth, often should be diagnosed by the radiologist, not the
well-corticated inferior border of the dens and pathologist. Listing a differential in such cases
the hypertrophied, densely corticated anterior often spurs the surgeon to perform a biopsy,
arch of C1. This latter finding presumably rep- when in fact no biopsy is necessary.
resents compensatory hypertrophy and indicates Perhaps the most commonly encountered le-
a long-standing condition.9 sion in this category is the nonossifying fibroma
Although unusual, osteopoikilosis, a benign (NOF). It is identical to a fibrous cortical defect,
familial process of multiple bone islands or small but the term is usually reserved for defects larger
areas of osteosclerosis, has caused confusion than 2 cm. Classically, NOFs are lytic lesions
because of its similarity to metastatic disease that are located in the cortex of the metaphysis
(Figure 4-26). Ordinarily osteopoikilosis has of a long bone, which have a well-defined and
such a characteristic appearance that it will not often sclerotic, scalloped border with slight cor-
be mistaken for another entity, and the pre- tical expansion (Figure 4-28) (see also Chapter 2).
dominance of sclerotic foci near the epiphyses They are almost exclusively found in patients
should help differentiate it from metastatic dis- younger than 30 years of age, suggesting that
ease (Figure 4-27). the natural history of the lesion is involution.
As they involute, they fill in with new bone
(Figure 4-29); hence they can have some in-
REAL BUT OBVIOUSLY BENIGN creased radionuclide activity on bone scans.
LESIONS They are most often mistaken for an area of
infection, eosinophilic granuloma, or aneurys-
Biopsies are frequently performed on some le- mal bone cyst. Patients with NOFs are asymp-
sions that should be recognized radiographically tomatic, and NOFs have never been reported to
A B
C D
FIGURE 4-26 ■ Diffuse metastatic disease mimicking osteopoikilosis. A and B, A CT scan through the pelvis and
hips shows diffuse sclerotic foci consistent with metastatic disease. One examiner thought that this might repre-
sent the sclerotic foci of osteopoikilosis, however, which is a benign familial process. C, An anteroposterior view
of the pelvis shows a similar appearance; however, a destructive lytic lesion is seen in the right proximal femur,
which makes metastatic disease more likely. This patient had metastatic prostate carcinoma. Compare this
with D in a patient with known osteopoikilosis, and it is easy to see how the two entities can be confused. Clinical
history is vital in the distinction.
B
FIGURE 4-28 ■ Nonossifying fibroma. A well-defined, FIGURE 4-29 ■ Resolving nonossifying fibroma. A, A
slightly expansile lytic lesion is seen in a long bone, minimally sclerotic, slightly expansile process is seen
which is characteristic for a nonossifying fibroma. in the posterior proximal tibia (arrows). This was be-
Unfortunately a biopsy of the lesion was performed, lieved by the surgeons to represent a focus of infection
and the diagnosis was surgically confirmed. or an osteoid osteoma, even though the patient was
asymptomatic. This is a characteristic appearance for a
disappearing or resolving nonossifying fibroma. The
postsurgical appearance, which went on to a patho-
logic fracture, is shown in B. Surgery confirmed a
nonossifying fibroma.
be associated with malignant degeneration. On This is characteristic of a bone island and help-
occasion a pathologic fracture can occur through ful in differentiating it from more aggressive
these lesions, but most surgeons do not advocate processes.
prophylactic curettage to prevent fracture, as Unicameral bone cysts are often prophylacti-
they would with unicameral bone cysts. NOFs cally packed, thereby preventing fracture with
can be quite large but invariably have a benign subsequent deformity. When these cysts occur in
appearance (Figure 4-30), and biopsy should be the calcaneus, however, they should be left alone.
avoided. The asymptomatic nature is imperative They always occur in the anteroinferior portion
to help distinguish them from most of the other of the calcaneus (Figure 4-33), an area that does
lesions in the differential, thereby precluding not receive undue stress. In fact, a pseudotumor
even giving a differential diagnosis. On occasion of the calcaneus in the identical position is seen
they are found to be multiple (Figure 4-31), yet because of the absence of stress and resulting
each lesion is so characteristic that they should atrophy of bony trabeculae (Figure 4-34). Calca-
be easily diagnosed. neal unicameral bone cysts are asymptomatic,
Bone islands are not a radiographic dilemma only rarely fracture, and should not suffer the
when they are 1 cm or smaller. Occasionally, same fate as their counterparts in long bones.
however, they grow to golf ball size and mimic Calcaneal unicameral bone cysts seem to involute
sclerotic metastases (Figure 4-32). Bone islands at a slower rate than those in long bones; hence
are always asymptomatic. Radiographically, two they can be seen in patients older than 30.
signs can be found to help distinguish giant Also, as they involute they occasionally undergo
bone islands from metastases: first, bone islands “lipidization,” which is fatty infiltration. If im-
usually are oblong, with their long axis in the aging or a biopsy is done during this stage, they
axis of stress on the bone (e.g., in a long bone are mistakenly diagnosed as an intraosseous
they align themselves along the axis of the di- lipoma. This is why many authors have called
aphysis); second, the margins of a bone island, calcaneal unicameral bone cysts lipomas. Annu-
if examined closely, will show bony trabeculae ally, around radiology oral boards time, resi-
extending from the lesion into the normal dents ask me to explain why some texts refer to
bone in a spiculated fashion.10 (See Chapter 2.) these lesions as unicameral bone cysts while
A B
FIGURE 4-30 ■ Nonossifying fibroma. A and B, This well-defined, minimally expansile lytic lesion of the proximal
tibia is characteristic for a nonossifying fibroma. It was believed by several radiologists to be a giant cell tumor;
however, it has a sclerotic border and does not abut the tibial articular surface. Even though the patient was
asymptomatic, a biopsy was performed and the diagnosis of a nonossifying fibroma was confirmed.
A B
FIGURE 4-31 ■ Multiple nonossifying fibromas. Multiple well-defined lytic lesions (arrows) are seen around the
knees in this patient on the anteroposterior (A) and lateral (B) views, each of which is characteristic for a non-
ossifying fibroma.
A B
FIGURE 4-32 ■ Giant bone island. A, A large sclerotic focus is seen in the right iliac wing, which was thought to
possibly represent an area of metastasis. B, Old films from 5 years earlier were obtained, which showed a similar
but much smaller process. This is characteristic for a growing bone island. Note in A how the lesion is somewhat
spherical or oblong in the lines of trabecular stress, which is characteristic of a bone island.
others call them lipomas. In fact, they are all uni- permeative process as opposed to the classic ap-
cameral bone cysts. Many benign fibro-osseous pearance of a sclerotic, serpiginous peripheral
lesions throughout the skeleton involute with border (Figures 4-35 and 4-36). In a patient with
fatty infiltration and are mistakenly called intraos- bone pain and a permeative bone lesion, many
seous lipomas. aggressive disorders head the differential list and
Early in the course of its development, a a biopsy soon ensues. If this process can be noted
bone infarct can have a patchy or a mixed lytic- to be multiple and in the diametaphyseal region
sclerotic pattern, or it may even resemble a of a long bone, especially if the patient has an
Copyright © 2014 Elsevier, Inc. All rights reserved.
72 4 “Don’t Touch” Lesions
A B
FIGURE 4-37 ■ Bone infarct. A, A plain film of the knee shows faint patchy sclerosis in the proximal tibia that was
at first thought to be infection or malignancy. B, MRI scan shows the characteristic serpiginous border seen with
bone infarct. MRI can on occasion better characterize the ill-defined early bone infarct, as in this example.
B C
FIGURE 4-38 ■ Pitt’s pit. A, A plain film of the left hip shows a well-defined lytic lesion in the lateral aspect of the
femoral neck. It has a sclerotic border. This is a characteristic appearance of a synovial herniation pit, also called
a Pitt’s pit. T1-weighted (B) and T2-weighted (C) magnetic resonance axial images through the hips show bilateral
lesions with low-signal T1 and high-signal T2, which is typical for Pitt’s pits.
underlying disorder such as sickle cell anemia or cystic changes in the underlying femoral neck—
systemic lupus erythematosus, areas of early a Pitt’s pit.
bone infarction should be considered. In several
instances the MRI appearance of an infarct has
saved patients from biopsy when the plain films CONCLUSION
were equivocal (Figure 4-37).11
A commonly encountered lytic lesion on the These are but a few of the many examples in
lateral aspect of the femoral neck was first de- skeletal radiology in which the well-trained
scribed by Michael Pitt as a synovial herniation radiologist can be of invaluable assistance to the
pit.12 It has taken on the appropriate eponym of clinician and the patient by helping avert a
a Pitt’s pit. It is believed to be caused by surface needless biopsy. Dozens of other examples are
erosion of the synovium and soft tissues around nicely shown in normal variant textbooks, which
the hip, but its exact etiology is unknown. It has are widely available. Because of the potential
a characteristic plain film and MRI appearance harm in performing a needless biopsy, these
primarily because of its location and benign examples are stressed. When these lesions are
appearance (Figure 4-38). Pitt’s pits are found encountered by the radiologist, a differential
much more frequently in patients with femoro- diagnosis should not be offered, because it will
acetabular impingement (FAI), a process in merely lead the surgeon to a biopsy in an at-
which the femoral neck abuts the rim of the tempt to reach a diagnosis. A biopsy of many of
acetabulum when the hip is abducted (see these entities not only is unnecessary but also
Chapter 13). It is thought this likely results in can be misleading.
Trauma
Radiology of trauma to the skeletal system emergency department, and it can present
is such a large topic that entire volumes have the most difficulty in interpretation. Usually
been devoted to it. Lee Rogers has written the a cross-table lateral view of the C-spine is
definitive work in his excellent book titled Radi- obtained first so as not to unduly move the
ology of Skeletal Trauma,1 and Jack and William patient who might have a cervical fracture.
Harris’s outstanding book on The Radiology of If the lateral C-spine film appears normal,
Emergency Medicine2 is a must-read for anyone the remainder of the C-spine series, which
dealing with a large emergency department may include flexion and extension views (if the
population. The leading orthopedic treatise on patient can cooperate), is obtained.
fractures is Rockwood and Green’s multivolume What do you look for on the lateral C-spine?
text.3 The following is merely an overview First, make certain that all seven cervical vertebral
of selected cases that residents and medical stu- bodies can be visualized. A number of fractures
dents should be exposed to and that can be are missed because the shoulders obscure the
studied in greater detail by referring to the texts lower C-spine levels (Figure 5-1). If the entire
just mentioned. cervical spine is not visualized, repeat the film
Before specific examples are given, the unini- with the shoulders lowered.
tiated or neophyte radiologist should keep a few What constitutes complete visualization of
key points in mind concerning radiology of the cervical spine? Many radiologists insist on
trauma. First, have a high index of suspicion. seeing the top of the T1 vertebral body on the
Every radiologist in the world has missed frac- lateral view, whereas others will pass a C-spine
tures on radiographs because they were not suf- lateral film if it includes any of the C7 body.
ficiently attuned to the fact that a fracture might Most textbooks say the lateral view should
be present. Often the history is either nonexis- show “C1 through C7.” What does “through
tent or misleading, and the anatomic area of C7” mean? I have no idea, but I was trained to
concern is therefore overlooked. When in doubt, accept a lateral C-spine film that included any
examine the patient! Orthopedic surgeons sel- of the C7 body. It can be difficult to image the
dom miss seeing fractures on radiographs be- T1 vertebral body in the majority of cases;
cause they have examined the patient, they know therefore I believe it is acceptable to accept a
where the patient hurts, and they have a high lateral C-spine film with any of the C7 vertebral
index of suspicion. Second, in every trauma case, body visible if the report includes the disclaimer
get two radiographs at 90 degrees to each other. that the C7–T1 disc space is not seen and clini-
A high percentage of fractures are seen only on cal correlation must be obtained to warrant
one view (the anteroposterior [AP] or the lateral) additional films or a computed tomography
and will therefore be missed unless two views are (CT) scan of that area. In fact, this is a moot
routinely obtained. Third, once a fracture is point in many practices where physicians rou-
identified, don’t forget to look at the rest of the tinely acquire CT scans throughout the C-spine
film. About 10% of all cases have a second find- instead of plain films. Complete CT evaluation
ing that often is as significant as, or even more of the cervical spine will probably evolve to be
significant than, the initial finding. Many frac- the standard of care, but it is not yet widely
tures have associated dislocation, foreign bodies, accepted as the normal routine.
or additional fractures, so be sure to examine the Next, evaluate five (more or less) parallel
entire film. lines for step-offs or discontinuity as follows
(Figure 5-2):
Line 1 is in the prevertebral soft tissue. It
SPINE extends down the posterior aspect of the
airway; it should be several millimeters
from the first three or four vertebral bodies
Examination of the Cervical Spine and then move further away at the laryn-
The cervical spine is one of the most com- geal cartilage; it should be less than one
monly radiographed parts of the body in a busy vertebral body width from the anterior
76 Copyright © 2014 Elsevier, Inc. All rights reserved.
5 Trauma 77
A B
FIGURE 5-1 ■ Shoulders obscuring C7. This patient came to the emergency department after being injured as a result
of diving into a shallow swimming pool. He had neck pain but no neurologic deficits. The initial radiograph obtained
of the C-spine (A) was interpreted as being within normal limits. However, because of high-riding shoulders only
five cervical vertebrae are visible. A repeat examination (B) with the shoulders lowered reveals a dislocation of
C5 on C6. To visualize C7 the shoulders were lowered even further. The C7 vertebral body must be visualized on
every lateral C-spine examination in a trauma setting.
A B 1 2 3 4
FIGURE 5-2 ■ Normal lateral cervical spine. A, Lateral radiograph of a normal cervical spine. B, Diagrammatic repre-
sentation of a lateral C-spine showing four parallel lines that should be observed in every lateral C-spine examination.
Line 1 is the soft tissue line that is closely applied to the posterior border of the airway through the first four or
five vertebral body segments and then widens around the laryngeal cartilage and runs parallel to the remainder
of the cervical vertebrae. Line 2 demarcates the anterior border of the cervical vertebral bodies. Line 3 is the posterior
border of the cervical vertebral bodies. Line 4, called the spinolaminal line, is drawn by connecting the junction of the
lamina at the spinous process. It represents the posterior extent of the central canal, which contains the spinal cord
itself. These lines should generally be smooth and parallel, with no abrupt step-offs.
2.5 mm
C-2
A B
FIGURE 5-3 ■ Normal C1–2. A lateral radiograph (A) and drawing (B) of the upper cervical spine, showing the normal
distance of the anterior arch of C1 to be less than 2.5 mm from the odontoid process of C2 (arrows).
8.0 mm
C-2
A B
FIGURE 5-4 ■ C1–2 dislocation. A lateral radiograph (A) and drawing (B) of the upper cervical spine in a patient
with trauma to the neck, which shows that the anterior arch of C1 is 8 mm anterior to the odontoid process of
C2 (arrows). This is diagnostic of a dislocation of C1 on C2, and indicates rupture of the transverse ligaments,
which normally hold these vertebral segments together.
The disc spaces are examined next to check radiologists cannot turn the film to their liking.
for any inordinate widening or narrowing, either Get used to viewing lateral spine films (lateral
of which could indicate an acute traumatic injury. chest films, too) in either anterior left or right
If a disc space is narrowed, it will usually be orientation; otherwise, you will find yourself
secondary to degenerative disease. Make certain disadvantaged in many situations.
that associated osteophytosis and sclerosis are
present, however, before assuming the narrowing
is from degenerative disease. Examples of Fractures,
An examination of the lateral C-spine as de- Dislocations, and Other
scribed earlier can be done in less than 1 minute.
If this view is normal, then the remainder of
Abnormalities
the examination can be completed, including A blow to the top of the head, such as when an
flexion and extension views. It is imperative that object falls directly on the apex of the skull, can
the patient initiate the flexion and extension cause the lateral masses of C1 to slide apart,
without help from the technician or anyone else. splitting the bony ring of C1. This is called a
A patient, if conscious and semialert, will not Jefferson’s fracture (Figure 5-5). It nicely illus-
injure himself or herself with voluntary flexion trates how a bony ring will not break in just
and extension and will have muscle guarding, one place but must break in several places. This
preventing motion, if an injury is present. Even rule is seldom violated. All the vertebral rings,
gentle pressure to aid in flexion or extension can when fractured, must fracture in two or more
cause severe injury if a fracture or dislocation is places. The bony rings of the pelvis behave the
present. same way. If you see only one fracture on
Learning to look at lateral spine films with the radiograph, you can be certain you are miss-
anterior facing either right or left is very impor- ing at least one more. A CT scan is excellent at
tant. Many radiologists can interpret images demonstrating the complete bony ring of C1
only facing one way and become almost unable and shows the fractures, as well as any associated
to function if the films are not placed on the soft tissue mass, much better than plain films. In
viewbox in their preferred orientation. This diagnosing a Jefferson’s fracture on plain film
is fine if they can control the film; however, the lateral masses of C1 must extend beyond the
in meetings where slides are shown, in books margins of the C2 body (see Figure 5-5, A). Just
and journals, and on oral board examinations, seeing asymmetry of the spaces on either side of
Copyright © 2014 Elsevier, Inc. All rights reserved.
80 5 Trauma
A B
FIGURE 5-5 ■ Jefferson’s fracture. A, An AP open-mouth odontoid view is suspicious for the lateral masses
of C1 being laterally displaced. However, because of overlying structures, this displacement is difficult to
appreciate. B, Therefore, a CT scan was obtained, which shows multiple fracture sites in the C1 ring (arrows).
This is called a Jefferson’s fracture. CT is routinely used in spinal trauma because of the obvious shortcomings
of plain films.
the dens is not enough to make the diagnosis, (supraspinous ligaments) to undergo a tremen-
because these spaces can be normally asymmet- dous force, pulling on the spinous process and
ric with rotation or with rotatory fixation of the avulsing it. This fracture can occur at any of the
atlantoaxial joint. lower cervical spinous processes (Figure 5-7).
What is rotatory fixation of the atlantoaxial A hangman’s fracture is an unstable, serious
joint? This is a somewhat controversial, little fracture of the upper cervical spine that is
understood process in which the atlantoaxial caused by hyperextension and distraction (e.g.,
joint becomes fixed and the C1 and C2 bodies hitting one’s head on a dashboard). This is a
move en masse instead of rotating on each other. fracture of the posterior elements of C2 and,
This condition is easily diagnosed with open- usually, displacement of the C2 body anterior to
mouth odontoid views. In the normal odontoid C3 (Figure 5-8). Patients with this type of frac-
view the spaces lateral to the dens (odontoid) are ture actually do better than one might think.
equal. With rotation of the head to the left, the They often escape neurologic impairment be-
space on the left widens; and with rotation to the cause of the fractured posterior elements of
right, the space on the right widens. With rota- C2, which, in effect, cause decompression and
tory fixation, one of the spaces is wider than the take pressure off the injured area. This is a sim-
other and stays wider even with rotation of the plistic explanation for a complex entity, but it
head to the opposite side (Figure 5-6). By itself, seems to be a reasonable answer to why these
this is a relatively innocuous malady usually patients often fare well.
treated with a soft cervical collar and/or gentle Severe flexion of the cervical spine can cause
traction. It is, however, occasionally associated a disruption of the posterior ligaments and ante-
with disruption of the transverse ligaments at rior compression of a vertebral body. This is
C1–2, and is then a serious problem. Rotatory called a flexion “teardrop” fracture (Figure 5-9).
fixation usually presents spontaneously or after A teardrop fracture is usually associated with
very mild trauma, such as that caused by sleeping spinal cord injury, often from the posterior
in an unusual position. portion of the vertebral body being displaced
Another relatively innocuous injury is a frac- into the central canal.
ture of the C6 or C7 spinous process, called If severe enough, and if associated with some
a clay-shoveler’s fracture. Supposedly workers rotation, the apophyseal joint ligaments will
shoveling sticky clay in Australia (I’ve also read rupture and the facet joints dislocate and then
England and North Carolina—this is a vital dis- override. This can result in locking of the facets
tinction, and some future researcher can perhaps in an overriding position, which in effect causes
straighten out this confusion) would toss the some stabilization to protect against further in-
shovelfuls of clay over their shoulders; once in a jury. This condition is called unilateral locked
while the clay would stick to the shovel, causing facets (Figures 5-10 and 5-11), but occasionally it
the ligaments attached to the spinous processes occurs bilaterally. When unilateral, the more
Copyright © 2014 Elsevier, Inc. All rights reserved.
