-i-
`
`Smith & Nephew Ex. 1069
`IPR Petition - USP 9,295,482
`
`

`
`
`
`MAGNETIC RESONANCE IMAGING
`
`OF THE KNEE
`
`Jerrold H. Mink, M.D.
`Director of Musculoskeletal Radiology
`Cedars—Sinai Medical Center
`Assistant Clinical Professor of Radiology
`University of California at Los Angeles
`School of Medicine
`Los Angeles, California
`
`Murray A. Reicher, M.D.
`Staff Radiologist
`Mercy Hospital and Medical Center
`San Diego, California
`Assistant Clinical Professor of Radiology
`University of California at Los Angeles
`School of Medicine
`Los Ange les, California
`
`John V. Crues III, M.D.
`Santa Barbara Cottage Hospital
`Santa Barbara, California and
`Assistant Clinical Professor of Radiology
`University of California at Los Angeles
`School of Medicine
`Los Angeles, California
`
`With a Special Contribution by
`
`James M. Fox, M.D.
`Associate Medical Director
`
`Center for Disorders of the Knee
`Southern California Sports Medicine and Orthopedic Medical Group
`Van Nuys, California
`
`Raven Press § New York
`
`-ii-
`
`

`
`Raven Press,
`
`I 185 Avenue of the Americas. New York. New York W036
`
`I98"? by Raven Press Books, Ltd. All rights
`*3
`reserved. This book is protected by copyright. No part of it may be
`reproduced. stored in a retrieval system. or transmitted, in any fonn or by any
`means, electronical, mechanical, photocopying, or recording. or otherwise.
`without the prior written permission of the publisher.
`
`Made in the United States of America
`
`Mink. Jerrold H.
`Magnetic resonance imaging of the knee.
`
`Includes bibliographies and index.
`2. Magnetic
`1. Knee—Diseases—Diagnosis.
`resonance imaging.
`1. Reicher, Murray A.
`ll. Crues.
`John V.
`III. Title.
`[DNLM:
`I. Knee-—patl1o]ogy.
`2. Nuclear Magnetic Resonance—diagnostic use.
`WE 870 M665m]
`RC95l.M5l5
`I98".-’
`ISBN (J-88167-332-3
`
`8643229
`
`617K582
`
`The material contained in this volume was submitted as previously
`unpublished material. except in the instances in which credit has been given to
`the source from which some of the illustrative material was derived.
`Great care has been taken to maintain the accuracy of the information
`contained in the volume. However. neither Raven Press nor the editors can be
`held responsible for errors or for any consequences arising front the use of the
`infomtation contained herein.
`Materials appearing in this book prepared by individuals as part of their
`official duties as U.S. Government employees are not covered by the above-
`mentioned copyright.
`
`9876543
`
`—iii—
`
`-iii-
`
`

`
`'
`joint disorders,
`tella, effusions, intra—articular loose bodies, osteochondral
`fractures, tibial plateau fractures, plicae syndromes, pop-
`liteal artery aneurysms, synovial cysts, and-tumors. Al-
`though there are not yet sufficient data to state unequivocally
`the role of MRI in evaluating most of these conditions, our
`clinical experience‘ and the few series thus far reported in
`the literature support the hypothesis that MRI will become
`a primary technique for evaluating many of these abnor-
`malities. The goal of this chapter is to review the patho-
`physiology of these disorders and to point out the proven
`and potential roles for MRI.
`
`CHAPTER 7
`A Spectrum of‘ Knee ]oint Disorders
`,_____________________,_,
`tal narrowing, marginal osteophytes, juxta-articular sclerosis,
`and subchondral cysts. Sclerosis usually occurs in the tibia
`or in both the femur and tibia, but rarely in the femur alone
`(70). Subchondral cysts usually are small and predominate
`in the tibia (70). Recently, single-photon—emission com-
`puted tomography (SPECT) has been shown to be more
`sensitive than either standard radiography or planar scintig—
`raphy for detecting osteoarthritis of the knee (23).
