Magnetic Resonance Imaging, Vol. 12, No. 5, pp. 703-709, 1994 Copyright 0 1994 Elsevier Science Ltd Printed in the USA. All rights reserved 0730-725X/94 $6.00 + .@I
`
`0730-725X(94)E0017-V
`
`l Original Contribution
`
`MR FEATURES OF OSTEOARTHRITIS OF THE KNEE
`
`FELIX FERNANDEZ-MADRID, * ROBERT L. KARVONEN, * ROBERT A. TEITGE,~ PETER R. MILLER,$ AND WILLIAM G. NEGENDANK* Departments of *Internal Medicine (Rheumatology), torthopedic Surgery, and SRadiology, Wayne State University School of Medicine, Detroit, MI 48201, USA
`
`osteoarthritis (OA) of the knee was surveyed using weight-bearing radiographs
`A
`and MR imaging to compare the relative value of these methods in disease evaluation. Fifty-two patients with a
`clinical and radiological diagnosis of OA of the knee of relatively short duration (87%: 14 yr) were compared
`to a reference group of 40 age- and sex-comparable subjects with no knee symptoms. All patients had a complete
`history, physical examination, standard anterior-posterior and lateral weight-bearing radiographs, T1-weighted,
`and FLASH MR images in both knees. The prevalence of MRI abnormalities was significantly greater in patients
`with OA of the knee in all radiographic grades (Kellgren and Lawrence) compared to the reference subjects. Sig
`nificant differences were encountered for synovial thickening (OA, 73%; reference, OVo), synovial fluid (6OVo;
`7%), meniscal degeneration (52%; 7%), osteophytes (67%; 12%), and subchoadral bone involvement (65%; 7%),
`even in the patients at the mild end of the osteoarthritic spectrum, indicating the exquisite sensitivity of MRI com-
`pared with weight-bearing radiographs.
`
`Keywords: Osteoarthritis; MRI; Synovial thickening; Meniscus; Osteophytes; Subchondral bone.
`
`INTRODUCTION
`
`MATERIAL AND METHODS
`
`Study Subjects
`
`Osteoarthritis (OA) is the most prevalent form of ar- thritis and a major cause of disability in the world’s population. l-3 Structural changes in the osteoarthritic joint have been difficult to study because of the rela- tive insensitivity of conventional radiographs.4 Conse- quently, with few exceptions,4s5 the literature on early OA is largely limited to studies of experimentally in- duced disease in animal models.6*7 Magnetic resonance imaging (MRI) is an excellent means to study noninvasively the anatomy of the knee joint to diagnose internal derangements.8*9 MRI has recently been used to study patients with advanced OA of the knee.lOpl’ Its ability to document articular cartilage erosions in the knee has been documented by us using histopathologic correlations’* and by others using arthroscopy.13 The goal of this work was to study a group of pa- tients with OA of the knees ranging from the mild to the severe end of the spectrum to examine the propor- tions and types of abnormalities detected by MRI and compare these with weight-bearing radiographs.