5 Trauma 81
A B
A B
FIGURE 5-8 ■ Hangman’s fracture. A, A lateral film of a patient with a hangman’s fracture shows an obvious
example of the posterior elements of the C2 vertebral body fractured and displaced inferiorly. B, The film
of another patient shows a subtle fracture through the posterior elements of C2 (arrow). A line drawn
through the spinolaminal lines of the posterior elements shows the C2 spinolaminal line to be offset in this
example.
C6
C7
FIGURE 5-9 ■ Flexion teardrop fracture. This patient FIGURE 5-10 ■ Unilateral locked facets. The C6–7 disc
suffered hyperflexion in a car accident and came to space is abnormally widened, and the C7 vertebral
the emergency department with severe neurologic body is posteriorly located in relation to C6. Also note
deficits. A lateral radiograph of the lower cervical the C7 facets, which are dislocated and locked on the
spine shows wedging anteriorly of the C7 vertebral C6 facets (arrow). When the facets are perched in this
body with some displacement of the posterior verte- manner, the condition is termed locked facets, which
bral line of C7 into the central canal. A small avulsion in this example is unilateral.
fracture off the anterior body is also noted.
C6
C7
A B
FIGURE 5-11 ■ Unilateral locked facets. A, The C6–7 disc space is slightly wider than normal and the C7 vertebral
body is narrower in this view than the C6 vertebral body because of abnormal rotation. B, The facets of C7 are
not well identified, but in the tomogram they are shown to be overriding and locked (arrow).
A B
FIGURE 5-12 ■ Seat belt fracture. A, Hyperflexion at the waist can cause anterior wedging of the vertebral body in
the lower thoracic or upper lumbar region. By itself, that is somewhat innocuous. B, However, here a horizontal
fracture through the right transverse process and pedicle (arrow) resulting from extreme traction during the flex-
ion injury can be seen. When fracture of the posterior elements occurs, this injury is considered unstable and
potentially debilitating. Any anterior wedging injury to a vertebral body should have the posterior elements of
that level closely inspected for interpedicular space widening.
figure of a Scottie dog, with the transverse pro- will have a defect or break. It often looks as if the
cess being the nose, the pedicle forming the eye, Scottie dog has a collar around its neck. Although
the inferior articular facet being the front leg, the often difficult to visualize with magnetic reso-
superior articular facet representing the ear, and nance imaging (MRI), spondylolysis should be
the pars interarticularis (which means the portion easily seen on CT scans (see Chapter 11).
of the lamina that lies between the facets) equal- The cause of a spondylolysis is said by some
ing the dog’s neck. If a spondylolysis is present, investigators to be congenital and by others to
the pars interarticularis, or the neck of the dog, be posttraumatic. Many believe that this is a
Copyright © 2014 Elsevier, Inc. All rights reserved.
84 5 Trauma
A B SPONDYLOLYSIS
FIGURE 5-13 ■ Spondylolysis. A, An oblique film of the lumbar spine shows a defect in the neck of the Scottie dog
at L5 (arrow), which is diagnostic of a spondylolysis. B, This drawing shows the findings more clearly. This has
been described as a collar around the Scottie dog’s neck.
stress-related injury from infancy that develops the initial trauma. I have seen more than a dozen
when toddlers try to walk and repeatedly fall lawsuits against radiologists who failed to men-
on their buttocks, sending stress to their lower tion minor anterior wedging of a vertebral body
lumbar spine. The significance of spondylolysis is that went on to further collapse with associated
just as controversial as its cause. More and more paraplegia. All that needs to be mentioned is that
clinicians are coming to the viewpoint that a a fracture of indeterminate age is present and
spondylolysis is an incidental finding with no requires clinical correlation. If the patient has
clinical significance in most cases. Certainly some pain in that location, the patient needs to wear a
patients have pain related to a spondylolysis and back brace until pain free. Old films can help
get relief after surgical stabilization, but such determine whether it is an old fracture. If no
cases are less common. pain is present on physical examination, it can be
If a spondylolysis is bilateral and the vertebral safely assumed to be an old fracture. It is not
body in the more cephalad position slips forward necessary to obtain a CT or MRI scan even if
on the more caudal body, spondylolisthesis is pain is present, because the treatment will be the
said to be present (Figure 5-14). Spondylolisthesis same regardless of what the CT or MRI scans
may or may not be symptomatic and by itself has reveal. No spine surgeon will operate on a stable
no clinical significance. If severe, it can cause spine fracture without kyphosis or neurologic
neuroforaminal stenosis and can impinge on the deficits, so the CT or MRI examination add
nerve roots in the central spinal canal. If it is nothing but time and expense.
symptomatic, it can be surgically stabilized. Patients who have fusion of their spine from
Anterior wedge compression fractures of the ankylosing spondylitis and, to a lesser extent,
spine are commonly seen (Figure 5-15), espe- from diffuse idiopathic skeletal hyperostosis
cially at the thoracolumbar junction, as a result (DISH; see Chapter 6) are at a very high risk of
of an old injury; they are often passed off by the spinal fractures from even relatively minor
radiologist, if they are mentioned at all, as inci- trauma. Patients with ankylosing spondylitis
dental findings. The problem with this is it’s typically have marked osteoporosis, which fur-
impossible to tell from a plain film if the fracture ther magnifies their risk of sustaining a frac-
is old or new, even if degenerative changes ture. A fused spine is more likely to fracture
are present (which are often not related to the than a normal spine, much in the same way that
fracture). If acute and left unprotected, a wedge a long glass pipette breaks more easily than
compression fracture can proceed to delayed a short one because it has a long lever arm.
further collapse with resulting severe neurologic A small force at one end is greatly magnified
deficits (Figure 5-16). This is called Kummell’s further down the lever arm. For that reason, a
disease and typically occurs 1 to 2 weeks after patient with ankylosing spondylitis should be
Copyright © 2014 Elsevier, Inc. All rights reserved.
5 Trauma 85
L4
Offset
L5
S!
A B Spondylolisthesis
FIGURE 5-14 ■ Spondylolisthesis. A, A lateral film of the lumbar spine shows the L5 vertebral body is slightly an-
teriorly offset on the S1 body as noted by the posterior margins (arrows). B, The drawing shows the findings more
clearly. Because the offset is less than 25% as measured by the length of the S1 end-plate, it is termed a grade 1
spondylolisthesis. A grade 2 offset is more than 25% but less than 50% of the length of the S1 end-plate.
L1
L1
A B
FIGURE 5-16 ■ Kummel’s disease. A, Very minimal anterior wedging of the L1 vertebral body is noted by comparing
the height of the anterior body versus the posterior height. This patient had been in an auto accident and com-
plained of back pain. No treatment for his back was given. B, After several weeks of continuing pain the patient
sought treatment for leg weakness, which proceeded to paraplegia. A spine film now shows progression of the
vertebral body collapse of L1. This almost certainly could have been avoided with simple bracing of the spine
after the initial injury.
L1
A B
FIGURE 5-17 ■ Spine fracture in ankylosing spondylitis. A, A lateral spine plain film after trauma shows fusion of
the spine anteriorly, which was secondary to ankylosing spondylitis. Minimal anterior wedging of the L1 vertebral
body, which was overlooked, is present. B, Two weeks later a CT scan of the spine was performed because of the
sudden onset of paralysis. This axial image through L1 shows a fracture of the posterior elements, which was
undoubtedly present on the initial visit to the emergency department. Patients with ankylosing spondylitis need
to be examined closely for any back pain after trauma, and CT or MRI scans should be obtained if any pain is
present.
FIGURE 5-20 ■ Mallet finger. A small avulsion injury is noted at the base of the distal phalanx, which is where the
extensor digitorum tendon inserts. This injury is termed a mallet finger or baseball finger. It is often caused by a
baseball striking the distal phalanx, causing the avulsion.
FIGURE 5-21 ■ Gamekeeper’s thumb. A small avulsion FIGURE 5-22 ■ Normal lateral radiograph of the wrist.
injury on the ulnar aspect of the first metacarpopha- The normal lateral view should show the lunate (L)
langeal joint (arrow) is diagnostic of a gamekeeper’s seated in the distal radius and the capitate (C) in turn
thumb. This is the insertion site for the ulnar seated in the lunate. A line drawn through the radius
collateral ligament and usually requires internal should connect all three structures. Compare this ra-
fixation. diograph with the drawing in Figure 5-23, A.
a perilunate dislocation into a lunate dislocation the median nerve if it gets impinged on by the
merely by manipulating the wrist. Therefore lunate.
strict classification of these entities is not recom- I strongly recommend that every radiologist
mended by everyone. For many surgeons it is a get in the habit of looking at the alignment of
moot point—they want rapid reduction of the the lunate and the capitate on every lateral wrist
capitate-lunate dislocation and don’t really care film. There’s not a whole lot else to look for on
which one happens to be more volarly or dor- the lateral view anyway. I see a missed perilunate
sally displaced. Failure to diagnose and treat this dislocation every few years, and they usually end
disorder can result in permanent impairment of up in litigation.
Copyright © 2014 Elsevier, Inc. All rights reserved.
5 Trauma 89
A B C
FIGURE 5-23 ■ Schematic depiction of normal lateral wrist (A), perilunate dislocation (B), and lunate dislocation (C).
(The dark shaded bone is the capitate; the cross-hatched filled bone is the lunate. Ventral is to the left.)
A B
FIGURE 5-25 ■ Lunate dislocation. A, The lateral radiograph of the wrist shows the lunate tipped off the distal radius,
whereas the capitate seems to be normally aligned in relation to the radius yet is dislocated from the lunate. Compare
this with the drawing in Figure 5-23, C. B, The AP view shows a pie-shaped lunate (L) rather than a lunate with a more
rhomboid shape. A pie-shaped lunate can be seen in a perilunate or lunate dislocation.
in radiology and sports medicine circles is that continue their participation, which can lead to a
of a professional athlete who participates in an nonunion of the fracture.
activity in which the butt of a club, bat, or racket Another wrist injury that is seen after a fall
is held in the palm. Overswinging can result in onto the outstretched hand is rotatory subluxation
the butt of the club levering off the hook of the of the scaphoid. This results from rupture of the
hamate. This has been seen in professional base- scapholunate ligaments, which allows the scaph-
ball players, tennis players, and golfers. Why oid to rotate dorsally. On an AP wrist radiograph
professionals? Amateurs usually are not strong a space is seen between the scaphoid and the
enough to exert enough force to lever the hook lunate (Figure 5-28) when ordinarily they
off and, if they do, will usually terminate that are closely opposed. This space has been called
activity, allowing healing. Professionals, however, the Terry-Thomas sign, after a famous British
Copyright © 2014 Elsevier, Inc. All rights reserved.
5 Trauma 91
FIGURE 5-28 ■ Rotatory subluxation of the scaphoid. FIGURE 5-29 ■ Scaphoid fracture. A coronal T1-weighted
An AP view of the wrist shows a gap or space image of the wrist in a patient with snuffbox tender-
between the scaphoid and the lunate (arrow). This is ness and a normal plain film shows a fracture of the
abnormal and represents the Terry-Thomas sign, mid-waist of the scaphoid (arrow).
which means that the ligaments between the lunate
and the scaphoid are ruptured. This is diagnostic of a
rotatory subluxation of the scaphoid.
FIGURE 5-33 ■ Colles’ fracture. A fracture of the distal FIGURE 5-35 ■ Plastic bowing deformity of the forearm.
radius with dorsal angulation is noted, which has These AP and lateral views of a child’s forearm show
been termed a Colles’ fracture (volar is to the left). the radius to be abnormally bowed. This condition has
been termed a plastic bowing deformity of the forearm
and occurs only in children.
A helpful indicator of a fracture about the el- be treated identically whether or not it is radio-
bow is a displaced posterior fat pad. Ordinarily graphically recorded. As long as there is no obvi-
the posterior fat pad is not visible on a lateral ous deformity or loose body present, it does not
view of the elbow because it is tucked away in the matter if the fracture is definitely identified in a
olecranon fossa of the distal humerus. When the patient with a posttraumatic painful elbow and a
joint becomes distended with blood secondary to visible posterior fat pad.
a fracture, the posterior fat pad is displaced out of Couldn’t an infection cause a joint effusion
the olecranon fossa and is visible on the lateral and a displaced posterior fat pad? Of course, but
view (Figure 5-38). Therefore, in the setting the clinical setting would not be to rule out a
of trauma, a visible posterior fat pad indicates fracture. In fact, any elbow effusion will cause a
a fracture. In an adult (epiphyses closed) the posterior fat pad to be visible.
fracture site is almost always the radial head The anterior fat pad also gets displaced
(see Figure 5-38, B). In a child (epiphyses open) with a joint effusion. Ordinarily it is visible as a
it is usually indicative of a supracondylar fracture small triangle just anterior to the distal humeral
(Figure 5-39). Often the fracture itself is not diaphysis on a lateral film (Figure 5-40). With
visualized, and heroic steps are taken by clini- an effusion it gets displaced superiorly and out-
cians and radiologists alike to demonstrate the ward from the humerus and has been called a
fracture. These steps include oblique views, sail sign because it resembles a spinnaker sail
special radial head views, tomograms, and even (see Figures 5-38 and 5-39). I have seen only
CT and MRI scans. These are costly and unnec- one example of a displaced anterior fat pad
essary attempts to document pathology that will without a visible posterior fat pad, and that was
A B
FIGURE 5-38 ■ Displaced fat pads about the elbow. A, On the lateral view of the elbow in this patient, the posterior
fat pad is visible (arrow) and the anterior fat pad is elevated and anteriorly displaced (curved arrow). These findings
indicate a fracture about the elbow, which in an adult should be in the radial head. B, An oblique view in this patient
shows the fracture of the radial head (arrow). Even without definitely seeing the fracture on the radiographs it
should be surmised to be present when the posterior fat pad is visualized in the setting of trauma. The elevated
and displaced anterior fat pad has been termed a sail sign because it resembles a spinnaker sail.
A B
FIGURE 5-47 ■ Dislocation of the hip. A, An AP plain film of the left hip shows dislocation of the femoral head,
which lies superior to the acetabulum. B, Fractures are easily identified on the CT scan. A cortical break through
the articular surface of the posterior acetabulum, as well as the dislocation, is identified.
A B
FIGURE 5-49 ■ Insufficiency fracture of the sacrum. A, Faint sclerosis is noted in the left part of the sacrum as
compared with the right in this patient complaining of pelvic pain. A radionuclide bone scan showed increased
isotope uptake on the left half of the sacrum, and metastatic disease was postulated. B, A CT scan through this
region demonstrates a cortical disruption (arrow) indicative of a fracture. This is a characteristic plain film and
CT appearance of an insufficiency fracture of the sacrum.
A B
FIGURE 5-52 ■ Avulsion off the ischium. A, An AP view of the pelvis shows calcification extending off the left
ischium (arrow) in a patient complaining of pain at this site. Note the irregular cortical surface, suggesting
periostitis. B, The CT scan shows dense calcification adjacent to the ischium (arrow). These findings are
characteristic for an ischial avulsion, and a biopsy should not be done.
In certain joints, however, erosions occur as a also behaves in this manner. Ordinarily when
result of degenerative joint disease. These joints erosions are a feature of an articular process,
include the temporomandibular joint (TMJ), the osteoarthritis (degenerative joint disease) is not
acromioclavicular (AC) joint, and the sacroiliac in the differential diagnosis. It should be if the
(SI) joint. These are easy to remember because TMJ, the AC joint, the SI joint, or the symphy-
they are the “letter joints.” The symphysis pubis sis is involved, however.
Copyright © 2014 Elsevier, Inc. All rights reserved.
102 5 Trauma
B
FIGURE 5-57 ■ Sacroiliac osteophytes. A, An AP view
of the pelvis in this marathoner shows dense sclero-
sis over both sacroiliac joints. B, A CT scan through
this area demonstrates dense bridging osteophytes
characteristic of degenerative disease.
subtle abnormalities. Stress fractures, however, fracture until proved otherwise. Occasionally
need to be considered in anyone with hip or a stress fracture will appear somewhat aggres-
leg pain, because overlooking the diagnosis sive, with aggressive periostitis and no definite
can lead to catastrophe. The most serious stress linearity to the sclerosis (Figure 5-63). If the
fracture—and fortunately one of the rarest—is fracture is mistaken for a tumor and a biopsy is
the femoral neck stress fracture. It has been performed, it can be confused with a malig-
divided into three types by orthopedic surgeons: nancy, which is then typically followed by
type 1—sclerosis without a fracture line evident radical therapy. Therefore, in such cases a
(Figure 5-59); type 2—a lucent fracture line biopsy should not be done under any circum-
without displacement (Figure 5-60); and type stances. If the clinical presentation is unusual
3—an evident displaced fracture. The prognosis for a stress fracture and the plain films are not
is best for a type 1 and worst for a type 3 fracture. diagnostic, take additional films 1 or 2 weeks
Many surgeons believe that type 2 and type 3 later. Sometimes CT and MRI scans will better
fractures require internal fixation, whereas a delineate the lesion. Stress fractures can be dif-
type 1 fracture requires non–weight bearing for ficult to diagnose radiologically early on but
at least 3 to 4 weeks. Many type 1 fractures pro- should be straightforward after several weeks.
gress to complete fractures with displacement A history of repetitive stress is not always
(type 3) with continued weight bearing; therefore obtained, so the diagnosis should not depend
these are considered very serious lesions. If a solely on the history.
femoral stress fracture is a clinical concern and An unusual stress fracture is a fibular stress
plain films are normal, an MRI scan should be fracture (Figure 5-64). The fibula is ordinarily
obtained. not thought of as a weight-bearing bone, but in
Stress fractures also occur in the distal certain people it must serve as such.
diaphysis of the femur and in the proximal, One final stress fracture that deserves mention
middle, and distal thirds of the tibia. All because it is frequently misdiagnosed clinically
of these stress fractures need to be treated and overlooked radiographically is the calcaneal
with the utmost caution, because complete stress fracture (Figure 5-65). It is often misdiag-
fractures are can occur with continued stress nosed clinically as a heel spur or plantar fasciitis,
(Figures 5-61 and 5-62). Sclerosis in a weight- and it can be a subtle radiographic finding.
bearing bone that has a horizontal or oblique An MRI scan can be helpful in cases in which
linear pattern should be considered a stress plain films are negative (Figure 5-66).
FIGURE 5-59 ■ Femoral stress fracture. An area of linear FIGURE 5-60 ■ Stress fracture of the femoral neck. A
sclerosis (arrows) is seen at the base of the femoral linear lucency with surrounding sclerosis is seen in
neck in a neophyte runner with hip pain. This finding the femoral neck in this jogger with hip pain. This is a
is diagnostic of a stress fracture of the femur. severe femoral neck stress fracture.
A B
FIGURE 5-61 ■ Stress fracture with completion. A, A linear lucency is seen in the anterior cortex of the tibia in this
runner, which is diagnostic for a stress fracture. B, This radiograph shows the result of continued exercise.
The stress fracture went on to a complete fracture, which illustrates why any stress fracture of a long bone should
be protected.