`MRI offers a unique potential for evaluating osteoarthritis
`because it is the only technique that allows visualization of
`the full width of the articular cartilage. Normal articular
`cartilage is seen as a thin rim of moderate signal intensity
`bordering the articular surfaces of the femur,
`tibia, and
`patella (see Chapter 3). Articular cartilage is easily visu-
`alized in contrast to the neighboring low-intensity cortical
`bone and menisci (10,19,37,44,62,63,65 ,82). Visualization
`of articular cartilage using MRI is potentially superior to
`that with arthrography, because arthrograms may be difficult
`to perform in older patients and allow depiction of only the
`surface of cartilage tangential to the X—ray beam. There
`have been published reports of knee joint articular cartilage
`thinning and erosion seen on MR images (10,19,37,62)
`(Figs. 7.1-7.3). MRI has been found to be accurate in
`delineating chondral defects as shallow as 1 mm in cadaver
`knees and patients (18).
`MRI may eventually prove useful for evaluating aging of
`articular cartilage (19). Degeneration of meniscal cartilage
`in the absence of a tear has already been demonstrated (82),
`as has degeneration and desiccation of intervertebral discs.
`The signal yielded by cartilage is dependent on the inter-
`action of water with negatively charged chondroitin sulfate
`and keratin sulfate polysaccharide chains (19,42). Altera-
`tions in these proteoglycans may be associated with changes
`in signal (42).
`
`SYNOVIAL DISORDERS
`There are numerous conditions that may cause synovial
`inflammation or thickening, including rheumatoid arthritis,
`gout, calcium pyrophosphate dihydrate deposition disease,
`pigmented villonodular synovitis, infection (tuberculous or
`fungal), hemophiliac arthropathy, synovial chondromatosis,
`synovial
`hemangioma,
`and
`lipoma
`arborescens
`(3,6,30,56,70). Both standard radiography and arthrogra-
`phy are limited for evaluation of these disorders, because
`the primary site of disease, the synovium, cannot be visu-
`alized directly using either method. MRI, which yields phys-
`iological and anatomic information, offers a potential tool
`for both early diagnosis and assessment of the response to
`
`OSTEOARTHRITIS
`The knee is the most common site of osteoarthritis (70).
`Patients generally present with symptoms of pain, tender-
`ness, diminished range of motion, warmth, effusions, or
`synovial cysts. As osteoarthritis progresses, a varus de-
`formity, soft-tissue atrophy, and knee joint instability fre-
`quently develop (70). Osteoarthritis begins with articular
`cartilage degeneration and erosion. Typical pathological
`changes include cartilage fibrillation, denudation of osseous
`surfaces, subchondral cystic lesions, and osteophytosis. Be-
`cause osteoarthritis is not truly an inflammatory disease,
`some authors advocate use of the term “osteoarthropathy.”
`The three compartments of the knee (the medial and lateral
`femorotibial compartments and the patellofemoral com-
`partment) are usually involved unevenly, i.e., the medial
`femorotibial and patellofemoral compartments are involved
`most frequently, and often both are involved. Involvement
`of the patellofemoral compartment usually predominates along
`the lateral facet of the patella. Isolated patellofemoral dis-
`ease occurs secondary to calcium pyrophosphate dihydrate
`deposition disease or hyperparathyroidism. Lateral com-
`partment osteoarthritis is less frequent and is rarely asso-
`ciated with disease of the patellofemoral compartment.
`Early disease of the articular cartilage is not visible on
`standard radiographs, but may be seen on arthrograms. Ar-
`thrography in early osteoarthritis depicts articular cartilage
`erosion, fragmentation, and imbibition of contrast material
`(6,30,70). Associated findings include meniscal degenera-
`tion, subchondral cysts, and popliteal cysts. Advanced dis-
`ease is manifested on standard radiographs as compartmen-
`
`______._____._____
`‘Over 2,500 MRI examinations of the knee have been performed at
`Cedars-Sinai Medical Center and the UCLA Center for the Health Sciences,
`Los Angeles, California, since July 1984.
`
`
`
`-123-
`
`

`
`124 / Cl-IAPTER7
`
`
`
`FIG. 7.1. Medial femorotlbial osteoarthritis. Magnified sagittal
`spin-echo image (TE = 20 msec, TR = 800 msec) of medial
`femorotibial compartment. Patient had undergone previous
`meniscectomy. Note thinning of tibial and femoral articular
`cartilage (arrows) and small erosions of underlying bones
`(curved arrows).
`
`therapy. As a general rule, obtaining both Tl— and T2-
`weighted images is recommended in evaluating patients with
`synovial disease. There have been only a few reports of the
`use of MRI for synovial disorders (19,37,44,69,80,83,90).