`Fifty-two consecutive patients seen at the Rheuma- tology Clinic at Wayne State University who met the criteria for classification of idiopathic OA of the knee developed by the American College of Rheumatol- ogy14 were studied. Radiographic” as well as clinical criteriaI were met by 70% of the OA patients. The re- maining patients with radiographic Grades 0 or 1, met four to six ACR criteria in the clinical classification tree derived from recursive partitioning analysis, where a score of 4 confirms the diagnosis of OA to a specific- ity of 88%. l4 These criteria include knee pain, morn- ing stiffness ~30 min, crepitus, bony tenderness, bony enlargement, no palpable warmth, erythrocyte sedi- mentation rate ~40 mm/h (Westergren), rheumatoid factor < 1:40, synovial fluid clear or viscous, and white blood cell count <2,OOO/cubic mm.14 All patients had a complete history and physical examination, and stan- dard anterior-posterior and lateral weight-bearing ra- diographs of the knees. Criteria for inclusion of patients in the study were: (1) clinical diagnosis of OA RECEIVED 8/30/93; ACCEPTED 2/9/94. PhD, Center For Rheumatic Diseases, Hutzel Hospital, 4707 Address correspondence to Felix Fernandez-Madrid, MD, St. Antoine, Detroit, MI 48201. 703
`
`-703-
`
`Smith & Nephew Ex. 1066
`IPR Petition - USP 9,295,482
`
`group of patients with idiopathic
`

`
`17
`
`ms, and four acquisitions were obtained in a 256 x 256 matrix with a 20 x 20 cm field-of-view. Using these parameters we have been able to obtain excellent con- trast between articular cartilage and adjacent struc- tures16 as well as accurate measurements of articular cartilage thickness. I2 Although synovial thickening was readily apparent in
`-weighted images, fast-scan (FLASH)” gradient-recalled echo images were also obtained in contiguous 3 mm sagittal slices through- out the knee to help differentiate synovial thickness from synovial fluid. The TR was 250 ms, TE 10 ms, flip angle
`
`40”,
`
`T,
`
`704 Magnetic Resonance Imaging 0 Volume 12. Number 5, 1994 of the knee; (2) pain elicited by activity or prolonged weight-bearing and relieved with rest; (3) knee stiffness lasting less than 15 min on awakening or after periods of inactivity (gelling phenomenon); (4) physical find- ings ranging from no gross joint deformities and restric- tion of motion limited to decreased flexion due to pain on passive motion in patients in the milder end of the spectrum to bulky, bony deformities with restriction of flexion and flexion contracture in those with more severe disease; (5) a weight-bearing radiographic gradeI from 0 to 4; (6) normal CBC and ESR and negative RF by latex fixation; (7) no history of trauma to the knees; and (8) no evidence of crystal-induced dis- ease or any kind of active inflammatory disorder. When synovial effusion was found on physical examination (11 patients), arthrocentesis was performed and the cell count ranged from 19 to 920 cells/cubic mm, predom- inantly mononuclear cells. Examination of the syno- vial fluid was in all cases negative for crystals. MRI was obtained in both knees of all patients in the OA and reference groups. Ninety-seven knees in 52 patients were analyzed and compared with 80 knees in 40 reference subjects. From a total of 104 knees, 7 were excluded from analysis because they previously had blunt trauma and/or had been subjected to arthro- scopic surgery and might have had OA secondary to trauma. Five patients with OA had only one knee stud- ied by MRI. The reference group consisted of 40 volunteer sub- jects with neither symptoms nor history of knee prob- lems. The physical examination of the knees in all reference subjects was normal. Conventional radio- graphs of the knees in 36 of these subjects showed no abnormalities. In four reference subjects radiographs could not be obtained. Reference subjects were re- cruited to achieve similarity of mean age, age range, and sex distribution. This group served to establish the range of apparent normal findings in MRI and as a ref- erence group to prevent overreading of abnormalities in patients with OA. MRI studies from study subjects and reference subjects were mixed and read blindly by two observers (W.G.N. and F.F-M.) by consensus. The patients are a sample with relatively early OA (disease duration 87%, 54yr; 70%, 53 yr; 48% 12 yr). The demographic features of patients and reference subjects are included in Table 1.
`-weighted spin-echo images obtained in 3 mm slices with 1.5 mm gaps. The TR was 600 ms, TE Table 1. Demographic features of study groups Age* Weight* Height * n M:F (yr) (pounds) (inches) Reference 40 15:25 49 i 15 156 f 34 67 f 3.0 (22-78) (108-250) (60-72) OA patients 52 17:35 55 rf: 14 175 & 32 67 * 3.9 (25-86) (125-259) (59-75) Pi NS NS 0.006 NS *Mean + SD, with range shown in parentheses. Wudent’s f-test comparisons between the reference and OA groups for age, weight, and height; chi-square analysis for sex distribution. NS, not significant.