A B
FIGURE 5-62 ■ Stress fracture with completion. A, Seen here is a faint linear sclerotic area (arrow), which
is characteristic for a stress fracture of the proximal tibia. B, This radiograph shows the result of continued
exercise in this patient: a complete fracture of the tibia and of the proximal fibula.
A B
FIGURE 5-63 ■ Stress fracture of the tibia. A, An irregular focus of sclerosis is seen in the posterior proximal
tibia with adjacent periostitis. There was concern that this might represent a primary bone tumor, and the
surgeons recommended a biopsy. B, An MRI scan was performed, however, showing a linear low-signal area
running obliquely across the tibia, which is characteristic of a stress fracture. No soft tissue mass was found.
The patient’s recent history included an increase in his jogging, and a stress fracture was diagnosed based on
these images.
FIGURE 5-64 ■ Fibular stress fracture. A linear band of FIGURE 5-65 ■ Calcaneal stress fracture. A linear band
sclerosis with adjacent periostitis (arrow) is seen in of sclerosis is seen in the posterior calcaneus
the distal fibula in this young female jogger. This is (arrows), which is diagnostic for a stress fracture of
diagnostic for a stress fracture of the fibula. the calcaneus.
A B
C
FIGURE 5-66 ■ MRI scan of a calcaneal stress fracture. A, A lateral plain film of the calcaneus in an elderly woman
with heel pain and a history of lung carcinoma shows only osteoporosis. B, A radionuclide bone scan reveals
diffuse increased uptake throughout the calcaneus. The question of whether to treat the calcaneus with radiation
for metastatic lung carcinoma or to do a biopsy first was a dilemma. C, An MRI scan was obtained to get a better
idea of where the met might be. This sagittal T1-weighted image shows a linear low-signal area (arrow), which is
characteristic for a stress fracture. Obviously, no biopsy or radiation was necessary.
Overt fractures in the lower extremity are a fracture. MRI examination has proved to be
uncommonly missed on radiographs; however, very useful in diagnosing hip fractures when
a few exceptions should be noted. Hip frac- plain films are negative (Figure 5-68), and even
tures in the elderly population can be difficult though expensive, they can actually reduce the
to detect (Figure 5-67), and a high index of overall costs by ensuring that no fractures are
suspicion should be maintained. A negative missed.5
plain film in an elderly patient with hip pain Another fracture that can be difficult to exclude
after trauma (even relatively mild trauma) does on routine plain films is a tibial plateau fracture
not exclude a femoral neck fracture. An imme- (Figure 5-69). Many emergency department phy-
diate MRI scan should be obtained to exclude sicians routinely take a cross-table lateral view of
Copyright © 2014 Elsevier, Inc. All rights reserved.
5 Trauma 107
A B
FIGURE 5-67 ■ Fracture of the hip. A, An AP view of the hip was obtained in an elderly man after he had fallen. It
was interpreted as normal, and the patient was discharged from the emergency department. B, Two weeks later
the patient returned to the emergency department because he was unable to walk. Another radiograph was
obtained and revealed a complete fracture through the femoral neck. In retrospect the fracture can be faintly seen
in A and should have been picked up initially. Fractures of the hip in the elderly can be difficult to see and should
be diligently searched for with additional views when the clinical setting is appropriate.
A B
FIGURE 5-68 ■ MRI scan of a hip fracture. A, A plain film of the hip in this elderly patient who has hip pain after a
fall does not show a fracture. B, A T1-weighted MRI scan shows linear low signal in the intertrochanteric region,
which is diagnostic for a hip fracture.
A B
C
FIGURE 5-69 ■ Tibial plateau fracture. A, An AP view of the knee shows no obvious abnormalities at first glance.
B, A CT scan with reformations of the knee, however, demonstrates a plateau fracture of the lateral tibia. Note
the rounded sclerosis, which in retrospect can be barely appreciated in A. C, A T1 coronal MRI scan shows a
tibial plateau fracture, which was barely discernible on plain films. MRI is an excellent imaging choice for subtle
fractures. Tibial plateau fractures are probably the most commonly missed fractures about the knee.
the knee to look for a fat/fluid level in the supra- surgeon in Napoleon’s army who would do fore-
patellar recess. This finding indicates that a frac- foot amputations in patients with gangrenous toes
ture has occurred and allowed fatty marrow to after frostbite. The Lisfranc’s fracture is a frac-
enter the joint; it is highly correlated with a tibial ture-dislocation of the tarsometatarsals. If the
plateau fracture. In the appropriate clinical setting dislocation is minimal, it can be easily overlooked.
CT or MRI examination may be necessary to A key to normal alignment is that the medial bor-
make the diagnosis. der of the second metatarsal should line up with
A serious fracture in the foot that can be the medial border of the second cuneiform and
missed radiographically when little or no dis- the medial border of the fourth metatarsal should
placement occurs is the so-called Lisfranc’s frac- line up with the medial border of the cuboid.
ture (Figure 5-70). It is named after a famous If they do not, a Lisfranc’s fracture-dislocation
Copyright © 2014 Elsevier, Inc. All rights reserved.
5 Trauma 109
20–40 DEGREES
REFERENCES
1. Rogers LF: Radiology of skeletal trauma, ed 3, New York,
2002, Churchill Livingstone.
2. Harris JH Jr, Harris WH: The radiology of emergency
medicine, ed 4, Baltimore, 2000, Lippincott Williams &
FIGURE 5-73 ■ Fracture of the base of the fifth metatar- Wilkins.
sal. This lateral plain film in a woman who sprained 3. Rockwood CA Jr, Green DP: Fractures in adults, ed 5,
her ankle shows the classic appearance of a fracture of Philadelphia, 2001, Lippincott Williams & Wilkins.
the base of the fifth metatarsal (arrow). 4. Dorsay TA, Major NM, et al: Cost-effectiveness of im-
mediate MR imaging versus traditional follow-up for
revealing radiographically occult scaphoid fractures.
AJR Am J Roentgenol 177(6):1257–1263, 2001.
5. Deutsch AL, Mink JH, Waxman AD: Occult fractures
of the proximal femur: MR imaging. Radiology 170:
113–116, 1989.
Arthritis
The radiologic study of arthritis can be ex- can involve the more proximal portion of the
tremely difficult for the inexperienced because hand and wrist, although this is unusual.
of the wide variety of patterns of disease, which Side-to-side symmetry of the arthropathy is
produces a tremendous amount of overlap occasionally helpful in selecting a differential
among the various diseases. What at first seems diagnosis (Box 6-1). Primary osteoarthritis and
to be simple characterization of disease entities rheumatoid arthritis are classically described as
is found by the more experienced observer to be bilaterally symmetric. Exceptions occur quite
broad generalizations that may or may not fit often, however, so bilateral symmetry in these
into any one category of disease. disorders is probably only on the order of
This chapter gives an overview of radiologic 80% to 90%. Rheumatoid arthritis is a com-
evaluation of arthritis with the caveat that it is, mon offender of the bilateral symmetry rule,
by necessity, a simplified version and in no way and one should not be surprised if rheumatoid
complete. If one is interested in greater detail arthritis is seen to be asymmetric in up to 25%
or more accuracy, I would urge reading either of cases.
Debbie Forrester’s excellent monograph1 on the Involvement of joints other than the hand and
subject or Anne Brower’s superb book.2 The wrist is not a common feature with most of the
definitive work on this subject is Don Resnick’s arthritides. In general, when a large joint such as
six-volume tome,3 but most can’t read even the the shoulder, hip, or knee is involved with arthri-
arthritis portion during a 4-year residency—it’s tis only a few entities need to be considered
best used as a reference. (Table 6-2). Although it must be emphasized
The majority of arthritides are most easily that almost any arthritis can affect almost any
examined and categorized by looking at their joint, the diseases listed in Table 6-2 probably
effect on the hands. Forrester recommends a will account for 90% or more of the large joint
search pattern that she calls the ABC’S, with the arthropathia.
A indicating alignment, B standing for bone Involvement of certain joints can often give a
mineralization, C standing for cartilage and in- clue as to the underlying disease process. For
cluding a search for erosions, and the S standing example, if the sacroiliac (SI) joints are involved,
for soft tissues. I would add to this search pattern the differential diagnosis is as listed in Table 6-3.
by making it the ABCD’S, with the D indicating Again, almost any arthritis can affect any joint,
distribution of the pathology. Although this is but if the SI joint is involved, using Table 6-3 for
implied in Forrester’s search pattern, I believe it the differential diagnosis will give a 95% or bet-
cannot be overemphasized. ter chance of having the right answer.
In general, if the distribution of the arthropa- The aforementioned differential diagnoses
thy can be determined, the differential diagnosis are to be considered generalizations and are,
becomes very short (Table 6-1). Although on for the most part (except when mentioned),
paper this sounds quite nice, it can on occasion probably not more than 75% to 85% accurate.
be difficult to accurately determine the distribu- They are a nice starting point, however, for
tion of the arthropathy. The distribution of the developing the differential diagnosis. I cannot
arthropathy is difficult to determine when it is overemphasize that the exceptions are exceed-
not clearly distal or proximal but is more gen- ingly common. There are probably more
eral, such as occurs with gout and sarcoid. It can missed diagnoses in the field of arthritis than
also be difficult to accurately determine the dis- in almost any other area of radiology. The re-
tribution when advanced disease is present, such mainder of this chapter gives a brief overview
as occurs with severe rheumatoid arthritis. With of arthritides with which most radiologists
severe rheumatoid arthritis the proximal nature should be familiar. Rather than provide an in-
of the pathology is not so apparent because of depth description of each process—which can
involvement with the metacarpophalangeal be obtained in any of the major radiology
joints and even the phalangeal joints. In a similar texts—I give salient discriminating points that
manner, when psoriatic disease, Reiter’s syn- might make it easier to differentiate one pro-
drome, or osteoarthritis is severely advanced, it cess from another.
Copyright © 2014 Elsevier, Inc. All rights reserved. 111
112 6 Arthritis
A
FIGURE 6-2 ■ Diffuse idiopathic skeletal hyperostosis.
A lateral film of the lumbar spine shows extensive
osteophytosis without significant disc space narrow-
ing or sclerosis. This is a classic picture for diffuse
idiopathic skeletal hyperostosis (DISH).
FIGURE 6-4 ■ Lack of bilateral symmetry in primary osteoarthritis. This patient has classic radiographic findings
of primary osteoarthritis in the left hand; however, the right hand shows only osteoporosis and soft tissue wasting
without evidence of osteoarthritis. The reason for the lack of bilaterality is that this patient has long-standing
right-sided paralysis, which has blocked the onset of the arthritic changes in the right hand.
HLA-B27 SPONDYLOARTHROPATHIES
A group of diseases that was formerly called
rheumatoid variants is now known as the serone-
gative HLA-B27-positive spondyloarthropa-
thies. What was wrong with rheumatoid variants?
It was short and concise. It has been replaced
FIGURE 6-7 ■ Rheumatoid arthritis. An erosive arthritis that primarily affects the carpal bones and the metacarpo-
phalangeal joints is seen that has osteoporosis and soft tissue swelling (note the soft tissue over the ulnar styloid
processes). It is bilaterally symmetric process in this patient, which is classic.
S
A
Involvement of the SI joints is common in the Reiter’s syndrome and psoriatic arthritis can
HLA-B27 spondyloarthropathies. The patterns exhibit unilateral or bilateral SI joint involve-
of involvement, like the patterns of involvement ment. It is said that it is bilateral about 50% of
of the spine, are somewhat typical for each dis- the time. It is often asymmetric when it is bilat-
order. Ankylosing spondylitis and inflammatory eral, but exact symmetry can be difficult to assess
bowel disease typically cause bilaterally symmet- on plain films. Therefore, when it is definitely
ric SI joint disease that is initially erosive and bilateral and not clearly asymmetric, I consider
progresses to sclerosis and fusion (Figures 6-15 the SI joints to be in the bilateral symmetric
and 6-16). It is extremely unusual to have asym- category. This means that if I have a case with
metric or unilateral SI joint disease in these two bilateral, symmetric SI joint disease, it could be
disorders. Another entity that can have bilateral caused by any of the four HLA-B27 spondyloar-
SI joint erosions is hyperparathyroidism. Sub- thropathies. If I have a case with unilateral (or
periosteal resorption along the SI joints mimics clearly asymmetric) SI joint involvement, I can
erosive changes. This is more commonly seen in confidently exclude ankylosing spondylitis and
children. inflammatory bowel disease, and I would consider
Reiter’s syndrome and psoriatic disease. In this
example I would have to also consider infection
and DJD (don’t forget that DJD can cause ero-
sions in the SI joints) (Figures 6-5 and 6-17) and
in older patients, gout. Computed tomography
(CT) is very helpful in examining the SI joints
and is considered by many to be the diagnostic
procedure of choice because of the unobstructed
view of the entire joint (Figure 6-18).
That is, in a nutshell, my approach to the SI
joints (see Table 6-3). Other considerations in
the differential are too uncommon for me to
worry about for the most part, and you shouldn’t
either.
Large joint involvement with the HLA-B27
spondyloarthropathies is uncommon (except for
FIGURE 6-15 ■ Ankylosing spondylitis. Bilateral, sym-
metric SI joint sclerosis and erosions are seen in this ankylosing spondylitis) but occurs often enough
patient with ankylosing spondylitis. Inflammatory to warrant learning about. In general, the ar-
bowel disease could have an identical appearance. thropathy will resemble rheumatoid arthritis
Although this is classic for these two disorders, it with the typical features thereof (Figure 6-19).
would not be that unusual for this appearance to be
seen with psoriatic disease or Reiter’s syndrome. It
The hips are involved in up to 50% of the
would be unlikely for infection or DJD to be bilateral patients with ankylosing spondylitis.
in this fashion.
Pseudogout (CPPD)
Calcium pyrophosphate crystal deposition dis-
ease (CPPD) causes much confusion among
radiologists as well as other specialists. It is actu-
ally quite simple if you don’t read all the con-
flicting literature. First, what do you call it? Is it
pseudogout or CPPD? Who cares? Call it ei-
ther, or call it both. Many academicians say that
it should be called pseudogout only when symp-
toms are present. Do we call lung cancer some-
FIGURE 6-24 ■ Gout. A sharply marginated erosion with thing else if the patient is asymptomatic? Of
an overhanging edge (arrow) is seen in the metatarso-
phalangeal joint in the great toe in this patient with
course not. For all practical purposes the terms
gout. This appearance and location are classic for pseudogout and CPPD are synonymous, and argu-
gout, whereas psoriasis and Reiter’s syndrome usu- ment over the issue is academic B.S.
ally involve the interphalangeal joint and do not have CPPD has a classic triad: pain, cartilage calci-
erosions that are this sharply marginated. fication, and DJD. The patient may have any
Copyright © 2014 Elsevier, Inc. All rights reserved.
6 Arthritis 123
A B
FIGURE 6-30 ■ Pseudo-Charcot’s joint in CPPD. This patient with CPPD shows severe joint destruction in the carpus
primarily at the radiocarpal joint. A, Large subchondral cysts or geodes are noted. Heterotopic new bone or joint
debris is also seen (arrow). B, Dislocation of the radiocarpal joint is seen, with the entire carpus lying volarly in
relation to the radius. The findings of severe joint destruction, heterotopic new bone, and dislocation are classic
for a Charcot’s joint. This patient, however, had sensation in this joint; therefore it is not a true Charcot’s or neu-
ropathic joint but a pseudo-Charcot’s joint, which is occasionally seen in patients with CPPD.
FIGURE 6-31 ■ Systemic lupus erythematosus. Marked soft tissue wasting, as noted by the concavity in the hypo-
thenar eminence, and ulnar deviation of the phalanges, as is seen primarily in the right hand in this patient,
are hallmarks of SLE.
FIGURE 6-32 ■ Scleroderma. Diffuse subcutaneous soft tissue calcification is seen throughout the hands and wrist
in this patient with scleroderma. Soft tissue wasting and osteoporosis are also present as well as bone loss
in multiple distal phalanges (acro-osteolysis) secondary to the vascular abnormalities are often present in this
disease.
also in the bones. In the skeletal system it has a large osteophytes commonly seen in this disorder.
predilection for the hands, where it causes lytic In fact, the osteophytes are often called droop-
destructive lesions in the cortex. These often ing because of the unusual way they hang off the
have a lacelike appearance (Figure 6-33). Sarcoid joint margin.
can also affect the joints in the hand, causing
DJD-like changes.
NEUROPATHIC OR CHARCOT’S JOINT
HEMOCHROMATOSIS The radiographic findings for a Charcot’s joint
are characteristic and almost pathognomonic. A
From 20% to 50% of patients with hemochro- classic triad has been described that consists of
matosis have a characteristic arthropathy in the joint destruction, dislocation, and heterotopic
hands that should suggest the diagnosis. Hemo- new bone (Figure 6-35). Multiple other findings
chromatosis is a disease of excess iron that gets have been described that do not seem to be as
deposited in tissues throughout the body, lead- useful as the classic triad.
ing to fibrosis and eventual organ failure. The Joint destruction is seen in every arthritis
characteristic arthropathy classically involves the encountered and therefore seems very nonspe-
second through fourth metacarpophalangeal cific; however, nothing causes as severe de-
joints. The radiographic changes are essentially struction in a joint as a Charcot’s joint. Early in
those of DJD (joint space narrowing, sclerosis, the development of a Charcot’s joint, the joint
and osteophytes; Figure 6-34). Up to 50% of destruction may merely appear to be joint space
the patients with hemochromatosis also have narrowing. It is extremely difficult to make
CPPD; therefore, when the hands are being the diagnosis this early. In the spine, instead
looked at, a search should be made for triangular of joint space destruction, there is disc space
fibrocartilage chondrocalcinosis. Another find- destruction (Figure 6-36).
ing that is often seen in hemochromatosis is Dislocation, like joint destruction, can be
called squaring of the metacarpal heads. They present in varying degrees. Early on, the joint
appear enlarged and blocklike as a result of the may have subluxation instead of dislocation.
Copyright © 2014 Elsevier, Inc. All rights reserved.
6 Arthritis 127
Heterotopic new bone has also been termed The shoulder can become a Charcot’s joint in
debris and consists of soft tissue calcification or patients with syringomyelia, which has a so-called
clumps of ossification adjacent to the joint. It, atrophic Charcot’s appearance. This refers to its
too, can be present in varying amounts. tendency to have no debris or heterotopic new
The most commonly seen Charcot’s joint is in bone, and the proximal humerus has a tapered
the foot of a diabetic patient. It typically affects appearance likened to a licked candy stick.
the first and second tarsometatarsal joints in a A pseudo-Charcot’s joint from CPPD (see
fashion termed a Lisfranc’s fracture (Figure 6-37). Figure 6-30) is encountered almost as commonly
Lisfranc was Napoleon’s surgeon, and he gained as a true Charcot’s joint from any other cause
fame for saving the lives of soldiers with gangre- with the exception of the Lisfranc type seen in
nous toes from frostbite by doing a forefoot diabetic patients.
amputation at the tarsometatarsal junction.
Tabes dorsalis from syphilis is seldom seen HEMOPHILIA, JUVENILE
today. I have encountered only two cases of a
Charcot’s joint in syphilis in the past 35 years, RHEUMATOID ARTHRITIS, AND
and I’ve been around some pretty raunchy resi- PARALYSIS
dents. More commonly seen is a Charcot’s joint
in a patient with paralysis who continues to use Why would three clinically disparate entities
the affected limb for support. such as juvenile rheumatoid arthritis (JRA),
Copyright © 2014 Elsevier, Inc. All rights reserved.