`The following is a brief review of the conditions that have
`been evaluated with MRI.
`
`
`
`FIG. 7.2. Patellofemoral osteoarthritis. Axial spin-echo pulse
`sequence with TE = 25 msec and TR = 800 msec. Femoral
`articular cartilage is narrowed over small osteophyte (curved
`arrow). Phase-encoding artifact (small arrows) is due to pop-
`liteal artery pulsation.
`
`-124-
`
`
`
`FIG. 7.3. Patellofemoral osteoarthritis. Sagittal spin-echo pulse
`sequence with TE = 25 msec and TR = 800 msec. Articular
`cartilage of patella and femur is absent, joint space is mark-
`edly narrowed, and osteophytes (arrows) are seen emanat-
`ing from patella and femur.
`
`
`
`Rheumatoid Arthritis
`
`Small groups of patients with rheumatoid arthritis have
`been evaluated with MRI (l9,37 ,69,80,90) (Fig. 7 .4). Find-
`ings that are known to occur in adult-onset rheumatoid ar-
`thritis include synovial thickening, articular erosions, sub-
`luxations, subchondral cysts, effusions, popliteal cysts, and
`osteoporosis. In contrast to the adult-onset form, in which
`articular erosion occurs early, joint space narrowing in ju-
`venile rheumatoid arthritis (JRA) is a late finding. Growth
`disturbances, such as metadiaphyseal constriction and bal-
`looning of the distal femoral and proximal tibial epiphyses,
`are prominent features of IRA (70). Other findings common
`in IRA include osteoporosis, flattening of the femoral con-
`dyles, widening of the intercondylar notch, “squaring” of
`the inferior margin of the patella, and marginal or central
`osseous erosions (35,70). Arthrography in patients with
`rheumatoid arthritis may be painful and generally has been
`applied only for delineation of popliteal cysts. Arthrographic
`findings are nonspecific and include enlargement of the joint
`cavity, nodular irregularity or corrugation of the synovial
`membrane, subchondral cysts, intra-articular filling defects,
`and lymphatic filling (70). MRI may painlessly depict sy-
`novial hypertrophy, cartilaginous erosions, and subchondral
`cysts. In addition, unlike arthrography or arthroscopy, MRI
`can disclose bone infarcts and delineate noncommunicating
`synovial cysts. Menisci smaller than normal have also been
`reported as MRI
`findings
`in JRA (80)
`(Fig. 7.5A)
`(see Chapter 8). Determination of the metabolic activity
`of synovial disease may be within the realm of MR
`spectroscopy.
`
`-124-
`
`

`
`SPECTRUM OF KNEE ]OINT DisoRDERs / 125
`
`
`
`
`
`FIG. 7.4. JRA in two different patients. A: Sagittal spin-echo
`sequence with TE = 60 msec and TR = 2,000 msec reveals
`thickened synovium (open arrows), thinned femoral articular
`cartilage (small black arrows), and bony erosion (curved ar-
`
`Pigmented Villonodular Synovitis
`
`Pigmented Villonodular synovitis (PVNS) is a monoar—
`ticular synovial disease that affects either the joints or tendon
`sheaths in young adults (27,30,34,48,56,70,89). The knee
`is involved in 80% of articular disease. Other joints that
`may be involved, in decreasing order of frequency, include
`
`
`
`FIG. 7.5. JRA. T2-weighted sequence (TE = 60 msec,
`TR = 2,000 msec)
`reveals low-intensity nodular pannus
`(curved arrows) outlined by high-intensity synovial fluid (small
`arrows). The posterior fluid appears brighter than that in the
`suprapatellar bursa because a posteriorly placed surface coil
`was used.
`
`rows). B: Another patient (TE = 20 msec, TR = 800 msec)
`demonstrates small meniscus (arrow), osseous erosions (curved
`arrows), and “balloon” distal femoral epiphysis.
`
`the hip, ankle, hands, feet, and elbow. The disease has two
`forms: diffuse and nodular. In the diffuse form, patients
`typically present with pain, decreased range of motion, joint
`swelling, warmth, and tenderness. The nodular form may
`cause knee locking, but usually is less symptomatic than
`the diffuse form, and may be clinically mistaken for a men-
`iscal tear (34,48,89). Hemorrhagic effusions are common;
`joint aspiration therefore usually yields dark bloody or Xan-
`thochromic fluid.