`
`and four acquisitions were obtained in a 256 x 256 matrix. Our “pilot” films were, in fact, a complete set in the coronal plane, but with a larger gap between slices than in the sagittal ones. We did not rec- ognize osteophytes not present also in sagittal images. We did not perform axial images. Because a major pri- ority of our project was to study both knees in each sub- ject, and because we needed FLASH studies to help characterize synovial proliferation, we were limited by total time taken for a study. For articular cartilage thickness the point chosen on the patella is readily de- fined in both sagittal and axial images. We do acknowl- edge that to fully assess degenerative changes in the patello-femoral compartment in a clinical study, axial planes are useful.
`The location of synovial proliferation (in particu- lar, invasion into the infrapatellar fat pad, intercon- dylar space, and anterior horn of the lateral meniscus), its irregular shape and internal heterogeneity, especially evident in FLASH images, serve to readily distinguish synovial proliferation from joint fluid. It is worth men- tioning, however, that our readings have been conser- vative, because almost certainly smaller degrees of synovial thickening cannot be detected. In this case, the prevalence of synovial proliferation in early OA may
`
`Analysis of A4R Images
`
`MR Imaging
`
`MR imaging was performed as previously de- scribed’2*‘6 at 1.5 Tesla on a Siemens instrument with a resonator designed to achieve uniform receptivity throughout the knee. The entire joint was evaluated on sagittal,
`
`T,
`
`-704-
`
`

`
`the knee 0
`
`FERNANDEZ-MADRID ET AL.
`
`705
`
`RESULTS
`
`MRI Characteristics of Reference Subjects
`and Patients with OA
`
`T,
`
`-weighted images by a scheme devised and validated in an earlier studyI to accom- modate established criteria for the diagnosis of menis- cal tears as well as the range of meniscal signals found in normal male and female subjects aged 22-78 yr.
`
`Tl
`
`Radiographs
`
`be even higher than we report here, further underscor- Grade 3, moderate joint space narrowing; Grade 4, se- ing our conclusions. vere joint space narrowing. Images were evaluated at the console with various window settings to obtain best contrast between artic- ular cartilage and adjacent structures and to ascertain whether linear lesions in menisci did or did not com- municate with the meniscal surface. Meniscal tears were diagnosed only if abnormalities extended to the surface. Menisci were graded in
`-weighted images as previously describedI for evidence of ero- sions on the femoral condyles and on the tibia1 plateau. Hyaline cartilage is readily contrasted against joint fluid. In our earlier study of fresh AKA specimens, we confirmed rigorously using histopathology that our analysis of hyaline cartilage in r, -weighted sagittal im- ages is valid. I2 The inherent error in articular cartilage thickness is due to resolution, not to chosen imaging planes or pulse sequences.” The subjects, classified according to radiographic grade,15 are grouped in Table 2. Eighty-five of the 97 knees studied had one or more MRI abnormalities, and 50 of the 52 patients with OA had one or more of these abnormalities in at least one knee. The proportion of abnormalities seen by MRI, increased significantly as the radiographic grade increased (Table 2). It is strik- ing that 62.5% of patients with radiographic Grade 0 and 85% of those with Grade 1 had one or more MRI abnormalities (Table 2). The prevalence of synovial thickening, synovial fluid, meniscal tears, and subchon- dral lesions, was significantly greater in the knees of patients with OA compared to the knees of the refer- ence subjects for all radiographic grades (Table 2).