128 6 Arthritis
A B
FIGURE 6-46 ■ Pigmented villonodular synovitis. Sagittal T1-weighted (A) and FSE T2-weighted (B) images of an
ankle with PVNS show a soft tissue mass emanating from the ankle joint, which is low signal on both sequences
and has very low-signal hemosiderin lining parts of the synovium. This is characteristic for PVNS.
A radiographic sign in the hip that does work of 5 to 10 mm is no-man’s-land. I usually call it
for indicating an effusion is called the teardrop an effusion if the distance is greater than 5 mm,
sign. Leonard Swischuk first brought this sign to realizing that I’m probably overcalling a few. I
my attention with regard to its application in am also aware that it doesn’t make any difference
pediatric patients. I have used it in adults as well if there is an effusion in the knee or not—the
with good results. The teardrop is an anatomic patient gets treated the same regardless. If it
landmark at the medial aspect of the hip joint were vital to the patient, you could aspirate the
(Figure 6-48) that is made up of several bony joint or obtain an MRI scan to find out. I should
structures bounding the acetabulum medially. emphasize that we never do an MRI examination
The teardrop measurement is the distance from just to see if there is fluid in the joint, because it
the medialmost part of the femoral head to the is completely nonspecific.
medialmost extent of the acetabulum (which is Shoulder effusions are difficult to detect un-
the teardrop). This measurement—inappropri- less they are massive enough to displace the
ately called the teardrop measurement—should humeral head inferiorly, as with a fracture and
be equal in both hips. An effusion will push the hemarthrosis (see Chapter 5). Fortunately, as
femoral head laterally and give the affected side with most other joints, treatment is not based on
a wider teardrop distance (Figure 6-49). The the presence or absence of an effusion, so it
teardrop distance is a valid indicator of an effu- hardly matters. The same is true in the ankle,
sion in children. It is valid in adults only when wrist, and smaller joints. Ultrasound is widely
no long-standing joint abnormality, such as available and is becoming popular for identifying
DJD or an old fracture, is present. A difference joint effusions, but, again, I question why bother
in the teardrop distance from one hip to the —it is unlikely to change the treatment.
other of as little as 1 mm is significant in the
appropriate clinical setting. It would be better
to aspirate a few normal hips rather than risk AVASCULAR NECROSIS
missing a hip infection that might destroy the
hip if diagnosed late. Avascular necrosis (AVN), or aseptic necrosis,
The radiographic sign for a knee effusion that can occur around almost any joint for a host of
seems to be the most reliable is the measurement reasons, including steroid use, trauma, and vari-
of the distance between the suprapatellar fat pad ous underlying disease states and may even occur
and the anterior femoral fat pad (Figure 6-50). A idiopathically. It is often seen in patients who
distance between these two fat pads of more have undergone renal transplant.
than 10 mm is definite evidence of an effusion. A The hallmark of AVN is increased bone density
distance of less than 5 mm is normal. A distance at an otherwise normal joint. Increased density at
a joint usually indicates DJD; however, if osteo-
phytes and joint space narrowing are not present,
Fovea Centralis another disorder should be considered.
The earliest sign of AVN is a joint effusion.
This often is not radiographically visible or is
so nonspecific as to not help with the diagnosis
unless the clinical setting had already raised
suspicion for AVN. The next sign for AVN is a
patchy or mottled density (Figure 6-51). In the
knee this density increase can occur through-
out an entire condyle, whereas in the hip, it
Tear Drop Dist.
often involves the entire femoral head. Next, a
subchondral lucency develops that forms a thin
line along the articular surface (Figure 6-52).
This lucent line has been described as being
an early indicator for AVN when, in fact, it is
a late finding. Also, the lucent line stage is of-
ten not present in the evolution of AVN.
Tear Drop
Therefore using the lucent line as one of the
FIGURE 6-48 ■ Drawing of the teardrop measurement. main criteria for AVN can lead to missing early
The teardrop measurement is the distance from the findings or missing the diagnosis completely.
medialmost aspect of the femoral head to the nearest
portion of the adjacent acetabulum (arrows). Widen- I would estimate that I see a lucent line in
ing of this distance as compared with the opposite hip only 20% or fewer of the cases of AVN in our
is indicative of a joint effusion. hospital.
Copyright © 2014 Elsevier, Inc. All rights reserved.
134 6 Arthritis
FIGURE 6-49 ■ Widened teardrop. The teardrop distance in this patient on the left side (arrows) is slightly wider
than that on the right side (arrows), which is indicative of a hip joint effusion. This patient had a hip joint infection
on the left side.
FIGURE 6-58 ■ Kienböck’s malacia. AVN of the lunate, FIGURE 6-60 ■ Freiberg’s infraction. Flattening, col-
or Kienböck’s malacia, is demonstrated in this patient. lapse, and sclerosis of the second metatarsal head, as
Note the increased density and partial fragmentation seen in this patient, is typical of AVN or Freiberg’s
of the lunate. infraction. It can also involve the third or fourth meta-
tarsal heads. Note the compensatory hypertrophy of
the cortex of the second metatarsal, which is invari-
ably found with this disorder.
REFERENCES
1. Forrester DM, Brown JC: The radiology of joint disease,
ed 3, Philadelphia, 1987, WB Saunders.
2. Brower AC: Arthritis in black and white, ed 2, Philadelphia,
1997, WB Saunders.
3. Resnick D: Diagnosis of bone and joint disorders, ed 4,
Philadelphia, 2002, WB Saunders.
FIGURE 7-1 ■ Mild osteoporosis. Mild cortical narrowing FIGURE 7-3 ■ Normal mineralization. Note the cortical
(arrows) at the mid-second metacarpal is noted in this width (arrows) of the mid-second metacarpal in this
patient with renal osteodystrophy. Compare this cortical patient with normal mineralization. The width of the
width with that of the normal width in Figure 7-3. cortex is easily greater than one third of the total
width of the metacarpal.
FIGURE 7-5 ■ Aggressive osteoporosis. A diffuse per- FIGURE 7-6 ■ Cortical holes in osteoporosis. This patient
meative pattern throughout the proximal femur is suffered a stroke and has aggressive osteoporosis
noted in this patient, who has recently had an amputa- secondary to disuse. What appears to be a diffuse
tion. Note that the cortices are riddled with holes, permeative pattern throughout the humerus is noted,
which would indicate that this is not a true intramed- on closer inspection, to be cortical holes, which in this
ullary process but an intracortical process. This is case resulted from the aggressive osteoporosis. This
distinctive for aggressive osteoporosis and causes a type of pattern, unfortunately, often leads to a biopsy
pseudopermeative pattern that can be mistaken for a to rule out multiple myeloma or other round cell
more sinister process. tumors.
FIGURE 7-14 ■ Hyperparathyroidism. Subperiosteal bone resorption is noted along the medial aspect of the proximal
tibias (arrows). This is pathognomonic for hyperparathyroidism.
A B
C
FIGURE 7-18 ■ PTH discitis. A, A lateral plain film of the lumbar spine reveals disc space narrowing with erosion of
the end-plates at the L2–3 level (arrow). This is a typical appearance of disc infection. B, A sagittal T1-weighted
image of the lumbar spine reveals disc space narrowing and low signal extending into the end-plates and vertebral
bodies adjacent to the L2–3 disc. C, A T2-weighted image shows high signal in the L2–3 disc and in the tissue
extending into the vertebral bodies. This is the classic magnetic resonance imaging appearance of disc infection
with involvement of the vertebral bodies. However, this patient has renal osteodystrophy, and hyperparathyroidism
can cause changes in the joints and the disc spaces identical to those caused by infection. Clinical correlation must
be used to avoid an unnecessary biopsy.
increased in density. This sounds straightfor- Renal Osteodystrophy. Anything that causes
ward enough but is, in fact, often difficult to do. HPT can cause osteosclerosis, but renal disease is
Technical factors can easily alter the apparent far and away the biggest offender. As mentioned
bone density and be misleading. Second, once previously, the sine qua non of renal osteodystro-
it is determined that diffuse osteosclerosis is phy is subperiosteal bone resorption, seen earliest
present, one merely has to list the disease enti- and most reliably at the radial aspect of the middle
ties that could be responsible. This is the easiest phalanges of the hands. Although most patients
step because it merely requires memorization. I with renal osteodystrophy will be osteopenic,
will supply a mnemonic to help your memory. about 10% to 20% will exhibit osteosclerosis, and
Last, one must look for radiographic findings the reasons for it are unknown.
that are specific for each disorder to rule them
out or in so as to narrow down the list of possi- Sickle Cell Disease. Like renal osteodystrophy,
bilities. The list of diseases that cause diffuse the reason for dense bones to occur in sickle cell
osteosclerosis is different with each textbook that disease is unknown. It occurs in only a small per-
you read. Many disorders have been reported centage of patients. Additional signs to look for are
to cause osteosclerosis, but you only need a list bone infarcts and H-shaped or step-off deformities
that is correct 95% to 98% of the time. Nobody of the vertebral body end-plates (Figure 7-19).
expects you to include the one reported case of These are also called fish vertebrae because of
humpback midget whale syndrome in your list. If their similarity to the vertebrae found in fish.
you absolutely cannot bear the thought of leaving
out a possibility in your differential diagnosis, you Myelofibrosis. Also called agnogenic myeloid
might as well just give your clinician the index metaplasia, myelofibrosis is a disease caused by
from Resnick’s book—it will be all-inclusive but progressive fibrosis of the marrow in patients
not really useful to the clinician (not to imply that older than 50 years of age. It leads to anemia
Resnick’s book is not useful—it is the Bible for
bone radiologists).
The entities I include in the differential for
diffuse osteosclerosis are the following:
• Renal osteodystrophy
• Sickle cell disease
• Myelofibrosis
• Osteopetrosis
• Pyknodysostosis
• Metastatic carcinoma
• Mastocytosis
• Paget’s disease
• Athletes
• Fluorosis
The mnemonic I use to remember them is
“Regular sex makes occasional perversions much
more pleasurable and fantastic.” Hey, it’s not a
great mnemonic, but in the twenty-first century
it’s not politically correct to be ribald, vulgar,
off-color, coarse, crude, erotic, lewd, or even,
sometimes, funny. If you want a filthy mnemonic,
make up your own, you insensitive pervert.
I will cover each of these topics in generali-
ties, trying to point out the features of each that
you should look for when deciding whether to
include or exclude them from the differential.
A nice feature of this mnemonic is it lists the
entities in their order of frequency. Not that FIGURE 7-19 ■ Sickle cell disease. Step-off deformities
osteopetrosis is more common than Paget’s are seen in the end-plates of multiple vertebral bodies
disease or metastatic carcinoma, but it’s more in this patient with sickle cell disease. Although the
bones do not show osteosclerosis, the visible
common to see osteopetrosis than either Paget’s trabeculae are somewhat coarsened. The step-off
disease or metastatic carcinoma presenting as deformities in the spine are characteristic for sickle cell
diffuse bony sclerosis. disease. These are also called fish vertebrae.
FIGURE 7-23 ■ Osteopetrosis. “Sandwich vertebrae” FIGURE 7-24 ■ Pyknodysostosis. Dense sclerosis is seen
are seen in the vertebral bodies in this patient with throughout the hand in this patient with pyknodysos-
osteopetrosis. This is virtually pathognomonic for tosis. A pathognomonic finding is seen in the distal
osteopetrosis when present and should not be con- phalanges, where the tufts are absent and the phalanges
fused with the ill-defined bands of sclerosis seen in are pointed and sclerotic.
the rugger jersey spine of hyperparathyroidism.
Three distinct phases are radiographically (Figure 7-30), whereas entities that cause osteo-
visible in Paget’s disease: a lytic phase, a scle- sclerosis as a result of marrow disease (sickle cell
rotic phase, and a mixed lytic-sclerotic phase disease and myelofibrosis) will spare the base of
(Figure 7-29). The lytic phase often has a sharp the skull because it contains no marrow. This
leading edge, called a flame-shaped or blade of point is best illustrated by thalassemia in which
grass leading edge. In a long bone, with the sole the calvarium is thickened by marrow hyperplasia
exception being the tibia, Paget’s disease always while the base of the skull is spared (Figure 7-31).
starts at the end of the bone; therefore, if a
lesion is present in the middle of a long bone Athletes. I see radiographs of professional ath-
and does not extend to either end, you can letes quite often and continue to be impressed by
safely exclude Paget’s disease. the degree of increased cortical thickness these
So why is Paget’s disease in this differential if people possess. No question about it: increased
it so rarely fits? Good question. You can proba- stress causes hypertrophy of bone as well as of
bly safely leave it out without missing too many muscle. I see radiologists routinely question the
cases, but it’s good to at least consider, because it
is so readily diagnosed once you do think of it.
Also, it does occasionally involve a large area,
such as the entire pelvis, and makes the observer
think that the entire skeleton might be involved.
When Paget’s disease involves the skull, it
often causes thickening of the base of the skull
Miscellaneous Conditions
There are a host of bony conditions, diseases, resonance imaging (MRI) is extremely valuable
and syndromes that do not fit conveniently into in demonstrating the presence and extent of
any of the preceding chapters yet should be AVN (Figure 8-2, B), even when plain films are
given some mention in an attempted overview of apparently normal. MRI is currently considered
musculoskeletal radiology. Many of these are the most efficacious way to evaluate a joint for
simply “Aunt Minnies” and only require you to AVN.2 It is useful in AVN not only of the hips
have seen them once or twice to recognize them. but also of the knee, wrist, foot, and ankle.
I have severely limited the things I have included
in this chapter—it could easily have dozens
of other entities, but none are very common. ENGELMANN’S DISEASE
Besides, I need to have something to add in
future editions. These are listed alphabetically Also known as diaphyseal dysplasia, Engelmann’s
for lack of a more scientific basis. disease is a congenital disorder manifested by
diaphyseal cortical thickening, which primarily
involves the long bones, particularly the lower
ACHONDROPLASIA extremity (Figure 8-5). Although it can be asymp-
tomatic and incidental, children with Engelmann’s
The most common cause of dwarfism is achon- disease can have a painful, waddling gait, and
droplasia, a congenital, hereditary disease of fail- the disease can progress dramatically, causing
ure of endochondral bone formation. The femurs severe medullary encroachment with subsequent
and humeri are more profoundly affected than anemia.
the other long bones, although the entire skeleton
is abnormal. The spine typically has narrowing of
the interpedicular distances in a caudal direction HYPERTROPHIC PULMONARY
(Figure 8-1), the opposite of normal, where the OSTEOARTHROPATHY
interpedicular distances get progressively wider as
one proceeds down the spine. I know of no other Hypertrophic pulmonary osteoarthropathy
entity that has narrowing of the interpedicular (HPO) is manifested by clubbing of the fingers
distance, and I have never seen a case of achon- and periostitis, usually in the upper and lower
droplasia without this narrowing. The long bones extremities (Figure 8-6), which may or may not
are short but have normal width, giving them a be associated with bone pain. It is most com-
thick appearance. monly seen in patients with lung cancer, but many
other etiologies have been reported, including
bronchiectasis, gastrointestinal disorders, and
AVASCULAR NECROSIS liver disease. The actual mechanism of formation
of periostitis secondary to a distant malignancy
The term avascular necrosis (AVN) refers to the or other process is unknown. The differential
lack of blood supply with subsequent bone death diagnosis for periostitis in a long bone without
and ensuing bony collapse in an articular sur- an underlying bony abnormality includes HPO,
face. This condition is also called osteonecrosis. venous stasis, thyroid acropachy, pachydermo-
The etiology of AVN is an extensive differential periostosis, and trauma (Box 8-2).
that most commonly includes trauma, steroids,
aspirin, sickle cell disease, collagen vascular dis-
eases, alcoholism, and idiopathic conditions MELORHEOSTOSIS
(Box 8-1).1 The radiographic appearance ranges
from patchy sclerosis (Figure 8-2, A) to articular Melorheostosis is a rare, idiopathic disorder
surface collapse and fragmentation (Figure 8-3). characterized by thickened cortical new bone
Just before collapse, a subchondral lucency is that accumulates near the ends of long bones,
occasionally seen (Figure 8-4); however, this is usually only on one side of the bone, and has
a late and inconstant sign of AVN. Magnetic an appearance likened to dripping candle wax
Copyright © 2014 Elsevier, Inc. All rights reserved. 153
154 8 Miscellaneous Conditions
MUCOPOLYSACCHARIDOSES
(MORQUIO’S, HURLER’S, AND
HUNTER’S SYNDROMES)
The mucopolysaccharidoses are a group of inher-
ited diseases characterized by abnormal storage
and excretion in the urine of various mucopoly-
saccharides, such as keratan sulfate (Morquio’s)
and heparan sulfate (Hurler’s). Affected patients
have short stature, primarily from shortened
spines, and characteristic plain film findings. In
the spine, patients with Morquio’s syndrome have
platyspondyly (generalized flattening of the verte-
bral bodies) with a central anterior projection or
FIGURE 8-1 ■ Achondroplasia. An anteroposterior (AP) “beak” off of the vertebral body as viewed on a
plain film of the spine in this patient with achondroplasia lateral plain film (Figure 8-8). In patients with
demonstrates narrowing of the interpedicular distance Hurler’s or Hunter’s syndrome, platyspondyly is
(arrows) in a caudal direction, which is characteristic
of this disorder. Ordinarily the interpedicular distance
seen with a “beak” that is anteroinferiorly posi-
widens in each vertebra in a caudal direction. tioned (Figure 8-9) (remember “middle” beak for
Morquio’s syndrome and “hind” beak for Hurler’s
and Hunter’s syndromes). The pelvis in these
BOX 8-1 Common Causes of disorders is similar in appearance to that seen
Avascular Necrosis in patients with achondroplasia with wide, flared
iliac wings and broad femoral necks. A character-
Trauma Collagen vascular diseases istic finding in the hands is a pointed proximal
Steroids Alcoholism fifth metacarpal base that has a notch appearance
Aspirin Idiopathic causes on the ulnar aspect (Figure 8-10).
Sickle cell disease
OSTEOID OSTEOMA
The etiology of osteoid osteoma is unknown. Is
it an infection (bacterial or viral), a slow-growing
tumor, a dessert topping? Nobody knows. It is a
painful lesion that occurs almost exclusively in
patients younger than 30 years of age that is
treated successfully with surgical excision or,
more commonly today, percutaneous radiofre-
quency ablation. Aspirin often gives dramatic
relief of the pain and can be used for conserva-
tive treatment in lieu of surgery. A classic clinical
picture for an osteoid osteoma is night pain
relieved by aspirin. However, many osteoid
osteomas do not have this presentation, and
most painful musculoskeletal lesions are worse
at night and relieved by aspirin, so this can lead
to confusion.
Radiographically, an osteoid osteoma is said
to have a typical appearance, but, in fact, it has
many different appearances, which can make
FIGURE 8-11 ■ Multiple hereditary exostosis. The knees diagnosis difficult.3 The classically described
are involved in virtually every case of multiple heredi- radiographic appearance is a cortically based
tary exostosis. They typically show not only multiple
exostoses (arrows) but also marked undertubulation sclerotic lesion in a long bone that has a small
(widening) in the metaphyses. lucency within it, called the nidus (Figure 8-13, A).
It is the nidus that causes the pain and the sur-
rounding reactive sclerosis. If the nidus is surgi-
cally removed or ablated percutaneously with
radiofrequency, complete cessation of pain is the
rule. Computed tomography (CT) and radionu-
clide bone scanning are often very helpful in
Copyright © 2014 Elsevier, Inc. All rights reserved.
158 8 Miscellaneous Conditions
A B
C
FIGURE 8-13 ■ Osteoid osteoma. A, An AP plain film of the femur in a child with hip pain shows an area of sclero-
sis medially near the lesser trochanter with a small lucency (arrow), which is the nidus of an osteoid osteoma.