`Pathological examination reveals fibrous synovial prolif-
`eration and infiltration by histiocytes, multinucleated giant
`cells, and hemosiderin—laden macrophages (27,34,56). The
`cause is unknown, but PVNS is postulated to be a benign
`inflammatory reaction to some unidentified agent (89).
`Typical radiographic findings in PVNS include a non-
`calcified synovial soft—tissue mass, subchondral and juxta-
`articular erosions and cysts, normal bone density, normal
`joint space, and lack of hypertrophic bone formation (27,70).
`Arthrography has been reported in both the diffuse and
`nodular forms (27,30,34,48,70). In the diffuse form, ar-
`thrograms disclose an enlarged joint cavity and irregular and
`nodular synovial masses with villous projections, an ap-
`pearance that may be mimicked by many synovial processes.
`In the nodular form, a circumscribed mass may be seen in
`the infrapatellar region, a finding that may be missed if
`coned—down views of the menisci omit this area.
`Only a single case of MRI of PVNS has thus far been
`reported (37). In that case,
`the clinical diagnosis was a
`meniscal tear, and standard radiographic findings were in-
`terpreted as normal. Sagittal T1-weighted images revealed
`a well-circumscribed mass within the infrapatellar fat pad
`(Fig. 7.6). The lesion was not seen at subsequent arthros-
`
`
`
`-125-
`
`-125-
`
`

`
` fi?_
`
`126 / CHAPTER 7
`
`
`
`FIG. 7.6. PVNS. Sagittal T1-weighted image (TE = 25 msec,
`TR = 500 msec) reveals homogeneous, well-circumscribed
`mass of intermediate signal intensity (arrow) within the in-
`frapatellar fat pad. (From ref. 37.)
`
`copy, but because of the MRI findings, a local incisional
`arthrotomy was performed, and the mass was located and
`excised. This case illustrates a potentially valuable role for
`MRI in evaluating lesions of the infrapatellar fat pad, an
`area difficult to examine by arthrography or arthroscopy. In
`theory, PVNS should yield low signal
`intensity on T2-
`weighted images because of the paramagnetic effect of he-
`mosiderin. If this hypothesis proves correct, MRI may allow
`PVNS to be distinguished from other lesions of synovial
`origin.
`Successful treatment of PVNS is dependent on local ex-
`cision, which is more easily accomplished with the nodular
`form. The diffuse form tends to recur in approximately 50%
`of patients (27) despite synovectomy.
`
`Hemophilia
`Knee joint arthropathy occurs frequently in patients with
`hemophilia secondary to repeated massive and often sub-
`clinical hemarthroses. The radiographic findings, which may
`be virtually indistinguishable from those of JRA, are known
`to underestimate the extent of disease (78). MRI may iden-
`tify hemarthroses, subchondral cysts, synovial inflamma-
`tion, and periarticular fibrosis in patients with hemophilia
`(44,90). Chronic periarticular and subchondral fibrosis yields
`low signal on T1— and T2-weighted images. Subchondral
`cysts and areas of synovial inflammation yield low signal
`on T1-weighted images, but high signal on T2—weighted
`images. Resolving hemarthroses may yield high signal on
`both T1— and T2-weighted images. In chronic hemophiliac
`arthropathy, the synovium is scarred and thickened, result-
`ing in low signal surrounding the knee (Fig. 7.7).
`
`FIG. 7.7. Hemophilia. Sagittal T1-weighted sequence reveals
`low-signal-intensity periarticular fibrosis. (Courtesy of David
`Stoller, M.D., Department of Radiology, University of Cali-
`fornia, San Francisco.)