`Plain anterior-posterior and lateral radiographs were obtained under weight-bearing conditions with a 40 inch source to detector distance. Radiographs were exam- ined by a radiologist (P.R.M.) and a rheumatologist (F.F.M.) who agreed on the assignment of radiographic grades by the method of Kellgren and Lawrence,” which scored knees as follows: Grade 0, no joint space narrowing; Grade 1, doubtful joint space narrowing; Grade 2, minimal but definite joint space narrowing; Osteophytes were found in 55 of 97 knees of OA pa- tients. An example of osteophytes seen by MRI is shown in Fig. 1. The prevalence of osteophytes was sig- nificantly greater in the knees of patients with OA com- pared to the knees of the reference subjects, except for radiographic Grade 0 (Table 2). Subchondral involve- ment and joint fluid were detected in 65% and 60% of OA patients, respectively. Anatomical abnormalities compatible with synovial thickening, not seen in the reference group, were present in 73% of OA patients. Synovial thickening typically appeared in or near the intercondylar region of the knee, in the infrapatellar fat pad, or.in the pos- terior joint margin. Frequently, when synovial thick- Table 2. MRI abnormalities in reference subjects and patients with OA Radiographic grade* Total knees Percentages of knees with: Synovial thickening Synovial fluid Meniscal tears Subchondral lesions Osteophytes Cartilage erosions 1 or more of above Reference subjects OA patients 0 0 1 2 3 4 80 24 20 27 21 5
`
`0
`
`25.0t 40.0t 77.8t 81.Ot 1OO.ot 3.7 41.7t 45.0t 66.7t 76.2t 1oo.ot 3.7 16.7$ 45.01 55.61 42.9-t 8O.Ot 3.7 20.8$ 40.0t 48.lt 71.4t 1oo.ot 6.3 16.7 50.0-i 63.0t
`1OO.ot 0 0 5.0 11.11. 38.lt 1OO.ot 18.7 62.5.t 85.Ot 96.37 1OO.ot 1OO.ot *Radiographic grade by method of Kellgren and Lawrence.” Chi-square comparison with reference subjects: tp < .OOl; $p < .027.
`
`90.4t
`
`-705-
`
`MR features of osteoarthritis of
`F.
`Articular cartilage was examined in
`

`
`706 Magnetic Resonance Imaging 0 Volume 12, Number 5, 1994 Fig. 1. MR image of eroded femoral articular cartilage in the weight-bearing region of the medial compartment (radiographic Grade 3) of the right knee of a 69-yr-old male patient with osteophytes (arrows) on the anterior and posterior margins of both the femoral condyle and tibia1 plateau. erring was present in the infrapatellar fat pad, densities appeared also in the adjacent anterior horn of the lateral meniscus. This pattern occurred in 28 knees (Fig. 2). Meniscal tears, apparently spontaneous, were present in 52% of the patients (Table 2). The extent of meniscal disease was far beyond that expected from the findings in an age- and sex-comparable reference group.16 The preferred localization of spontaneous meniscal tears was in the posterior horn of the medial meniscus (Fig. 3). In 14 of the 27 patients the meniscal disease was bilateral. Cartilage erosions were found in 15% of patients with OA (Table 2). An example of eroded articular car- tilage is shown in Fig. 1. Cartilage erosions were sig- nificantly greater in the OA patients only for Grades 2 to 4 (Table 2).
`
`Incidental findings in the 40 reference subjects (Ta- ble 2) included tears of the posterior horn of the me- dial meniscus in three subjects and mild osteophytes in five subjects. Two of the latter had trace effusions. These five subjects may have early but asymptomatic OA. DISCUSSION An association between meniscal disease and OA has previously been suspected from the joint changes of- Table 3. MRI abnormalities in painful and asymptomatic knees of OA patients Asymptomatic knees Painful knees
`
`MRI Characteristics of the Reference Group
`
`Asymptomatic
`
`and Painful Knees
`
`(p <
`
`Total number of knees 19 78 Nineteen of the 97 knees of the OA patients were asymptomatic. Nevertheless, except for cartilage ero- sions, the proportions of all MRI abnormalities in the asymptomatic knee group were significantly greater
`.OOl) than in the reference group. The prevalence of MRI abnormalities in the painful knees was not sig- nificantly different than that in the asymptomatic knees, except for a greater number of painful knees with synovial fluid (Table 3). Percentages of knees with: Synovial thickening Synovial fluid Meniscal tears Subchondral lesions Osteophytes Cartilage erosions 1 or more of above 47.4 61.5 26.3 67.9* 26.3 46.2 47.4 47.4 42.1 60.3 5.3 11.5 73.7 89.7 Chi-square comparison with asymptomatic knees: *p < .OOl.