Osteomyelitis could have this identical appearance. B, A radionuclide bone scan shows increased uptake in the
proximal femur, which corresponds to the reactive new bone seen on the plain film. In addition, however, note
the second smaller area of increased radionuclide uptake within the larger area (arrow). This corresponds to the
nidus itself. This pattern on a bone scan is called the double-density sign. C, A CT scan of the femur shows the
sclerosis medially and the lucent nidus (arrow) to better advantage. The CT and the bone scan give the surgeon
a more precise anatomic location of the nidus than the plain film.
demonstrating the exact location of the nidus as sclerotic. Up to 80% of osteoid osteomas are
(Figure 8-13, B and C). located intracortically, with the remainder being
If the nidus of an osteoid osteoma is located in in the intramedullary part of a bone. Rarely, an
the medullary rather than the cortical portion of osteoid osteoma will present in the periosteum,
a bone or if it is located in a joint, there is much causing tremendous periostitis.
less reactive sclerosis present. This gives the le- The nidus itself is usually lucent but often
sion a different overall appearance than the more develops some calcification within it. It then has
common cortical lesion in that it does not appear the appearance of a sequestrum as is seen in
Copyright © 2014 Elsevier, Inc. All rights reserved.
8 Miscellaneous Conditions 159
A B
C
FIGURE 8-14 ■ Osteoid osteoma. A, A lateral plain film of the tibia in this child with leg pain shows cortical thickening
in the posterior diaphysis. No lucency in the sclerotic area could be identified. B, A radionuclide bone scan reveals
uptake corresponding to the area of sclerosis in the tibia, with a more marked area of uptake centrally (arrow), which
is the double-density sign of an osteoid osteoma. C, The surgical specimen shows the nidus (arrow) as a faint
lucency within the sclerotic bone.
A B
A B
REFERENCES
1. Mankin H: Nontraumatic necrosis of bone (osteonecrosis). 4. Helms CA, Hattner RS, Vogler J III: Osteoid osteoma:
N Engl J Med 326:1473–1479, 1992. radionuclide diagnosis. Radiology 151(3):779–784, 1984.
2. Mitchell D, Kressel H, Arger P, et al: Avascular necrosis of 5. Korompilias AV, Karantanas AH, Lykissas MG, Beris AE:
the femoral head: morphologic assessment by MR imaging, Bone marrow edema syndrome. Skeletal Radiol 38:425-436,
with CT correlation. Radiology 161:739–742, 1986. 2009.
3. Marcove R, Heelan R, Huvos A, et al: Osteoid osteoma:
diagnosis, localization, and treatment. Clin Orthop Relat
Res 267:197–201, 1991.
out a meniscocapsular separation. This was the Many acceptable variations of this protocol exist.
second or third time that week that the resident Many centers, for various reasons, prefer not to
had been so instructed, so he asked, “Why don’t use FSE images and instead use gradient-echo.
we just routinely do a T2 coronal—we don’t see
anything on the T1 coronal, and we often have
to hedge or repeat the examination with a T2 MENISCI
coronal?” I told him to worry about dictating the
cases, and I and others would take care of setting The normal meniscus is a fibrocartilaginous,
up the protocols. Well, 2 weeks later he rotated C-shaped structure that is uniformly low in
to another service, and we began doing T2 coro- signal on both T1- and T2-weighted sequences
nal images as part of our standard knee protocol. (Figure 9-1). With T2* sequences the menisci
Uppity residents! will usually demonstrate some internal signal.
Axial images were initially used by the tech- With T1-weighted images any signal within
nicians as a scout view. They were then found to the meniscus is abnormal, except in children,
be useful for viewing the patellofemoral carti- in whom some signal is normal and represents
lage and for identifying and characterizing fluid normal vascularity. Meniscal signal that does
collections. As in the coronal images, to afford not disrupt an articular surface is representa-
an opportunity to see any pathology, T2 images tive of intrasubstance degeneration (Figure 9-2),
must be obtained. which is myxoid degeneration of the fibrocar-
We have found that fat suppressing the sagittal tilage. It most likely represents aging and
T1 or proton-density meniscus-sensitive sequence normal wear and tear. It is not believed to be
increases the dynamic range of signal in the symptomatic and cannot be diagnosed clinically
meniscus and makes meniscal pathology more or with arthroscopy. Some choose, therefore,
conspicuous. It gets rid of all the distracting high not to mention intrasubstance degeneration
signal in the marrow, making it easier to visualize in the radiology interpretation. A grading scale
the meniscus. for meniscal signal that is widely used is the
The use of FSE sequences with a short TE following (Figure 9-3): grade 1—rounded or
(FSE proton-density) has been reported in sev- amorphous signal that does not disrupt an
eral studies to be useful for meniscal tears, yet articular surface; grade 2—linear signal that
others have reported a decreased sensitivity. does not disrupt an articular surface; and grade
What is the truth? The truth is that every pub- 3—rounded or linear signal that disrupts an
lished report I can find shows a sensitivity of articular surface (Figure 9-4). Grades 1 and 2
around 80% for FSE proton-density sequences, are intrasubstance degeneration and should
whereas conventional spin-echo sequences have not be reported as “grade 1 or 2 tears.” This
a sensitivity closer to 95%. It is very controver- is a radiology grading scale, which is not
sial, but basically everyone’s results are the same known widely by orthopedic surgeons; hence,
(80% sensitivity), only the conclusions differ.4 the term tear often leads to an unnecessary
If you’re willing to decrease your sensitivity for arthroscopy (arthroscopy is not indicated for
meniscal tears from 95% to 80% to save 3 minutes intrasubstance degeneration). Only grade 3 is
(the only advantage of FSE), then your time is a meniscal tear.
more valuable than your diagnostic accuracy. Get When high signal in a meniscus disrupts the
another job! superior or inferior articular surface, a meniscal
The protocol I currently recommend consists tear is diagnosed (see Figure 9-4). Care must be
of a sagittal proton-density weighted spin-echo taken to be sure that the signal actually disrupts
series with fat suppression and sagittal, coronal, the articular surface of the meniscus before call-
and axial FSE T2 with fat suppression (Table 9-1). ing a tear. When high signal approaches the
Fat Sat, Fat suppression; FOV, field of view; Nex, number of excitations.
A B
FIGURE 9-1 ■ Normal meniscus. A, A T1-weighted sagittal image through a normal lateral meniscus demonstrates
uniform low signal in the meniscus. This is a section through the body of the meniscus, because it has a bowtie
configuration. Two sections of the body should be seen in each meniscus with 4- or 5-mm thick slices. B, In the
same T1-weighted sequence, this sagittal image demonstrates uniform low signal in the anterior and posterior
horns of this normal lateral meniscus. (Anterior is to the left.)
B
FIGURE 9-2 ■ Intrasubstance degeneration. A, Faint
intermediate signal can be seen in the posterior horn
of this meniscus (arrow), which does not disrupt the
articular surface of the meniscus. This is intrasub-
stance degeneration. B, Linear high signal is present
in the posterior horn of the meniscus (arrow). The
signal does not disrupt the articular surface; therefore
this represents intrasubstance degeneration.
B
FIGURE 9-8 ■ Displaced fragment in bucket handle tear. FIGURE 9-10 ■ Radial tear. Sagittal images through the
(A) Sagittal and (B) coronal T2-weighted images body of the medial meniscus in a patient with a radial
through the intercondylar notch in a patient with a tear showed a normal bowtie configuration in the
bucket handle tear reveals the displaced free fragment most medial image, with the next adjacent image
or “handle” (arrows) just anterior to the posterior (shown here) having a small gap (arrow), which indi-
cruciate ligament. cates a disruption of the free edge of the meniscus.
tear and, as the next section describes, the sign wedged, flat, and others. It is unknown whether it
can be used to diagnose a discoid meniscus. is congenital or acquired, but most are found in
There are four pitfalls to be aware of in applying children or young adults. It is seen laterally in up
the bowtie sign (Box 9-1). First, if the knee and to 3% of the population, with a discoid medial
the menisci are very small, as in a child’s knee, meniscus being rare. A discoid meniscus is be-
only one bowtie may be observed without a lieved to be more prone to tear than a normal
bucket handle tear being present. However, meniscus and can be symptomatic even without
there will be only two or three sagittal images being torn. Although they are easily identified on
that demonstrate the anterior and posterior coronal images by noting meniscal tissue extend-
horns. A normal-sized knee will have two bow- ing into the tibial spines at the intercondylar notch
ties and three or four images that show the ante- (Figure 9-11), they are most reliably diagnosed by
rior and posterior horns. Also, in a small knee noting more than two consecutive sagittal images
both the medial and lateral menisci will have that show the meniscus with a bowtie appearance
only one bowtie image, and bucket handle tears (Figure 9-12). Hence, the bowtie sign can be used
involving both the medial and lateral menisci are to diagnose a bucket handle tear (fewer than two
very rare. bowties) or a discoid meniscus (more than two
The second pitfall in the bowtie sign is seen in bowties). If slices thinner than 4 or 5 mm are used,
older patients—those older than age 60. Patients the bowtie sign can be adjusted to whatever slice
older than 60 years often have worn down the thickness is used.
inner free edges of their menisci so only one The lateral meniscus often has what appears
sagittal image of the body is seen, followed by to be a tear on the anterior horn near its upper
four or five images that show the anterior and margin, which is a pseudotear from the insertion
posterior horns, usually a good sign for a bucket of the transverse ligament (Figure 9-13). This
handle tear. This, unlike the pitfall described can easily be differentiated from a real tear by
in children or small knees, does not necessarily following it medially across the knee in Hoffa’s
occur in both menisci. How do I differentiate fat pad to where it inserts onto the anterior horn
this from a real bucket handle tear? First, degen- of the medial meniscus. Although less common
erative disc disease is always present, and, second, than on the lateral meniscus, a pseudotear from
no displaced meniscal fragment can be found. the insertion of the transverse ligament onto
A third pitfall to be aware of in using the bow- the anterior horn of the medial meniscus can
tie sign is that the sign works only if the entire be seen.
meniscus is covered with sagittal images. If the Meniscal cysts occur in about 5% of cases
technologist doesn’t begin the sagittal images at and can cause pain even if the meniscus is not
the far medial or lateral aspects of the knee, the torn. The cause is unknown, but they occur
meniscus will not be imaged in its entirety. One more frequently in discoid menisci. If the me-
will quickly learn to appreciate whether or not niscus is not torn, the surgical approach used by
the entire meniscus is covered. some is percutaneous with decompression and
The fourth pitfall is seen in knees that have
been operated on previously and in which a partial
meniscectomy has been performed. This can be
differentiated from a bucket handle tear by the
lack of a displaced fragment (no handle of the
bucket). In fact, all four of the pitfalls mentioned
can be recognized by the inability to see a dis-
placed meniscal fragment.
A discoid meniscus is a large disclike meniscus
that can have many different shapes—lens-shaped,
A B
C
FIGURE 9-12 ■ Discoid lateral meniscus. Three consecutive 5-mm thick T2-weighted images through the lateral
meniscus, beginning with the most lateral (A) and extending medially (B and C); each shows the meniscus to have
a bowtie configuration. Because only two images should have a bowtie shape, indicative of the body of the
meniscus, this is diagnostic of a discoid lateral meniscus (Figure 9-11 is a coronal image of the same knee).
(Anterior is to the left.)
packing performed, whereas if a meniscus tear (called a parameniscal cyst) or into the joint via
is associated with the cyst, it is approached in- a meniscus tear. Decompression into a parame-
traarticularly. Hence, accurate diagnosis of a niscal cyst does not indicate a meniscus tear.
tear is imperative. The intrameniscal portion of A meniscus tear, by definition, has to disrupt the
the cyst typically does not get fluid-bright in articular surface of the meniscus. Although a
signal on T2 sequences (Figure 9-14), which has meniscus with a meniscal cyst is more likely to
misled many radiologists into discounting the tear than an otherwise normal meniscus, up to
presence of a cyst. A meniscal cyst will enlarge 40% are not torn.7 Many reports cite that me-
the meniscus and give it a swollen appearance nisci with cysts are torn up to 98% of the time;
unless it decompresses into the soft tissues this is simply not true.
Copyright © 2014 Elsevier, Inc. All rights reserved.
172 9 Magnetic Resonance Imaging of the Knee
CRUCIATE LIGAMENTS
MRI scanning of the cruciate ligaments is more
accurate than is an MRI scan of the menisci,
with accuracy reported near 100% in several
published series.8 The normal anterior cruciate
ligament (ACL) is seen in the intercondylar
notch as a linear, predominantly low-signal struc-
ture on T1-weighted or proton-density-weighted
images that often shows some linear striations
near its insertion onto the medial tibial spine B
when viewed on sagittal images (Figure 9-15, A).
T2-weighted images are imperative for obtain- FIGURE 9-14 ■ Meniscal cyst. A, A sagittal proton-density-
ing the highest accuracy in diagnosing ACL weighted image through the medial meniscus shows
tears, because fluid and hemorrhage will often a swollen anterior horn filled with increased signal
(arrow). B, A T2-weighted image shows high signal
obscure the ligament on T1-weighted images similar to fluid in the parameniscal portion, whereas
(Figure 9-15, B). the intrameniscal signal is only intermediate.
A torn ACL is most often simply not visual-
ized (Figure 9-16), although sometimes the ac-
tual disruption will be seen (Figure 9-17). Partial report. It is an incidental finding with little or no
tears or sprains of the ACL are manifested by clinical significance. It is found in about 1% of
high signal and/or laxity in an otherwise intact all knees. Even though it often gets misdiag-
ligament. The diagnosis of a partial tear or sprain nosed as an ACL tear on MRI examination, it
is generally not critical on MRI because the virtually never confuses a surgeon into operating
treatment afforded the patient depends primar- because the knee is stable.
ily on the diagnosis of a complete tear. In most The normal posterior cruciate ligament
instances the arthroscopist cannot tell a partial (PCL) is a gently curved, uniformly low-signal
tear from an intact ACL. structure (Figure 9-19), which is infrequently
An entity that has been mistaken for a torn torn and even less frequently repaired by sur-
ACL is an ACL cyst (Figure 9-18). The cause is geons. When torn, it typically has a thickened
unknown, but it basically represents fluid in and appearance with diffuse intermediate signal
around the fibers of the ACL, giving it a swollen, throughout (Figure 9-20). It typically appears
drumstick appearance on sagittal images. It has intact but has undergone a plastic deformation
been mistakenly diagnosed as a tumor with sub- that renders it unstable, much like overstretch-
sequent radical removal (oops!) in one published ing the elastic in your socks. In only a third of
Copyright © 2014 Elsevier, Inc. All rights reserved.
9 Magnetic Resonance Imaging of the Knee 173
A B
FIGURE 9-15 ■ Normal anterior cruciate ligament. A, A sagittal T1-weighted image through the intercondylar notch
shows the normal appearance of the ACL (arrows). B, A sagittal T2-weighted image with fat suppression through
the intercondylar notch in another patient shows a normal ACL.
Ligament of
Wrisberg
Ligament of
Humphry
FIGURE 9-22 ■ Ligament of Wrisberg. This coronal T1-
weighted image shows an obliquely oriented struc-
ture extending from the medial femoral condyle to Posterior Cruciate
the lateral meniscus. This is a normal ligament of B Ligament
Wrisberg. FIGURE 9-23 ■ Ligament of Wrisberg. A, A sagittal T1-
weighted image through the intercondylar notch shows
a rounded low-signal structure posterior to the PCL,
which is the meniscofemoral ligament of Wrisberg
(arrow). B, This drawing shows the relationship of the
ligaments of Wrisberg and Humphry to the PCL.
! !
A B
FIGURE 9-24 ■ Pseudotear from ligament of Humphry insertion. A, A sagittal proton-density fat-suppressed image
through the lateral meniscus reveals an apparent tear of the posterior horn (arrow), which is the insertion of the
ligament of Humphry onto the meniscus. B, On the image through the intercondylar notch, a ligament of
Humphry (arrow) is seen anterior to the PCL. The ligament of Humphry could be followed on adjacent images,
from anterior to the PCL to its insertion on the posterior horn of the lateral meniscus.
COLLATERAL LIGAMENTS
The medial collateral ligament (MCL) origi-
nates on the medial femoral condyle and inserts
on the tibia. It is closely applied to the joint and
is intimately associated with the medial joint
capsule and the medial meniscus. The MCL is
uniformly low in signal on all imaging sequences. FIGURE 9-25 ■ Grade 2 sprain of the medial collateral
Injuries to the MCL usually occur from a valgus ligament. A T2-weighted coronal image reveals high
stress, a blow to the lateral part of the knee. signal in the soft tissues adjacent to the MCL (arrows),
A grade 1 injury represents a mild sprain and which represents edema and hemorrhage from a
is diagnosed on MRI by noting fluid or hemor- sprain of the MCL. The MCL is somewhat attenuated
(area between the arrows), indicative of a grade 2
rhage in the soft tissues medial to the MCL. The sprain.
ligament is otherwise normal. A grade 2 injury
is a partial tear and is seen as high signal in
and around the MCL on T2 coronal sequences. recognized on a T2 coronal image by noting
The ligament is intact, although the deep or joint fluid extending between the medial menis-
superficial fibers may show minimal disruption cus and the capsule. It is essential to use T2
(Figure 9-25). A grade 3 injury is a complete sequences because a T1-weighted image may
disruption of the MCL. It can be best appreci- not detect the fluid between the meniscus and
ated on T2 images (Figure 9-26). It is unusual the capsule (Figure 9-27). They can be over-
for a surgeon to operate on an MCL even if it is looked at arthroscopy if they involve only the
a complete disruption. MCL partial tears, and superficial fibers of the capsule because they are
even complete tears, heal quite nicely simply then essentially extracapsular. It is an important
with immobilization. diagnosis to make because it involves a very
A meniscocapsular separation occurs when vascular portion of the meniscus; hence, it will
the medial meniscus is torn from its attachment readily heal with immobilization or with sutur-
to the joint capsule. It occurs most commonly ing by the surgeon. If overlooked and continued
at the site of the MCL and often occurs con- activity occurs, it can lose the vascular interface
comitantly with an MCL injury. It is easily and never heal.
Copyright © 2014 Elsevier, Inc. All rights reserved.
9 Magnetic Resonance Imaging of the Knee 177
A B
A B
FIGURE 9-34 ■ Semimembranosus tibial collateral ligament bursa. A, A sagittal FSE T2-weighted image with fat
suppression through the medial aspect of the knee shows a fluid collection (arrows) at the joint line that is
adjacent to the posterior horn of the medial meniscus. This is characteristic of a semimembranosus MCL bursa.
B, A coronal FSE T2-weighted image with fat suppression shows this bursa at the joint line with a comma-
shaped appearance.
REFERENCES
1. Ruwe P, Wright J, Randall R, et al: Can MR imaging 7. Anderson J, Helms CA, Conner G: New observations
effectively replace diagnostic arthroscopy? Radiology on meniscal cysts. Skeletal Radiol 39:1187–1191,
183:335–339, 1992. 2010.
2. Crues JI, Mink J, Levy T, et al: Meniscal tears of the knee: 8. Lee J, Yao L, Phelps C, et al: Anterior cruciate liga-
accuracy of MR imaging. Radiology 164:445–448, 1987. ment tears: MR imaging compared with arthroscopy
3. Mink JH, Deutsch AL: Magnetic resonance imaging of and clinical tests. Radiology 166:861–864, 1988.
the knee. Clin Orthop Relat Res 244:29–47, 1989. 9. Rodriquez W Jr, Vinson EN, Helms CA, Toth AP:
4. Helms CA: The meniscus: recent advances in MR MRI appearance of PCL tears. AJR Am J Roentgenol
imaging of the knee. AJR Am J Roentgenol 179(5): 191:W155–159, 2008.