`
`CHONDROMALACIA PATELLA
`
`Chondromalacia patella is characterized by premature de-
`generation and erosion of the patellar cartilage and is the
`most frequent cause of knee pain in adolescents and young
`adults (6,30,70). The disorder is caused by a mechanical
`abnormality of the patellofemoral articulation that results in
`abnormal stress as the patella glides over the femur. Pro-
`posed predisposing conditions include hypoplasia of the lat-
`eral femoral condyle, flattening of the posterior ridge of the
`patella, patella alta or baja, genu valgum, and lateral tilt of
`the patella (6). A recent report, however, suggests that there
`is no association between patellar malalignment and chon-
`dromalacia (28). Patients usually present with pain and crep-
`itus over the anterior portion of the knee, exacerbated by
`knee flexion or stress such as climbing stairs. The disorder
`may be difficult to differentiate clinically from a meniscal
`tear or plica syndrome. Early pathological changes include
`edema, softening, and fissuring of the patellar cartilage sur-
`faces (6,1l). As the disease progresses, the cartilage be-
`comes irregular, but the width may actually thicken sec-
`ondary to edema (6). In late stages, the patellar cartilage
`
`-126-
`
`

`
`FIG. 7.8. Chondromalacia patella. Sagittal T1 -weighted images
`in two different patients. A: Patellar cartilage is thinned and
`has concave contour (small arrows); femoral cartilage is also
`
`wears thin, and the adjacent femoral cartilage may also be
`eroded. The medial facet of the patella and the junction of
`the medial and odd facets are the most frequently involved
`sites (28,70).
`Standard radiography is insensitive for detecting Chon-
`dromalacia. The most consistent abnormality is osteoporosis
`of the patella (70). Arthrography may be employed to outline
`
`narrowed (curved arrow). B: Patellar cartilage is absent in
`phytes are present
`more advanced case; small patellar osteo
`(arrows).
`
`the patellar cartilage, but can be inaccurate, with false-
`positive findings in up to 14% and false—negative findings
`in up to 55% of patients (12,39,46,56,85). The combination
`of CT and arthrography is more accurate than arthrography
`alone (12,l3,50,68), but
`is costly and time—consuming.
`SPECT may be very sensitive in detecting patellofemoral
`disease, but is nonspecific.
`
`FIG. 7.9. Joint effusion. Parasagittal T1-weighted image (A)
`with TE = 28 msec and TR = 500 msec reveals moderate-
`intensity effusion in suprapatellar bursa (arrow) and behind
`
`B: With T2-weighting, fluid yields
`meniscus (curved arrow).
`high signal and is seen in same locations. (From ref. 37.)
`
`
`
`-127-
`
`-127-
`
`

`
`lZ8 / CHAPTER7
`
`
`
`FIG. 7.10. Joint effusion with fat—serum—sediment levels in
`suprapatellar bursa. Sagittal spin-echo pulse sequences with
`TE = 20 msec and TR = 2,000 msec (A) and TE = 60 msec
`and TR = 2,000 msec (B). Fat droplets (small arrows) float
`
`on serum (open arrow), which is seen layering above hem-
`orrhagic sediment (curved arrow). The‘ cause of effusion was
`unsuspected, radiographically occult “dent” fracture of femoral
`condyle (white arrow).
`
`The normal multiplanar anatomy of the patellofemoral
`articulation has been demonstrated with MRI (see Chapter
`
`3) (65). MRI appears capable of noninvasively depicting
`patellofemoral anatomy as well as CT—arthrography. The
`patellar articular cartilage is best visualized on axial images.
`Although one report indicated that MRI can distinguish be-
`tween swollen or irregular cartilage or absence of cartilage
`and bony patellar defects in patients with chondromalacia
`patella (91), the accuracy of MRI in diagnosing this con-
`dition is yet to be determined in a controlled study. MRI is
`certainly less costly and less time-consuming than CT-ar-
`thrography or arthroscopy.
`We have observed approximately a dozen cases in which
`MRI in patients with patellofemoral pain has depicted thin—.
`ning of the patellar cartilage suggesting chondromalacia (Fig.
`7.8). Only one case has thus far been proven surgically,
`however. In evaluating patients with patellofemoral pain,
`MRI has the capability of noninvasively imaging other dis-
`orders that may be clinically considered in the differential
`diagnosis, including meniscal tears, thickened plicae, and
`patellofemoral osteoarthritis. Because chondromalacia pa-
`tella is generally treated nonoperatively, MRI may obviate
`diagnostic arthroscopy by excluding other diagnostic con-
`siderations.
`
`on both Tl- and T2—weighted images (44). The site and
`composition of synovial fluid provide clues to the underlying
`cause. For example, the presence of fat within synovial fluid
`indicates the presence of an acute fracture (Figs. 7.10 and
`7.11). Because MR images are generally acquired with the
`patient supine, fat droplets, which yield high signal on both
`
`
`
`JOINT EFFUSIONS
`
`Normal and acutely hemorrhagic synovial fluid yields
`moderate signal on T1-weighted MR images and high signal
`on T2—weighted images (7,8,10,3l,37,44,63,64,83) (Fig.