`
`-706-
`
`

`
`MR
`
`F. FERNANDEZ-MADRID ET AL.
`
`101 Fig. 2. (A) T,-weighted and (B) FLASH MR images through the lateral compartment of the right knee (radiographic Grade 2) of a 50-yr-old woman with OA, showing nodular synovial thickening (arrows) within the lower margin of the infrapatellar fat pad. A subchondral bone lesion can be seen in the anterior femoral condyle. ten observed after total meniscectomy.” Recent MRI studies of McAlindon et aLi0 and Chan et al.” found a high frequency of severe meniscal degeneration or tears in patients with advanced OA of the knee. Our results are in agreement with the reports of McAlindon et al.” and Chan et al.” but, in addition, we have ob- served a high frequency of meniscal changes in the knees from subjects at the mild end of the spectrum of OA. Changes in menisci are known to occur in patients with idiopathic gonarthrosis who are 1 or 2 decades older than our patients. 19,20 In our study, the fre- quency of meniscal disease observed at the mild end of the spectrum of primary OA in the absence of a his- tory of trauma was clearly greater than expected from studies of normal knees over a wide age range.16,2’,22 The subchondral bone is uniformly abnormal in specimens from advanced OA obtained at the time of reconstructive surgery. 23 Cancellous subchondral bone, by virtue of its plasticity, functions as a major shock attenuator.24 Our finding of a large proportion of subchondral bone lesions in early stages of OA of the knee supports the contention that physical changes in the subchondral bone may be important in the patho- genesis of OA. The role of inflammation in the pathogenesis and progression of OA is unclear.2-20,25 Although most studies recognize that synovial inflammation is a fre- quent phenomenon in OA, the current consensus is that inflammation is a secondary event occurring as a con- sequence of the release and phagocytosis of cartilage breakdown products in the joint.2,26*28 MRI offers the opportunity to define the extent and Fig. 3. MR image of the midportion of the medial compart- ment (radiographic Grade 1) showing a Grade 3 oblique me- niscal tear in the posterior horn of the left knee of a 63-yr-old male patient.
`
`-707-
`
`features of osteoarthritis of the knee 0
`

`
`708 Magnetic Resonance Imaging 0 Volume 12, Number 5, 1994 localization of synovial membrane involvement early in the course of OA. In this work we have found a high prevalence of synovial thickening even at the mild end of the spectrum of OA. Our interpretation of these MRI findings is supported by the recent report of an identical MRI pattern observed in an antigen-induced arthritis in the rabbit knee29 and in human villonodular synovitis, 3o shown in both cases to be due to synovitis. Although we found a greater proportion of MRI ab- normalities in the painful knee, the difference was not significant. In our study, the involvement of both knees was the rule, in agreement with the report of Cushna- ghan and Dieppe.31 The demonstration that synovial thickening, small amounts of synovial fluid, spontaneous meniscal dis- ease, and subchondral bone defects occur in a large number of patients with idiopathic OA of the knee in the mild end of the spectrum (Table 2), indicate that MRI shows a high prevalence of abnormalities not ev- ident in radiographs.