1115–1122, 2002. 10. Covey DC: Injuries of the posterolateral corner of the
5. De Smet AM, Norris MA, Yandow DR, et al. MR diag- knee. J Bone Joint Surg Am 83-A:106–118, 2001.
nosis of meniscal tears of the knee: importance of high 11. Murphy B, Smith R, Uribe J, et al: Bone signal ab-
signal in the meniscus that extends to the surface. AJR Am normalities in the posterolateral tibia and lateral
J Roentgenol 161:101–107, 1993. femoral condyle in complete tears of the anterior
6. Helms CA, Laorr A, Cannon WD Jr: The absent bow tie cruciate ligament: a specific sign? Radiology
sign in bucket-handle tears of the menisci in the knee. 182:221–224, 1992.
AJR Am J Roentgenol 170(1):57–61, 1998.
A B
FIGURE 10-3 ■ Internal rotation hiding partial tear of the supraspinatus tendon. A, An oblique coronal FSE T2-
weighted image shows an apparently normal supraspinatus tendon inserting onto the greater tuberosity (arrow).
B, One slice anteriorly, the bicipital groove can be identified with the anterior fibers of the supraspinatus tendon
just lateral to the groove lifted off of the greater tuberosity (arrow). This is a partial tear of the rotator cuff at its
anterior-most portion.
information to the oblique coronal FSE T2- T2 weighting.6 Some tendon degeneration (tendi-
weighted sequence. nopathy) can be seen in asymptomatic shoulders
If signal in the critical zone is brighter on the in patients of all ages; hence, it needs to be corre-
T2-weighted images, it is abnormal and repre- lated with the clinical picture. If the signal gets
sents a partial tear if it is fluid bright. A partial tear brighter on T2-weighted images, it must be con-
can also be present if the cuff has focal thinning of sidered pathologic—a partial tear. If intermediate
the tendon (Figure 10-4). signal in the cuff tendons is accompanied by fusi-
Myxoid or fibrillar degeneration of the supra- form or focal thickening, myxoid degeneration is
spinatus tendon is commonly found in autopsy present (Figure 10-5). Surgeons will debride this
specimens and increases with age. The majority when it is prominent.4
of asymptomatic shoulders in patients older than
age 50 are believed to have some tendon degen-
eration in the supraspinatus tendon, which has
been termed tendinopathy. This is seen as inter-
mediate to high signal in the critical zone on
T1-weighted images that does not increase with
FIGURE 10-4 ■ Partial tear of the supraspinatus tendon. FIGURE 10-5 ■ Tendinosis. An oblique coronal FSE T2-
An oblique coronal FSE T2-weighted image with fat weighted image shows intermediate signal in a supra-
suppression shows thinning of the supraspinatus ten- spinatus tendon (arrow) that has fusiform swelling.
don (arrow), which is a partial articular-sided cuff tear. This is myxoid degeneration or tendinosis.
If disruption of the supraspinatus tendon can impingement syndrome. In the proper clinical
be seen, obviously a full-thickness tear is present setting an anterior acromioplasty will relieve the
(Figure 10-6). In these cases fluid is invariably symptoms of impingement syndrome and prevent
present in the subacromial bursa. Care should be a more serious full-thickness cuff tear. Many be-
taken to look for retraction of the supraspinatus lieve it is imperative that the surgeon also remove
muscle, because marked retraction will obviate any AC joint undersurface irregularity, if present,
some types of surgery. or a failed surgery can be expected. This is some-
what controversial, with many surgeons subscrib-
ing to the intrinsic degeneration theory for cuff
Partial Tears pathology, while others claim impingement from
Partial cuff tears have marked clinical signifi- bony abnormalities such as spurs being the pri-
cance because most agree that they will not heal mary etiologic agent for cuff disease. Those
on their own if they are greater than 25% of the surgeons who believe intrinsic degeneration is
cuff thickness.7 Although we generally can’t be the primary problem do not remove bony spurs;
so precise as to what percentage of the cuff is they debride the abnormal cuff tissue.4
involved, we can usually identify partial cuff The acromion has been classified into three
tears. If there is an irregularity or thinning of the types as seen on the sagittal images. The normal
cuff on either the bursal side of the cuff or on the appearance, type 1, is an acromion with a flat or
joint side, I will describe it as small, medium, or slightly convex undersurface. A type 2 acromion
large (near full thickness). Bursal-sided partial has a concave undersurface, and a type 3 acro-
tears are exceedingly uncommon, 25 to 30 times mion has a concave undersurface with an ante-
less frequent than articular-sided partial tears. rior osteophyte. Although it has been reported
The most common cuff tear encountered is an that a type 3 acromion is seen in up to 80% of
articular-sided partial tear called a rim rent. It cuff tears, a type 2 acromion is seen in 20%, and
occurs at the insertion of the fibers of the cuff a type 1 acromion has no association with a torn
onto the greater tuberosity, most commonly cuff, this has not been substantiated and is cer-
anteriorly at the insertion of the supraspinatus tainly not my experience. I often see cuff tears
tendon (Figure 10-7). Rim rent tears comprise with a type 1 acromion, and I often see no cuff
20% to 40% of all cuff tears.5 tear when a type 3 acromion is present. I do not
routinely report the type of acromion, because
our surgeons don’t use the information for any-
Bony Abnormalities thing. More importantly, I try to tell them if
The undersurface of the anterior acromion and the acromion is low lying or if it is anteriorly
the AC joint should be examined for osteophytes or laterally sloping, because it can cause bony
or irregularities that can be responsible for impingement in these instances.
A B
FIGURE 10-6 ■ Torn supraspinatus (two examples). A, This oblique coronal FSE T2-weighted image with fat suppres-
sion shows a gap in the supraspinatus tendon (arrow). B, This oblique coronal FSE T2-weighted image with fat
suppression in a different patient shows a gap in the supraspinatus tendon (arrow).
A B
FIGURE 10-12 ■ Labroligamentous tear. A, An axial FSE T2-weighted image with fat suppression through the infe-
rior glenoid shows detachment of the inferior glenohumeral ligament from the anterior labrum (arrow). B, One
slice more inferiorly shows the stripping of the ligament extending down the labrum (arrow).
A B
FIGURE 10-14 ■ Sublabral foramen. A, An axial FSE T2-weighted image with fat suppression through the glenohu-
meral joint at the level of the coracoid shows a space between the anterior labrum and the bony glenoid (white
arrow), which is a sublabral foramen. A detached labrum could have this appearance, but this is in the anterosu-
perior part of the joint. The middle glenohumeral ligament (black arrow) is seen as a separate structure anterior
to the labrum. B, A few slices inferiorly, the labrum is seen attached firmly to the anterior glenoid.
the glenoid, a location that is virtually immune It can be impinged by an abnormal acromion
from isolated labral pathology. A much less com- in the same way the supraspinatus tendon is
mon variant is the Buford complex. It consists impinged, resulting in tendinosis, a partial tear,
of an absent anterosuperior labrum with a thick- or a complete tear, or it can undergo myxoid
ened, cordlike middle glenohumeral ligament degeneration similar to the cuff. In tenosynovitis,
(Figure 10-15). It is found in only about 1% to fluid can be seen in the tendon sheath surround-
3% of shoulders. ing an otherwise normal tendon. Because fluid
in the glenohumeral joint can normally fill the
biceps tendon sheath, this diagnosis is difficult
BICEPS TENDON to make with MRI examination alone. If the ten-
don is enlarged or has signal within it, tendinosis
The long head of the biceps tendon runs in the is present (Figure 10-16). If the tendon is not
bicipital groove between the greater and lesser seen on one or more of the axial images, it is
tuberosities and inserts onto the superior labrum. disrupted (Figure 10-17) or dislocated. When
B
FIGURE 10-15 ■ Buford complex. A, An axial FSE T2-
weighted image with fat suppression through the upper
part of the joint shows the anterior labrum separated
from the bony glenoid (arrow). B, Lower in the joint the
anterior labrum is firmly attached to the glenoid, but a
thick, cordlike middle glenohumeral ligament is present FIGURE 10-17 ■ Ruptured biceps tendon. An axial gradi-
(arrow). This is a Buford complex. In (A), the anterior ent-echo image shows the biceps tendon sheath filled
labrum is absent and a thick middle glenohumeral liga- with fluid (arrow) but with no tendon. This indicates a
ment is simulating the anterior labrum. torn biceps tendon.
dislocation occurs, the tendon can be seen to anterior to posterior, just medial to the coracoid
lie anteromedial to the joint (Figure 10-18), and process. It gives off a branch that innervates the
it is inferred that the superior fibers of the sub- supraspinatus muscle as it courses posteriorly in
scapularis are torn. the suprascapular notch. It then innervates the
infraspinatus as it runs inferiorly through the
spinoglenoid notch in the posterior scapula. A
SUPRASCAPULAR NERVE not uncommon finding is a ganglion in the spi-
ENTRAPMENT noglenoid notch that impresses the infraspinatus
portion of the nerve with resultant pain. Atrophy
The suprascapular nerve is made up of branches and/or edema of the infraspinatus muscle is seen
from the C4, C5, and C6 roots of the brachial on MRI (Figure 10-19).8 These are always asso-
plexus. It runs superior to the scapula, from ciated with a torn or detached posterior labrum.
A B
FIGURE 10-19 ■ Ganglion in spinoglenoid notch. A, An axial FSE T2-weighted image with fat suppression reveals
a high-signal mass posterior to the scapula in the spinoglenoid notch. Note the neurogenic edema in the infra-
spinatus muscle (arrow). B, A sagittal image shows the infraspinatus edema (arrow) as well as atrophy (compare
the size of the infraspinatus muscle with the adjacent teres minor and note the difference). This is a ganglion that
has impressed the suprascapular nerve, causing shoulder pain and atrophy of the infraspinatus muscle.
This has been reported almost exclusively in several patients who have had needless surgery
males who are athletic, particularly weight lift- for presumed cuff pathology when the real prob-
ers. It is important to see on preoperative MRI, lem was quadrilateral space syndrome. Generally,
because the ganglion is extracapsular and cannot no surgery is necessary, because physical therapy
be seen during arthroscopy. Symptoms in these is usually successful in breaking up the fibrous
patients can clinically mimic those of having a bands or scar tissue that cause this entity.
rotator cuff tear; hence, MRI examination is
critical to making this diagnosis.
PARSONAGE-TURNER SYNDROME
QUADRILATERAL SPACE SYNDROME After I look at the oblique sagittal T1-weighted
images for fatty atrophy, I look at the oblique sag-
My search pattern for a shoulder MRI begins ittal FSE T2-weighted fat-suppressed images for
with the oblique sagittal T1-weighted images to muscle edema. In about 1% of cases neurogenic
look for fatty atrophy in any of the cuff muscles. edema is found in muscle groups, which corre-
If the infraspinatus is smaller than the other sponds to a particular nerve (i.e., supraspinatus/
muscles or has fatty infiltration, the aforemen- infraspinatus ! suprascapular nerve; teres minor/
tioned suprascapular nerve entrapment second- deltoid ! axillary nerve). This is characteristic for
ary to a ganglion in the spinoglenoid notch is Parsonage-Turner syndrome (Figure 10-21). It’s
the likely diagnosis. If the teres minor has fatty not pathognomonic because a traumatic nerve
atrophy (Figure 10-20), the only diagnosis I’m injury (e.g., a brachial plexus injury) could have a
aware of is quadrilateral space syndrome. This similar appearance. It becomes pathognomonic
most commonly occurs from fibrous bands or once the clinical presentation is provided. If there
scar tissue in the quadrilateral space impinging is no history of trauma or of an insidious onset and
on the axillary nerve. The quadrilateral space if the onset is sudden, with severe pain, followed in
lies between the teres minor superiorly, the teres a day or two with profound weakness, the edema
major inferiorly, the long head of the triceps pattern is virtually pathognomonic for Parsonage-
medially, and the diaphysis of the humerus later- Turner syndrome.
ally. The axillary nerve traverses the quadrilat- The etiology of Parsonage-Turner syndrome is
eral space and innervates the teres minor and unknown, but it seems to have an association with
deltoid muscles; however, the deltoid is never prior vaccinations, viral illness, or general anesthe-
involved in quadrilateral space syndrome. We sia in about one third of cases. It is bilateral
have found quadrilateral space syndrome in in about 10% to 15% of cases. It affects people
about 1% of our shoulder MRI scans. These of all ages and both sexes and is self-limited.
patients can present clinically similar to those
who have a rotator cuff tear, and I have seen
It can affect either the axillary or suprascapular 2. Palmer WJ, Brown JH, Rosenthal DI: Rotator cuff: evalu-
nerve, or both simultaneously. For that matter, ation with fat-suppressed MR arthrography. Radiology
188:683–688, 1993.
it can affect any of the brachial plexus nerves, 3. Neviaser R, Neviaser T: Observations on impingement.
including the long thoracic and phrenic nerves. Clin Orthop Relat Res 254:60–63, 1990.
I have seen unnecessary shoulder, brachial plexus, 4. Budoff JE, Nirschl RP, Guidi EJ: Debridement of partial-
and cervical spine surgery performed before the thickness tears of the rotator cuff without acromioplasty.
Long-term follow-up and review of the literature. J Bone
correct diagnosis of Parsonage-Turner syndrome Joint Surg Am 80:733–748, 1998.
was made. 5. Vinson EN, Helms CA, Higgins LD: Rim-rent tears of
Parsonage-Turner syndrome was first de- the rotator cuff: a common and easily overlooked partial
scribed in the radiology literature in 1998, indi- tear. AJR Am J Roentgenol 189:943–946, 2007.
cating we all missed it on MRI examination for 6. Kjellin I, Ho CP, Cervilla V, et al: Alterations in the supra-
spinatus tendon at MR imaging: correlation with histo-
more than 15 years.9 That’s because we did not pathologic findings in cadavers. Radiology 181:837–841,
routinely fat suppress our shoulder images until 1991.
the early 1990s, and the edema in the muscles 7. Fukuda H: The management of partial-thickness tears of
was not conspicuous enough to be picked up on the rotator cuff. J Bone Joint Surg Br 85:3–11, 2003.
8. Fritz R, Helms C, Steinbach L, Genant H: Suprascapular
non-fat-suppressed sequences. nerve entrapment: evaluation with MR imaging. Radiology
182:437–444, 1992.
9. Helms CA, Martinez S, Speer KP: Acute brachial neuritis
REFERENCES (Parsonage-Turner-syndrome): MR imaging appearance—
1. Zlatkin MB, Iannotti JP, Roberts MC, et al: Rotator cuff report of three cases. Radiology 207:255–259, 1998.
tears: diagnostic performance of MR imaging. Radiology
172:223–229, 1989.
BROAD-BASED FOCAL
FIGURE 11-4 ■ Schematic of types of disc bulges. The
broad-based disc bulge (left) is typical for a bulging
annulus fibrosus. A focal disc bulge (right) is more
consistent with a protrusion.
bulge on a CT or MRI scan does not mean it is fragment. Missing a free disc fragment is one of
clinically significant. the leading causes of failed back surgery, and
One of the most widely used classifications identifying it will guide the surgeon to search for
has the terms protrusion, extrusion, and extruded it and remove it. I have seen several cases in
as the basis for describing the type of disc bulge which the terms extrusion and extruded were mis-
present. I have seen this terminology cause mis- used or not understood and patients had free
adventures for patients because it was either fragments left behind (the surgeon didn’t realize
used incorrectly by the radiologist or not under- the term extruded meant a free fragment), and in
stood by the surgeon. In this classification a others, the surgeon searched for a free fragment
protrusion is a focal disc bulge with a wide neck when there was none because the term extruded
or base (Figure 11-5, A); an extrusion is a focal was incorrectly applied. Because there is no
bulge with a narrow neck or base (Figure 11-5, B); clinical or surgical difference between having a
and an extruded disc is a free disc fragment protrusion or an extrusion and surgical mistakes
(Figure 11-5, C). Surgeons don’t treat a patient can occur when this terminology is used, I do
with a protrusion any differently than one with not let my residents and fellows use these terms.
an extrusion; it’s an artificial distinction based MRI has a high degree of accuracy in delin-
solely on the neck width of the bulge. If that’s all eating disc protrusions and showing if neural
it were, so what, but it’s much more significant, tissue is impressed (Figure 11-6). MRI scans
because of the term extruded, which is a free can also show if annular fibers of the disc are
FIGURE 11-5 ■ Disc protrusions. A, A focal disc protrusion with a wide neck or
base (arrow). This has been termed a protrusion. B, A focal disc protrusion with
a narrow neck or base (arrow). This has been termed an extrusion. C, A piece
of disc that has broken off as a free fragment. This has been termed an
extruded disc.
A B
FIGURE 11-6 ■ Disc protrusions. A, An axial T1-weighted image shows a focal disc protrusion (arrows). B, An axial
T2-weighted image shows a broad-based disc bulge (arrows). Because these are both showing impression of the
thecal sac, they could each cause symptoms.
disrupted (Figure 11-7)—a so-called HIZ (high- to the parent disc or is really “free.” As long
intensity zone). Although annular tears can’t be as disc material is above or below the level of
diagnosed on the basis of CT examination, clini- the disc space, it really does not matter if it is
cians do not currently treat them surgically. The attached or not. The key element is recognizing
annulus is innervated by the sinuvertebral nerve, that disc material is present away from the level
which goes to the dorsal root ganglion and can of the disc space (caudally or cranially) so that the
mimic a focal disc protrusion at that level. They surgeon will be aware that he or she may have to
can cause back pain and even sciatica (buttock increase his or her exposure to find and account
and leg pain), but they typically resolve with for the additional disc material, whether it is
conservative management. attached to the parent disc or not.
Free fragments are diagnosed on MRI scans
by noting disc material cephalad or caudal to
Free Fragments the disc space (Figure 11-8). Free fragments
A type of disc abnormality that is critical to diag- may migrate either cranially or caudally with no
nose is the free fragment or sequestration. Missed apparent preference.
free fragments are one of the most common Axial images often show the free fragment
causes of failed back surgery.6 The preoperative more conspicuously than the sagittal images
diagnosis of a free fragment means the surgeon (Figure 11-9); therefore contiguous axial images
needs to explore more cephalad or caudally dur- without large skip areas or gaps are imperative to
ing the surgery to remove the free fragment. prevent missing free fragments.
Because free fragments can be very difficult to A conjoined root, which is a normal variant of
diagnose clinically, imaging is critical in the two roots exiting the thecal sac together or in an
evaluation of the spine for any patient contem- anomalous manner (seen in 1% to 3% of the
plating surgery. At times it can be difficult to be population)7 (Figure 11-10), or a Tarlov cyst, a
absolutely certain as to whether or not a disc that normal variant in which a nerve root sleeve is
lies above or below the disc space is still attached dilated, can have a similar appearance to a free
Lateral Discs
Discs will occasionally protrude in a lateral direc-
tion, causing the nerve root that has already exited
the central canal to be stretched (Figure 11-11).
Although not common (less than 5% of cases),
B these are frequently overlooked and are known to
FIGURE 11-9 ■ Free fragment more evident on axial be a source of failed back surgery.8 Because they
images. The axial image (A) reveals an obvious large affect the previously exited root, they can clinically
free disc fragment (arrow). It is not as conspicuous on mimic symptoms of a disc protrusion from one
the sagittal image (B). The free fragment can be seen level more cephalad (Figure 11-12). For example,
posterior to the L4 vertebral body (arrow).
in a patient with multilevel disc disease and
symptoms referable to the L3–4 disc, the disc
protrusion is usually a posterior L3–4 bulge that
fragment but can almost always be differentiated impresses the L4 nerve root. However, a lateral
from disc material by the signal staying isoin- disc at L4–5 could impress the L4 nerve root
tense to the thecal sac on both T1 and T2 se- and cause the same symptoms. If not noticed,
quences. It is critical to identify a conjoined root surgery could be performed at the L3–4 disc—
or a Tarlov cyst and not confuse them for a free the wrong level. Unfortunately, I have seen this
fragment. Surgeons will often change their pro- on several occasions. Also, it is important to
cedure and certainly their amount of surgical notify the surgeon that the disc is lateral to the
Copyright © 2014 Elsevier, Inc. All rights reserved.