`7.9). Resolving hemorrhagic effusions may yield high signal
`
`FIG. 7.11. Osteochondral fracture with‘associated fat—fluid
`level in joint effusion. Sagittal spin-echo pulse sequence with
`TE = 20 msec and TR = 800 msec. High-intensity fat glob-
`ules (arrow) float on synovial fluid within suprapatellar bursa
`(white arrow). Small osteochondral fracture is present in an-
`terior aspect of femoral condyle (curved arrow).
`
`
`
`-128-
`
`-128-
`
`

`
`SPECTRUM or KNEE ]o1Nr DISORDERS / 129
`
`FIG. 7.12. infrapatellar bursitis. T1—weighted im-
`ages (TR = 500 msec, TE = 28 msec) reveal
`moderate-intensity fluid (arrows) in infrapatellar
`bursa on sagittal (A) and axial (B) images. C:
`Larger effusion of infrapatellar bursa (arrows) seen
`in another patient (TE = 28 msec, TR = 500
`msec). (From ref. 37.)
`
`T1— and moderately T2—weighted images, tend to rise to the
`ventral aspect of the suprapatellar bursa. A serum—sediment
`level may be seen if the patient lies still in the supine position
`for several minutes prior to imaging (Fig. 7.10). Infrapa—
`tellar bursitis may be easily diagnosed by the presence of
`an effusion localized to the infrapatellar bursa (Fig. 7.12).
`MRI may thus distinguish this entity from other clinically
`considered diagnoses such as PVNS or a patellar tendon
`injury. Demonstration of synovial fluid surrounding the os-
`teochondral fragment in osteochondritis dissecans suggests
`that the fragment is loose (see Chapter 6), a conclusion that
`has important prognostic and therapeutic implications (see
`Fig. 6. 10B). Therefore, although standard radiography may
`be up to 90% accurate in determining the presence of a
`
`tappable fluid collection in the suprapatellar bursa (77), MRI
`offers the additional advantages of detecting effusions in
`other sites, characterizing the composition of the fluid, and
`determining the underlying cause. Furthermore, small knee
`joint effusions that do not distend the suprapatellar bursa
`may be missed on physical examination or standard ra-
`diography, but detected using MRI (37).
`
`POPLITEAL CYSTS
`
`Popliteal cysts, or Baker cysts, are bursal collections of
`synovial ‘fluid that usually occur in the semimembranosus— A
`gastrocnemius bursa, but can occur in other locations, in-
`cluding the bursa beneath the popliteal tendon,
`the bursa
`
`
`
`-129-
`
`-129-
`
`

`
`130 / CHAPTER 7
`
`between the lateral head of the gastrocnemius and the distal
`biceps femoris muscle, and the tibiofibular joint cavity
`(6,30,36,47,70). Although 30%—50% of adults have been
`found to have communications between the knee joint and
`the semimembranosus-gastrocnemius bursa,
`these com-
`munications are seldom seen in younger people (47), sug-
`gesting that they are acquired through trauma or degener-
`ation of the posterior joint capsule. Cysts that may or may
`not be clinically palpable are typically associated with chronic
`knee joint effusions. In adults, they usually are sequelae of
`other knee joint abnormalities. Common causes include in-
`temal derangement (meniscal or cruciate ligament tear or
`loose body), osteoarthritis, and rheumatoid arthritis or JRA.
`Less common causes include other chronic arthritides, chon—
`dromalacia, granulomatous synovitis, osteochondritis dis-
`secans, PVNS, and septic arthritis. Idiopathic popliteal cysts
`occur in children. Occasionally, popliteal cysts may rupture,
`causing synovial fluid to dissect between the soleus and
`gastrocnemius muscles. Local pain and inflammation result,
`mimicking thrombophlebitis. Further complicating the clin-
`ical assessment in these cases is the fact that cysts may
`directly compress the popliteal vein and cause deep venous
`obstruction (84). Other conditions that may clinically mimic
`popliteal cysts are benign lipomas, popliteal artery aneu-
`rysms, malignant tumors, and chronic hematomas (75).
`Traditional methods for imaging popliteal cysts include
`ultrasonography and arthrography (6,30,38,49,70,84,88).