198 11 Lumbar Spine: Disc Disease and Stenosis
A B
FIGURE 11-11 ■ Lateral disc. A, A sagittal T1-weighted MRI scan through the left neuroforamen shows a low-signal
structure in the L4 neuroforamen (arrow), which is a lateral disc protrusion. B, Axial T1-weighted (upper) and
T2*-weighted (lower) images show the lateral disc (arrows) in the left neuroforamen.
FIGURE 11-13 ■ Central canal stenosis. An axial T2- FIGURE 11-15 ■ Ligamentum flavum hypertrophy. Inward
weighted image shows marked compression of the bulging of the ligamentum flavum (arrows) is shown on
thecal sac in an anteroposterior direction, diagnostic this MRI scan. Central canal stenosis from ligamentum
of central canal stenosis. flavum hypertrophy is common.
Neuroforaminal Stenosis
DJD of the facets with bony hypertrophy is the
most common cause of neuroforaminal stenosis;
however, encroachment on the nerve root in the
neuroforamen can be seen with free disc frag-
ments, with postoperative scarring, and from a
lateral disc protrusion.
The neuroforamen is best evaluated on axial
images just cephalad to the disc space. The disc
space lies at the inferior portion of the neurofo- FIGURE 11-17 ■ Neuroforaminal stenosis. Marked nar-
ramen, and the exiting nerve root lies in the rowing of the neuroforamen (arrow) is seen on this
superior or cephalad portion of the neurofora- sagittal T1-weighted image.
men. Although the neuroforamen can be clearly
seen on sagittal MRI scans (Figure 11-16), care Lateral Recess Stenosis
must be taken to evaluate the entire neurofora-
men and not just the 4 or 5 mm of one sagittal The lateral recesses are the bony canals in which
image. A normal-appearing neuroforamen on a the nerve roots lie after they leave the thecal sac
sagittal image does not exclude neuroforaminal and before they enter the neuroforamen. The
stenosis (the axial images must be evaluated), lateral recesses are bounded by the neurofora-
whereas a stenotic foramen seen on a sagittal men caudally and cranially. They are triangular
image (Figure 11-17) can be counted on as being in shape when viewed on axial images, and the
reliable for pathology. nerve root can be identified as a low-signal
rounded structure on all imaging sequences.
Hypertrophy of the superior articular facet from
DJD is the most common cause of encroachment
on the lateral recesses (Figure 11-18), although,
as with the neuroforamen, disc fragments and
postoperative scarring can cause nerve root
impingement.
A B
C
FIGURE 11-19 ■ Postoperative scar. A, A T1-weighted axial image shows a left-sided mass that could be either
postoperative scar tissue or a recurrent disc. B, A T2-weighted image through the same level shows the soft
tissue mass is wrapping around the thecal sac rather than impressing it. This is typical for scar tissue, not
recurrent disc protrusion. C, After administration of Gd-DTPA intravenously, enhancement of the scar tissue
adjacent to the thecal sac is seen. The diagnosis of postoperative fibrosis could be made without the admin-
istration of gadolinium.
Modic et al.10 and was termed type 2 changes the routine lumbar spine MRI and include
(generally referred to as Modic type 2). Modic annular tears, pars breaks (spondylolysis), and
type 1 changes are seen as low-signal bands facet disease. In addition, sacroiliac joint and
parallel to the end-plates on T1-weighted im- sacral abnormalities can mimic disc disease
ages that get brighter on T2-weighted images and should be looked for on every spine MRI
(Figure 11-22). This represents an inflamma- scan (Figure 11-24). Tendinosis or partial tears
tory or granulomatous response to degenerative at the hamstring insertion on the ischial tuber-
disc disease. The type 1 changes were reported osity is often misdiagnosed as sciatica with
in 4% of cases and must be distinguished from resultant L-spine MRI examination. I have seen
disc space infection (Figure 11-23). In disc space several patients who had focal disc protrusions
infection the disc should get bright on the that were surgically removed as the putative
T2-weighted images, whereas it is unusual for cause of sciatica when in fact the “sciatica” was
a degenerative disc to have high signal on T2- secondary to an abnormality at the hamstring
weighted images. Modic type 3 changes are insertion; the disc was an incidental, asymptom-
parallel bands of low signal adjacent to the end- atic bulge. Disc surgery in those cases, obvi-
plates on both T1- and T2-weighted images. ously, was unnecessary and did not improve
Type 3 changes represent bony sclerosis seen on their pain.
plain films. Last, a lumbar spine MRI scan typically
includes at least part of the kidneys. A renal
cell carcinoma can cause back pain that can be
Mimics mistakenly attributed to a disc protrusion
Multiple entities can mimic disc disease clini- and, if overlooked on the L-spine MRI scan,
cally and, if not noted, can result in unnecessary can result in a tragic outcome for the patient
disc surgery. Many of these can be identified on (Figure 11-25).
A B
FIGURE 11-22 ■ Type 1 marrow changes. A, A sagittal T1-weighted image in a patient with degenerative disc disease at
L3–4 shows faint bands of low signal parallel to the L3–4 end-plates (arrows). B, A sagittal fast spin-echo T2-weighted
image with fat suppression shows bands of high signal adjacent to the L3–4 end-plates. This represents granulation
tissue seen with degenerative disc disease and has been called type 1 marrow change. It can be differentiated from a
disc infection by the low signal of the disc on the T2-weighted image.
A B
FIGURE 11-23 ■ Disc infection. A, A sagittal T1-weighted image shows bands of low signal in the vertebral bodies
adjacent to the L4–5 end-plates. B, On a T2-weighted (gradient-echo) image, the vertebral body/end-plate signal
increase is faintly seen (because it is a gradient-echo sequence). However, note the high signal in the disc, which
makes this consistent with a disc infection rather than type 2 signal of a degenerative disc.
B
FIGURE 12-2 ■ Normal ankle anatomy. A, This drawing of
the tendons around the ankle at the level of the tibiotalar
joint shows the relationship of the flexor tendons poste-
riorly and the extensor tendons anteriorly. B, An axial
T1-weighted image through the ankle just above the
tibiotalar joint shows the normal anatomy. A, Achilles
tendon; D, flexor digitorum tendon; H, flexor hallucis
tendon; P, peroneal tendons; T, posterior tibial tendon;
TA, tibialis anterior tendon.
A B
FIGURE 12-4 ■ Torn posterior tibial tendon. Axial T1-weighted (A) and T2-weighted (B) images through the ankle
in this patient with chronic pain reveal a distended posterior tibial tendon sheath (arrows) with no low-signal
tendon identified within. This is a tear of the posterior tibial tendon.
TUMORS
There are a few tumors that have a predilection
for the foot and ankle.4 Up to 16% of synovial
sarcomas occur in the foot. Desmoid tumors are
commonly seen in the foot. Giant cell tumors of
tendon sheath are often found in the tendon B
sheaths of the foot and ankle (Figure 12-10). FIGURE 12-10 ■ Giant cell tumor of the tendon sheath.
They are characterized by marked low signal in Axial proton-density (A) and T2-weighted (B) images
the synovial lining and in the tendons on T1- and reveal a mass surrounding the flexor hallucis longus
T2-weighted images, similar to the appearance tendon (arrow), which is confined by the tendon
sheath. Although high-signal fluid is present, a large
of pigmented villonodular synovitis in a joint. amount of low-signal material is lining the distended
The differential diagnosis for calcaneal tu- tendon sheath. This low signal is hemosiderin, which
mors is similar to that of the epiphyses: giant is typically found in a giant cell tendon of tendon
cell tumor, chondroblastoma, infection, as well sheath. Pigmented villonodular synovitis in a joint has
as a unicameral bone cyst (Figure 12-11). This an identical appearance.
differential works more than 95% of the time
for epiphyseal lesions; however, it may be less
Copyright © 2014 Elsevier, Inc. All rights reserved.
212 12 Magnetic Resonance Imaging of the Foot and Ankle
than 50% inclusive in the calcaneus—but it’s a a treatable lesion (i.e., a mass) in many cases.
good starting point. An MRI examination for tarsal tunnel syndrome
Soft tissue tumors in the medial aspect of the is becoming increasingly requested as surgeons
foot and ankle can press on the posterior tibial learn how valuable it can be in identifying the
nerve, resulting in tarsal tunnel syndrome.5 source of the symptoms.
Clinically, patients with tarsal tunnel syndrome Anomalous muscles in the foot or ankle are
have pain and paresthesia in the plantar aspect reported to be present in up to 6% of the popula-
of the foot. In the aforementioned mnemonic, tion. These can be mistaken for a tumor, and a
“Tom, Dick, and Harry,” the and is for artery, biopsy may be unnecessarily performed. MRI
nerve, and vein. It is the position of the poste- scans will show these “tumors” to have imaging
rior tibial nerve. The nerve is easily compressed characteristics identical to those of normal muscle
in the tarsal tunnel, which is bounded medially (Figure 12-13) and to be sharply circumscribed.
by the flexor retinaculum—a strong fibrous Accessory soleus and peroneus quartus muscles
band that extends across the medial ankle joint are the most common accessory muscles encoun-
for about 5 to 7 cm in a superior to inferior tered around the foot and ankle.
direction. Ganglions and neural tumors, both
of which can look similar on T1- and T2-
weighted images, often lie in the tarsal tunnel LIGAMENTS
(Figure 12-12) and compress the posterior tibial
nerve, resulting in pain and paresthesia on the MRI is not the best way to diagnose acute ankle
plantar aspect of the foot extending into the ligament abnormalities. The clinical evaluation
toes. Tarsal tunnel syndrome often occurs is usually straightforward, and no diagnostic
secondary to trauma, fibrosis, or idiopathically, imaging of any type is necessary. Nevertheless,
all of which may not respond to surgical inter- in clinically equivocal cases, when the examina-
vention; hence, MRI is valuable in delineating tion is ordered for other reasons, or in cases of
Copyright © 2014 Elsevier, Inc. All rights reserved.
12 Magnetic Resonance Imaging of the Foot and Ankle 213
FIGURE 12-13 ■ Anomalous muscle. An axial T1-weighted image of both ankles in this patient complaining of a
mass in the right ankle shows an anomalous muscle (arrow), a peroneus quartus, lateral to the flexor hallucis
longus muscle, which is responsible for the mass the patient feels.
chronic lateral ankle pain, the ligaments can be up the anterior capsule of the joint (Figure 12-16).
clearly evaluated with high-quality MRI.6 The anterior talofibular ligament is usually
The deltoid ligament lies medially as a broad torn without other ligaments being involved;
band beneath the medial flexor tendons. It can be however, if the injury is severe enough, the next
seen extending from the medial malleolus to the ligament to tear is the calcaneofibular ligament.
talus and calcaneus. It has multiple parts that are Even with very severe trauma, the posterior
not easily separated with MRI. It is considered talofibular ligament rarely tears. The anterior
injured when it has increased T2 signal or dis- talofibular ligament always tears before the cal-
ruption of the linear fibers. Surgery to repair a caneofibular ligament. A torn anterior talofibu-
torn deltoid is uncommon, whereas reconstruc- lar ligament will be absent in some chronic tears
tion of the lateral ligament complex is commonly but can be thickened (Figure 12-16, B); it can
performed. be attenuated or have a gap also (Figure 12-17).
The lateral collateral ligament complex is
responsible for more than 90% of all ankle
ligament injuries. It is made up of two parts:
Sinus Tarsi Syndrome
(1) a superior group, the anterior and posterior An entity that has a high association with torn
tib-fib ligaments that make up part of the syn- lateral collateral ligaments in the ankle is sinus
desmosis (Figure 12-14), (2) and an inferior tarsi syndrome. Clinically affected patients have
group, the anterior and posterior talofibular lateral ankle pain and a feeling of hindfoot insta-
ligaments and the calcaneofibular ligament bility. Up to 80% of these patients have torn
(Figure 12-15). The anterior and posterior tib- lateral collateral ligaments and up to one third of
iofibular ligaments can be seen on axial images patients who tear their lateral collateral liga-
at the level of the dome of the talus. The ante- ments have been reported to have sinus tarsi
rior and posterior tib-fib ligaments are seen on syndrome.7 In the past, diagnosis has relied on
the axial images just below the tibiotalar joint clinical suspicion and injection of lidocaine into
and emanate from a concavity in the distal fibula the sinus tarsus, which causes resolution of the
called the malleolar fossa (Figure 12-15, B). The pain. Treatment is varied but can include a joint
most commonly torn ankle ligament is the ante- fusion.
rior talofibular ligament. It is easily identified The sinus tarsi is the space that lies between
when a joint effusion is present because it makes the talus and the calcaneus and opens up in a
Copyright © 2014 Elsevier, Inc. All rights reserved.
214 12 Magnetic Resonance Imaging of the Foot and Ankle
Anterior
Anterior
Tib-Fib
Tib-Fib
Posterior
Tib-Fib TIBIA
Posterior
A B Tib-Fib
FIGURE 12-14 ■ Schematic of lateral collateral ligaments. A, This drawing of the ankle in a lateral view shows
how the anterior and posterior tib-fib ligaments extend off of the fibula and course superiorly to the tibia.
B, A drawing in the axial plane shows that the fibula has a flat or convex surface at the origin of these
ligaments.
Anterior
Anterior Posterior talofibular
talofibular talofibular
TALUS
Posterior
A Calcaneofibular B talofibular
FIGURE 12-15 ■ Schematic of lateral collateral ligaments. A, This drawing of the ankle in a lateral view shows
how the anterior and posterior talofibular ligaments and the calcaneofibular ligament extend off of
the fibula and course inferiorly. These ligaments arise off of the fibula more distally than the anterior and
posterior tib-fib ligaments. B, A drawing in the axial plane shows that the anterior and posterior talofibu-
lar ligaments arise from the level of the distal fibula, which has a concave medial surface, the malleolar
fossa.
A B
FIGURE 12-16 ■ Anterior talofibular ligament. A, An axial T2-weighted image through the distal fibula at the level
of the malleolar fossa (the concave medial surface of the fibula) shows an intact anterior talofibular ligament
(arrow), which makes up part of the joint capsule at this level. Note the high-signal joint fluid adjacent to the
ligament. B, This axial T2-weighted image at the level of the malleolar fossa reveals a thickened anterior talofibular
ligament (arrow). The marked thickening of the ligament indicates a chronic process with scarring.
TALUS
Interosseous
ligament
Cervical CALCANEUS
ligament
SAGITTAL
FIGURE 12-18 ■ Schematic of the sinus tarsi. A sagittal
schematic of the sinus tarsi show the positions of the
cervical and the interosseous ligaments. The cervical
ligament lies more anteriorly and laterally in relation
to the interosseous ligament.
A
A
RIGHT
B
FIGURE 12-22 ■ Osteomyelitis. Axial T1-weighted (A) and
T2-weighted (B) images through the forefoot in this
diabetic patient shows low T1 signal (arrow, A) and high
T2 signal (arrow, B) in the lateral sesamoid. Cortical
disruption is present which is diagnostic for osteomyelitis.
which were discussed earlier in this chapter) has The synovium becomes thickened and scarred
been reported; both of these conditions can be a from an inversion ankle injury and causes a pain-
cause of chronic lateral ankle pain by themselves. ful, mechanical block to dorsiflexion. It is easily
Another cause of chronic lateral ankle pain that identified with MRI by noting low-signal scar
has a high association with chronic disruption of tissue in the joint just deep to the anterior
the anterior talofibular ligament is anterolateral talofibular ligament where one should see
impingement or lateral gutter syndrome. The joint fluid (Figure 12-23). It is treated by ar-
lateral gutter is simply the lateral joint space be- throscopically debriding the scar tissue and, if
tween the distal tibia and talus medially and the necessary, reconstructing the lateral collateral
fibula and lateral collateral ligaments laterally. ligaments.
A B
FIGURE 12-23 ■ Anterolateral impingement. A, This axial T1-weighted image through the ankle reveals absence of the
anterior talofibular ligament (arrow). B, The corresponding T2-weighted image shows low-signal scar
tissue deep to the expected location of the anterior talofibular ligament (arrow), which indicates anterolateral impinge-
ment syndrome.
Miscellaneous Magnetic
Resonance Imaging
There are several additional areas in which mag- significance of a torn TFC. That is because
netic resonance imaging (MRI) is useful but not torn TFCs (and torn intercarpal ligaments, for
well enough developed or of enough widespread that matter) are found with a high frequency
use to have an entire chapter devoted to them. in older patients who do not have wrist pain or
Included in this group are MRI examinations of dysfunction. Nevertheless, in a young patient
the wrist, hip, elbow, and bone marrow. with a painful, torn TFC, most hand surgeons
would surgically intervene if conservative care
was ineffective. For this reason, imaging may
WRIST play a role.
The normal TFC is predominantly low
MRI examination of the wrist has been slower signal on all imaging sequences and seen to be
to develop than that of other joints. Similarly, triangular in shape, with the base attaching
wrist arthrography has not enjoyed the same to the ulna and the apex attaching onto the
popularity as that of the knee or shoulder. radius (Figure 13-1). A detached or torn
Nevertheless, MRI of the wrist has some defi- TFC is best seen in the coronal plane with
nite utility. It is useful in evaluating the carpal T2 or gradient-echo sequences and is usually
bones for fractures and avascular necrosis accompanied by joint fluid in both the distal
(AVN). It seems to have some use for evaluat- radioulnar and the proximal carpal joints
ing the triangular fibrocartilage (TFC) and the (Figure 13-2).
intercarpal ligaments.1
Avascular Necrosis
Imaging Techniques The wrist has several bones that have a propen-
Thin-section (2 to 3 mm) T1- and T2-weighted sity to undergo AVN. The lunate is commonly
images in both an axial and a coronal plane are affected and is known as Kienböck’s malacia.
typically used with a dedicated wrist coil or a It is seen as uniform low signal on T1- and
small surface coil. Some recommend sagittal T2-weighted images (Figure 13-3). As is found
images as well. A small field of view (FOV; 5 to with AVN in other joints, MRI examination can
8 cm) should be used for maximal resolution. be useful in showing AVN when plain films are
Three-dimensional volumetric coronal images normal.
with thin (1 to 2 mm) slices are used in many The proximal pole of the scaphoid often un-
centers to replace the T2-weighted images. dergoes AVN after a fracture is sustained. MRI
These are especially useful for examining the can demonstrate the AVN earlier than plain
TFC and the intercarpal ligaments.2 films, allowing earlier treatment (Figure 13-4).
Subtle or occult fractures of the scaphoid (or any
other carpal bone) can be identified with MRI
Pathology examination. MRI should be considered when
clinical suspicion of a fracture is high and plain
Triangular Fibrocartilage
films are negative, because a missed fracture of
The TFC lies between the distal ulna and the the scaphoid can lead to AVN. We have found it
carpal bones and is thought to have some cost-effective to obtain an MRI scan of the wrist
shock-absorbing function. It can tear or be- to rule out a fracture in a patient with trauma,
come detached and cause significant wrist pain pain in the snuffbox, and a negative plain film, as
and dysfunction. Tears of the TFC can be opposed to casting the patient for a week and
diagnosed with arthrography or with MRI, having him or her return for a follow-up x-ray
although it is somewhat controversial as to the examination.3
A B
FIGURE 13-4 ■ Scaphoid AVN. Coronal images in a patient who sustained a fracture of his scaphoid several months
prior shows low signal in the proximal pole on the T1-weighted images (A), which is high in signal on the
T2-weighted images (B). This is worrisome but not diagnostic of AVN. If the T2-weighted images were low in
signal, it would indicate AVN; however, with high T2 signal, the proximal pole might still be viable.
tendinosis, which can result in a “trigger outcome for the patient. MRI scans cannot dif-
thumb.” This is where the thumb catches on ferentiate a sterile tenosynovitis from an infectious
attempted flexion and then suddenly gives way. tenosynovitis.
This is called de Quervain’s syndrome. MRI
shows the fluid distending the tendon sheath
(tenosynovitis) and will show the tendinosis, if HIP
present (Figure 13-8).
Tenosynovitis in any of the flexor tendons MRI has proved to be useful in a number
(Figure 13-9) is considered a surgical emergency of abnormalities of the hip, including AVN,
by most hand surgeons, and when seen on MRI, it fractures (see Chapter 5), idiopathic transient
deserves a phone call to the surgeon. An infection osteoporosis of the hip (ITOH; see Chapter 8),
in a flexor tendon sheath can easily spread to the torn acetabular labra, and femoroacetabular
common flexors in the wrist and lead to a terrible impingement (FAI).
Osteonecrosis (AVN)
AVN can be diagnosed with great sensitivity
with MRI. It has a characteristic appearance,
with involvement of the anterosuperior portion
of the femoral head. The area of AVN typically
is surrounded by a low-signal serpiginous border
(Figure 13-10). AVN can be diagnosed earlier
and more reliably based on MRI scans than on
plain films or with nuclear medicine.
Acetabular Labrum
The acetabular labrum is analogous to the
glenoid labrum of the shoulder. It can tear or
detach, resulting in a painful, snapping hip.
FIGURE 13-8 ■ De Quervain’s syndrome. This axial T2-
The labrum can be arthroscopically debrided
weighted image through the wrist at the distal radius or repaired, with multiple surgeons reporting a
in a patient complaining of a trigger thumb shows good clinical outcome.
distention of the common tendon sheath surrounding The acetabular labrum is best visualized
the abductor pollicis longus and the extensor pollicis with a magnetic resonance arthrogram.4 Imag-
brevis tendons (arrow). The fluid is not very T2 bright,
suggesting synovitis. The abductor pollicis longus ing should be done with a small FOV and only
is somewhat thickened, with some increased signal one hip studied. Bilateral examinations with a
indicative of tendinosis. large FOV and nonarthrogram studies have an
Copyright © 2014 Elsevier, Inc. All rights reserved.
13 Miscellaneous Magnetic Resonance Imaging 223
ELBOW
MRI examination of the elbow has been shown
to be useful in diagnosing collateral ligament
tears and biceps tendon tears and identifying
loose bodies. It can show high T1 signal in the
flexor tendons as they insert on the medial epi-
condyle and in the extensor tendons on the
FIGURE 13-11 ■ Torn acetabular labrum. This coronal
T1-weighted, fat-suppressed arthrogram of the hip
lateral epicondyle in so-called golfer’s and tennis
shows fluid between the bony acetabulum and the elbow. Tendinosis of the flexor and extensor ten-
labrum (arrow), indicating a detached labrum. dons is the most common abnormality seen on
Copyright © 2014 Elsevier, Inc. All rights reserved.
224 13 Miscellaneous Magnetic Resonance Imaging
BONE MARROW
MRI readily shows the bone marrow throughout
the skeleton. Its appearance varies with age and
location within the skeleton. For instance, a
young person’s skeleton has more red (hemato-
poietic) marrow, whereas an elderly person’s has
more fatty marrow. The axial skeleton (spine and
pelvis) has more red marrow than the peripheral
FIGURE 13-13 ■ Partial tear of the common extensors. skeleton.
A coronal T2-weighted image of the elbow shows One of the best tips for differentiating red
fluid signal at the insertion of the common extensor marrow from an infiltrative process is that red
tendons onto the lateral epicondyle (arrow). This indi- marrow is always higher in signal on T1-weighted
cates a partial tear of the tendons.
images than adjacent muscle, or, in the lumbar
spine, higher in signal than the discs (Figure 13-15).
elbow MRI scans. Partial tears of the flexor or ex- An infiltrative process, such as tumor or infection,
tensor tendons show high T2 signal (Figure 13-13). only rarely is higher in signal than muscle or disc
Bony and cartilaginous abnormalities can also be on a T1 sequence. Even the most severe anemia
seen with MRI. with marked red marrow hyperplasia will obey
The imaging protocol typically includes ax- this rule unless there is iron overload, as in hemo-
ial and coronal T1- and T2-weighted images. chromatosis. Care must be taken not to confuse
Sagittal images are optional in some centers; diffuse increased bone density with abnormal
A B
FIGURE 13-14 ■ Ulnar collateral ligament. Coronal T1-weighted, fat-suppressed arthrograms of the elbow show the
normal ulnar collateral ligament (A) (arrow) and one that is torn off of the attachment to the medial epicondyle (arrow).
FIGURE 13-15 ■ Normal red marrow. This sagittal T1- FIGURE 13-16 ■ Increased red marrow in anemia. This
weighted image of the lumbar spine in a 25-year-old sagittal T1-weighted image through the lumbar spine
shows the appearance of normal red marrow in the in a 40-year-old woman with anemia shows a marked
vertebral bodies. Note that the signal of the marrow is increase in the red marrow. Note that it remains
higher than that of the discs. higher in signal than the discs.
Differential Diagnoses
BUBBLY OR LYTIC LESIONS DENSE BONES (REGULAR SEX MAKES
(FEGNOMASHIC) OCCASIONAL PERVERSIONS MUCH
Fibrous dysplasia MORE PLEASURABLE AND
Enchondroma, eosinophilic granuloma FANTASTIC)
Giant cell tumor
Nonossifying fibroma Renal osteodystrophy
Osteoblastoma Sickle cell disease
Mets and myeloma Myelofibrosis
Aneurysmal bone cyst Osteopetrosis
Solitary bone cyst Pyknodysostosis
Hyperparathyroidism (brown tumor), Mastocytosis
hemangiomas Mets—breast and prostate
Infection Paget’s disease
Chondroblastoma; chondromyxoid fibroid Athletes
Fluorosis
BONY SEQUESTRUM
CPPD-ASSOCIATED DISORDERS
Infection
Primary hyperparathyroidism Eosinophilic granuloma
Gout Osteoid osteoma
Hemochromatosis Fibrosarcoma
Lymphoma
DENSE BASE OF THE SKULL
Fibrous dysplasia
Engelmann’s disease
Von Buchem’s disease
Paget’s disease
Meningioma
Pyknodysostosis
Osteopetrosis
A Arthritis (Continued)
ABCs. See Aneurysmal bone cysts collagen vascular diseases, 124
Abrams, Herbert, 1 crystal-induced, 111 See also Calcium pyrophosphate
Achilles tendons, MRI for, 206–207, 207f dihydrate crystal deposition disease
Achondroplasia, 153 gout, 121–122, 122f
ACL (anterior cruciate ligament), MRI for, 172, 173f, 174f differential diagnosis, 111
Acromegaly, 124 hemophilia, 127–128, 129f, 130f
Adamantinoma, 11f HLA-B27 spondyloarthropathies, 115–121
Aggressive osteoporosis, 141, 141f ankylosing spondylitis, 84–85, 86f, 117–118, 119f, 120f
Aneurysmal bone cysts (ABCs), 22–23, 22f Reiter’s syndrome, 117, 118, 119, 121, 121f
cortical destruction, 32, 33f syndesmophytes, 117–118, 118f
MRI, 22, 23f variants, 115–117
osteoblastomas, 19 joint effusions, 131–133, 134f
primary type, 22–23 diagnosis, 132
secondary type, 22–23 teardrop measurements, 133f, 134f
Ankle large joint involvement, 112t
MRI for, 206 See also Tendons, MRI for osteoarthritis
anatomy, 206, 207f DISH, 112–113, 113f
anterolateral impingement, 218f geodes, 115, 115b, 115f
chronic lateral ankle pain, 216–217, 217b joint space narrowing, 112
osteomyelitis, 217f letter joints, 114, 114b
stress fractures, 217f primary, 113–114, 113f, 114f
Ottawa ankle rules, 4 secondary, 113
after trauma, 4f in shoulder, 112f
unnecessary examinations, 4 SI joints, 102f, 114f
Ankylosing spondylitis, 84–85, 86f, 117–118, 119f, 120f subchondral cysts, 115, 115b, 115f
Anterior cruciate ligament (ACL), MRI for, 172, 173f, 174f paralysis, 127–128
Anterior wedge compression fracture, 84, 85f physical examination, 111
Arm, trauma to psoriatic, 118, 118f, 119, 119f, 120f
displaced fat pads, 95–96, 95f, 96f CT, 120f
fractures, 92–98 PVNS, 131, 131f, 132f
Colles’, 92, 93f rheumatoid, 115, 115b
Galeazzi’s, 93, 94f CPPD, 115
glenoid, 98f of hip, 116f
Monteggia’s, 92–93, 94f juvenile, 127–128, 129f
plastic bowing deformity of forearm, 92, 93f migration routes, 116f
shoulder secondary degenerative disease, 117f
Bankart deformity, 96–97 of shoulder, 117b, 117f
dislocations, 96–98, 97f, 98f sarcoidosis, 124–126, 127f
Hill-Sachs deformity, 96–97 SI joints, 111, 112t
normal view, 97f osteoarthritis, 102f
pseudodislocations, 62f, 63f side-to-side symmetry, 111
Arthritis bilateral, 112b
anatomical distribution, 111 SLE, 124, 125f
in hands and wrists, 112t synovial osteochondromatosis, 128–131
AVN, 133–136 without calcification, 131f
Freiberg’s infraction, 136, 137f development, 128–129
geodes, 136, 136f MRI, 129–131
of hip, 134f, 135f muscular dystrophy, 130f
indications, 133 presentation, 129–131
Kienböck’s malacia, 136, 137f Atypical synovial cysts, 52f
Köhler’s disease, 136, 137f Avascular necrosis (AVN), 153, 154f, 155f
Legg-Perthes disease, 136 arthritis, 133–136
MRI, 135 Freiberg’s infraction, 136, 137f
OCD, 131f, 135, 136f geodes, 136, 136f
Osgood-Schlatter disease, 136 of hip, 134f, 135f
Scheuermann’s disease, 136, 138f indications, 133
of shoulder, 135f Kienböck’s malacia, 136, 137f
Chronic lateral ankle pain, 216–217, 217b “Don’t touch” lesions (Continued)
Clay-shoveler’s fracture, 80, 81f geodes, 57, 60f, 61f
Coccyx, unnecessary examinations, 2–3 MRI, 55
Collagen vascular diseases, 124 myositis ossificans, 55, 55f, 56f, 57f
Colles’ fracture, 92, 93f os odontoideum, 65–68, 67f
Computed tomography (CT) scans osteogenic sarcoma, 56f
cortical destruction, 32 osteophytosis, 57–58
lumbar spine, 194f pseudocyst of humerus, 66f
NOFs, 18 pseudodislocation of shoulder, 62f, 63f
osteoid osteoma, 157–158 from trauma, 57
for osteoporosis, 139 Dorsal defect of patella, 56f, 64
posttraumatic lesions, 55 MRI scans, 61f
psoriatic arthritis, 120f
skull, 1
Contusions, 180, 181f E
Cortical desmoid EG. See Eosinophilic granuloma
parosteal osteosarcoma, 40f Elbows, MRI for
posttraumatic lesions, 56–57, 58f, 59f extensor tears, 224f
Cortical destruction, 32, 32f ulnar collateral ligament, 224f
ABCs, 32, 33f Enchondroma, 12–14, 12f
CT, 32 bone infarcts, 12, 12f
Costochondritis, 59–63, 63f chondrosarcoma, 13
CPPD. See Calcium pyrophosphate dihydrate crystal Maffucci’s syndrome, 13–14
deposition disease MRI, 13
Crystal-induced arthritis, 121–124. See also multiple, 13–14
Calcium pyrophosphate dihydrate crystal Ollier’s disease, 13–14, 13f
deposition disease periostitis, 13
gout, 121–122, 122f Engelmann’s disease, 153, 155f
C-spine. See Cervical spine Eosinophilic granuloma (EG), 10f, 14–15, 14f, 15f
CT scans. See Computed tomography scans appearances, 14
Cystic angiomatosis, 25f bony sequestrum, 15
Cysts. See specific cysts monostotic, 14
presentation, 15b
Ewing’s sarcoma, 35, 36, 37, 41–42, 41f, 42f
D differential diagnosis, 41–42
De Quervain’s syndrome, 222f Examinations. See Unnecessary examinations
Degenerative joint disease. See Osteoarthritis
Desmoid tumors, 43–45, 45f, 46f
cortical desmoid parosteal osteosarcoma, 40f F
Diffuse idiopathic skeletal hyperostosis (DISH), 84–85 FAI. See Femoroacetabular impingement
osteoarthritis, 112–113, 113f Fast spin-echo (FSE) sequences, 166, 182
Disc disease, 193–198, 194f, 204f FEGNOMASHIC, 7–8
mimics, 203 definition, 7–8
renal cell carcinoma, 204f discriminators, 8
sacroiliitis, 204f GCT, 8
Discitis, 146f unicameral bone cyst, 8
Discogenic vertebral disease, 57–58, 61f Femoral stress fractures, 102–103, 103f
Discoid meniscus, 170, 170f, 171f osteoporosis, 140f
DISH (diffuse idiopathic skeletal hyperostosis), 84–85 Femoroacetabular impingement (FAI)
Disuse osteoporosis, 141 Pitt’s pit, 74
“Don’t touch” lesions in wrist, 223, 223f
benign, 68–74 FHL (flexor hallucis longus), 208, 209f
bone infarcts, 71–74, 72f, 73f Fibrous dysplasia, 8–11, 10f
bone islands, 70, 71f adamantinoma, 11f
lipidization, 70–71 assessment, 8
NOF, 68–70, 69f, 70f, 71f chondrosarcoma, 11
Pitt’s pit, 74, 74f classical description, 9
pseudocyst of calcaneus, 72f differential diagnosis, 8, 9
unicameral bone cysts, 70–71, 72f McCune-Albright Syndrome, 9–11
biopsies, 64–65, 68 monostotic, 9, 11f
chondroblastoma, 64–65, 66f periostitis, 8–9
differential diagnosis, 55 polyostotic, 9–11
dorsal defect of patella, 56f, 64 Fibular stress fractures, 103, 105f
MRI scans, 61f Field of vision (FOV), 165
normal variants, 63–68 Flexion teardrop fractures, 80, 82f
osteopoikilosis, 68, 68f, 69f Flexor hallucis longus (FHL), 208, 209f
posttraumatic, 55–63 Fluorosis, 152
avulsion injuries, 56, 57f, 58f Foot fractures, 108–109, 110f
cortical desmoid, 56–57, 58f, 59f Lisfranc’s, 108–109, 109f, 127
CT scans, 55 Forrester, Debbie, 111
fractures, 58, 62f FOV (field of vision), 165
PCL (posterior cruciate ligament), MRI for, 172–175, 174f Rib, unnecessary examinations, 2
Periostitis, 8–9, 13 Rogers, Lee, 76, 92–93
benign, 33–35, 34f Rolando’s fracture, 85, 87f
malignant tumors, 32–35 Rotator cuff, 183–186
appearance, 32–33 Rotatory fixation of atlantoaxial joint, 80, 81f
Pigmented villonodular synovitis (PVNS), 49–50, 51f
arthritis, 131, 131f, 132f
Pitt, Michael, 74 S
Pitt’s pit Sacral insufficiency fractures, 63, 64f
benign lesions, 74, 74f Sacroiliac (SI) joints
FAI, 74 arthritis, 111, 112t
Plasma cell granuloma, 28 HPT, 145f
Plasmacytoma, 21f, 49 osteoarthritis, 102f, 114f
Plastic bowing deformity of forearm, 92, 93f Sacroiliitis, 204f
Plica, 179, 179f Sarcoidosis, 124–126, 127f, 159, 163f
Polyostotic fibrous dysplasia, 9–11 Scheuermann’s disease, 136, 138f
Posterior cruciate ligament (PCL), MRI for, 172–175, 174f Schwannomas, 28, 53f
Posttraumatic lesions, 55–63 Seat belt fracture, 81, 83f
avulsion injuries, 56, 57f, 58f Secondary osteoarthritis, 113
cortical desmoid, 56–57, 58f, 59f Secondary osteoporosis, 139
CT scans, 55 Shoulders
fractures, 58, 62f AVN, 135f
geodes, 57, 60f, 61f MRI
MRI, 55 acromion, 185, 186
myositis ossificans, 55, 55f, 56f, 57f anatomy, 182, 183f
os odontoideum, 65–68, 67f biceps tendon, 189–190, 189f, 190f
osteogenic sarcoma, 56f bony abnormalities, 185–186
osteophytosis, 57–58 Buford complex, 187–189, 189f
pseudocyst of humerus, 66f coronal images, 182, 183f
pseudodislocation of shoulder, 62f, 63f FSE, 182
from trauma, 57 glenoid labrum, 170f, 186–189, 187f, 188f, 190f
Primary lymphoma of bone, 45, 46f Hill-Sachs lesion, 186, 187f
Primary osteoarthritis, 113–114, 113f, 114f imaging protocols, 182–183
Primary osteoporosis, 139 impingement syndrome, 183
Pseudo-Bennett’s fracture, 85 os acromiale, 186f
Pseudo-Charcot joint, 125f, 126–127, 128f, 129f Parsonage-Turner syndrome, 191–192, 191f
Pseudocyst of calcaneus, 72f partial tears, 185, 186f, 188f
Pseudocyst of humerus, 66f quadrilateral space syndrome, 191, 191f
Pseudogout. See Calcium pyrophosphate dihydrate crystal rotator cuff, 183–186
deposition disease suprascapular nerve entrapment, 190–191
Pseudotumor of hemophilia, 28 supraspinatus tendon, 183–184, 184f, 185, 185f
Psoriatic arthritis, 118, 118f, 119, 119f, 120f tendinosis, 184f
CT, 120f osteoarthritis, 112f
PVNS. See Pigmented villonodular synovitis rheumatoid arthritis, 117b, 117f
Pyknodysostosis, 148–149, 149f SI joints. See Sacroiliac joints
Pyoarthritis, 26f Sickle cell disease, 147, 147f
MRI, 225f
Sinus
Q CT scan, 2
Quadrilateral space syndrome, 191, 191f sinusitis, 1–2
unnecessary examinations, 1–2
Waters’ view, 2f
R Sinus tarsi syndrome, 213–214, 215f, 216f
Radial tears, 169f Skull
Radiology, 1 CT scan, 1
The Radiology of Emergency Medicine (Harris and fracture, 2f
Harris), 76 MRI, 1
Radiology of Skeletal Trauma (Rogers), 76 unnecessary examinations, 1
Radionuclide scans, metastatic bone surveys, 3–4 SLE (systemic lupus erythematosus), 124
Reiter’s syndrome, 117, 118, 119, 121, 121f Slipped capital femoral epiphysis, 163, 164f
Renal cell carcinoma, 22f Smith’s fracture, 81, 92, 93f
disc disease, 204f Soft tissue tumors, 49–54
Renal osteodystrophy, 147 atypical synovial cysts, 52f
Resnick, Don, 111 Baker’s cysts, 50–52
Rheumatoid arthritis, 115, 115b, 116f hemangiomas, 50, 51f
CPPD, 115 hematomas, 35f, 52–53
of hip, 116f hemorrhagic cysts, 41f
juvenile, 127–128, 129f misdiagnosis of, 52–53
migration routes, 116f PVNS, 49–50, 51f
secondary degenerative disease, 117f schwannomas, 28, 53f
of shoulder, 117b, 117f synovial osteochondromatosis, 49–50, 50f