`Sonography is a safe, painless, and inexpensive method,
`but it may miss small cysts and cannot disclose the under-
`lying intra—articular cause (38,70). Arthrography is more
`sensitive than ultrasound (38), but it may miss cysts that
`
`
`
`do not readily communicate with the joint (70,88) and has
`been reported to have a sensitivity of under 50% (88). More
`recently, CT has been advocated and found to be superior
`to arthrography, especially for detecting cysts in unusual
`locations (45,75).
`There have now been several reports of popliteal cysts
`revealed by MRI (37,80,83). Simple cysts are well circum-
`scribed and yield low—to—moderate signal on Tl—weighted
`images and high signal on T2-weighted images. A dissecting
`cyst has been reported with high signal on both Tl- and T2-
`weighted images (83). In our experience, MRI has revealed
`underlying meniscal tears (Fig. 7.13), rheumatoid arthritis
`(Fig. 7.14), cysts in atypical locations (Figs. 7.14 and 7.15),
`and cysts containing osteochondral fragments (Fig. 7.16).
`MRI offers several advantages in evaluating popliteal cysts.
`It is as painless and safe as ultrasound, yet as capable of
`disclosing underlying intra—articular disorders as arthrog-
`raphy. Based on data derived from CT, MRI should prove
`more sensitive than arthrography in detecting atypical cysts
`(75). An understanding of knee joint anatomy and the im-
`aging characteristics of synovial fluid allows the diagnosis
`of a popliteal cyst to be made with complete specificity with
`MRI. Finally, detection of associated deep venous compres-
`sion is entirely within the realm of MRI (21,29,5l,87).
`
`FRACTURES
`
`Trauma to the knee joint may result in tibial plateau
`fractures, stress fractures, and chondral or osteochondral
`fractures. These may be radiographically occult and can
`
`
`
`FIG. 7.13. Popliteal cyst. Sagittal spin-echo pulse sequence
`with TE = 25 msec and TR = 800 msec. Large fluid collec-
`tion of moderate signal intensity (arrows) is present in sem-
`imembranosus-gastrocnemius bursa. Posterior horn of the
`medial meniscus is truncated (curved arrow).
`
`FIG. 7.14. Ascending popliteal cyst in JRA. Sagittal spin-echo
`pulse sequence with TE = 20_ msec and TR = 2,000 msec.
`With this pulse sequence, synovial fluid yields moderate sig-
`nal. Large joint effusion is seen (small arrows), and there is
`an extracapsular collection above the gastrocnemius muscle
`origin (open arrows) that is filled with dark nodular pannus.
`
`-130-
`
`-130-
`
`

`
`FIG. 7.15. Synovial cyst. Coronal spin-echo pulse sequence
`with TE = 25 msec and TR = 800 msec. Moderate-intensity
`synovial cyst (white arrows) is present adjacent to tibiofibular
`articulation (black arrow).
`
`clinically masquerade as a meniscal or ligamentous injury.
`The following sections discuss potential and proven capa-
`bilities of MRI for assessing fractures.
`
`Tibial Plateau Fractures
`
`Thorough assessment of the degree of comminution and
`displacement of fragments is required to properly plan ther-
`apy for fractures of the tibial plateau. Most fractures involve
`the lateral tibial plateau and result from an abduction or
`valgus strain combined with an axial or compression force
`(59). The anterior portion of the lateral femoral condyle is
`thus driven into the lateral tibial plateau. Most authors agree
`that open reduction and internal fixation are indicated when
`there is associated instability of the anterior cruciate or me-
`dial collateral ligament (20,57). The amount of acceptable
`depression of the articular surface remains debatable (1-7
`mm) (20,57). Late complications include joint instability,
`valgus malalignment, and osteoarthritis (3,59).
`Standard radiography often underestimates the degree of
`comminution and gives misleading impressions of the de-
`gree of plateau depression (59). For this reason, CT with
`sagittal or coronal reformatting of images has been advo-
`cated (59). Planar tomography has also been employed to
`delineate tibial plateau fractures, which are not evident on
`standard radiographs (3,4). Arthrography and arthrotomog—
`raphy may be employed to evaluate the integrity of the
`proximal tibial articular surface (3 ,70).
`MRI has also been applied to detect tibial plateau fractures
`in patients presenting with acute knee pain (41,81). Frac-
`tures may appear as linear segments or homogeneous areas
`of decreased signal intensity on T1-weighted images. The
`
`FIG. 7.16. Popliteal cyst. Sagittal spin-echo pulse sequence
`withTE = 25 msec and TR = 800 msec.Withthis T1 —weighted
`sequence, synovial fluid yields moderate signal and is seen
`filling gastrocnemius-semimembranosus bursa (arrows). Os-
`teochondral fragments (curved arrows) within cyst yield high
`signal because true ossification results in formation of med-
`ullary bone that contains marrow fat. There is an associated
`meniscal tear (open arrow).
`
`
`
`FIG. 7.17. Tibial plateau fracture. Sagittal spin-echo pulse
`sequence with TE = 20 msec and TR = 800 msec. Although
`this was an old, healed comminuted fracture, linear low-sig nal
`regions are seen along old fracture lines (arrows). Small step-
`off in articular cartilage (open arrow) of tibial plateau is readily
`apparent.
`
`
`
`-131-
`
`-131-
`
`

`
`j
`
`132 / CHAPTER7
`
`
`
`FIG. 7.18. Tibial plateau fracture. Sagittal spin-echo
`pulse sequence with TE = 28 msec and TR = 777
`msec. Posterior tibial plateau fracture (arrows) re-
`sults in avulsion of the site of attachment of the
`posterior cruciate ligament (curved arrow).
`
`decreased signal from medullary bone along the fracture has
`been attributed to edema, acute hemorrhage, and/or a hy-
`pervascular response (10,81). In the acute or subacute set-
`ting, the fracture region may produce increased signal on
`T2—weighted images. Old fracture lines retain low signal
`with T1- and T2-weighting (Fig. 7.17). The normal adult
`physis may be seen as a low-signal line (see Chapter 8) and
`should not be mistaken for a fracture.
`
`MRI offers several potential advantages in imaging tibial
`plateau fractures. Like CT, MRI provides a computed tom-
`ographic view of the knee that may facilitate a three—di-
`mensional understanding of the degree of comminution and
`orientation of fracture fragments. The spatial resolution on
`axial MR images now approaches that of CT; sagittal or
`coronal MR images have better spatial resolution than re-
`formatted CT images. In addition, MRI can delineate the
`tibial articular cartilage (Fig. 7.17) and be used to assess
`associated disruptions of the menisci, cruciate ligments, or
`collateral ligaments (Fig. 7.18) (see Chapters 4 and 5). In
`
`
`
`
`
`FIG. 7.19. insufficiency fracture of the proximal tibia. Coronal
`spin-echo pulse sequence with TE = 25 msec and TR = 800
`msec. Nondisplaced fracture of proximal tibia is depicted as
`low—intensity line (arrow). Standard radiographic findings were
`initially normal, but later films demonstrated a typical insuf-
`ficiency fracture.
`
`FIG. 7.20. Osteochondral loose body. Sagittal T1 -weighted
`MR image reveals oval,
`low—intensity fragment
`(arrow)
`embedded in infrapatellar fat pad. Loose body was not de-
`tected at prior arthroscopy, but after MRI, localized incisional
`arthrotomy was performed and the loose body excised. (From
`ref. 62.)
`
`-132-
`
`-132-
`
`

`
`SPECTRUM OF KNEE JOINT DISORDERS / 133
`
`
`
`
`
`
`
`FIG. 7.21. Osteochondral bodies. A: Standard radiograph reveals fluffy
`calcification in posterior aspect of knee. The differential diagnosis in-
`cluded the possibilities of a loose body and synovial chondrosarcoma.
`B:Sagitta| T1 -weighted MRI (TR = 500 msec, TE = 28 msec) depicts
`two well-circumscribed, faceted,
`low-intensity osteochondral bodies
`(white arrows). C: Surgical specimen provides precise correlation.
`(From ref. 62.)
`
`patients presenting with normal radiographic findings and
`knee pain, MRI may detect fat droplets within synovial fluid,
`indicating the presence of an occult fracture (81). The major
`disadvantage of MRI is that cortical bone yields low signal
`because of its low proton density and extremely short T2.
`Although the contrast with adjacent medullary bone and soft
`tissues makes cortical bone visible on MR images, small
`fragments may not be as conspicuous as on CT. Although
`its current high cost probably precludes primary use of MRI
`for evaluating tibial plateau fractures, MRI may be valuable
`for evaluating patients suspected of