`We thank Drs. Jose Granda, Steven Plomaritis, and Jerome Ciullo for referring patients for this study, and Annette Skornia for expert technical assistance. This work was supported in part by the Vaitkevicius Magnetic Resonance Center, Harper Hos- pital, and Hutzel Hospital, Detroit, Ml. 1. 2. 3. 4. 5. 6. REFERENCES Moskowitz, R.W. Clinical and laboratory findings in os- teoarthritis. In: D. J. McCarty (Ed). Arthritis and Allied Conditions. 10th edition. Philadelphia: Lea and Febiger; 1985: pp. 14081432. Peyron, J.G. Epidemiological aspects of osteoarthritis.
`
`Mink, J.H.; Reicher, M.A.; Crues, J.V. Magnetic Res- onance Imaging of the Knee. New York: Raven Press; 1987. 9. 10. 11. Stoller, D. W.; Martin, C.; Crues, J.V.; Kaplan, L.; Mink, J.H. Meniscal tears: Pathologic correlation with MR im- aging.
`
`Radiology 163:731-735; 1987.
`
`McAlindon, T.E.M.; Watt, I. ; McCrae, F.; Goddard, P.; Dieppe, P.A. Magnetic resonance imaging in osteoarthri- tis of the knee: Correlation with radiographic and scin- tigraphic findings. Ann.
`Dis. 50:14-19; 1990. Chan, W.P.; Lang, P.; Stevens, M.P.; Sack, K.; Majum- dar, S.; Stoller, D.W.; et al. Osteoarthritis of the knee: Comparison of radiography, CT, and MR imaging to as- sess extent and severity.
`
`Rheum.
`
`Am. J. Roentg. 157:799-806;
`
`1991.
`
`Ann.
`
`Rheum. Dis. 49~672-675; 1990.
`
`12. 13. 14. Karvonen, R.L.; Negendank, W.G.; Fraser, S.M.; Mayes, M.D.; An, T.; Fernandez-Madrid, F. Articular cartilage defects of the knee: Correlation between mag- netic resonance imaging and gross pathology.
`
`Radiology 166:865-872; 1988.
`
`Arthritis Rheum. 29: 1039-
`
`Tyrrell, R.L.; Gluckert, K.; Pathria, M.; Medic, M.T. Fast three-dimensional MR imaging of the knee: Com- parison with arthroscopy.
`Altman, R.; Asch, E.; Bloch, D.; Bole, G.; Borenstein, D.; Brandt, K.; et al. Development of criteria for the clas- sification and reporting of osteoarthritis: Classification of osteoarthritis of the knee.
`
`1049; 1986.
`
`Ann. Rheum. Dis. 16:494-502; 1957.
`
`Orthop.
`
`15. 16. Kellgren, J.H.; Lawrence, J.S. Radiologic assessment of osteoarthritis.
`Negendank, W.G.; Fernandez-Madrid, F.R.; Heilbrun, L.K.; Teitge, R.A. Magnetic resonance imaging of me- niscal degeneration in asymptomatic knees. J.
`
`Res. 8:311-320; 1989.
`
`Magn. Reson. 67:258-266; 1986.
`
`J. Bone Joint Surg. 30B:664; 1948.
`
`Clin. Rheum. Dis.
`Cooke, T.D.V. Pathogenetic mechanisms in polyarticu- lar osteoarthritis.
`
`11:203-238; 1985. Mankin, H.J.; Brandt, K.D.; Shulman, L.E. Workshop on etiopathogenesis of osteoarthritis. J.
`
`Rheumatol.
`
`17. 18. 19. 20. Haase, A.; Frahm, J.; Matthaei, D.; Hanicke, W.; Mer- boldt, K-D. FLASH imaging. Rapid NMR imaging using low flip-angle pulses. J.
`13: 1127-1160; 1986. 21. Brunner, M.C.; Flower, S.P. ; Evancho, A.M.; Allman, EL.; Apple, D.F.; Fajman, W.A. MRI of the athletic knee. Findings in asymptomatic professional basketball and collegiate football players.
`
`Invest. Radiol. 24:72-
`
`75; 1989.
`
`Radiology 177:463-
`
`22. Kornick, J.; Trefelner, E.; McCarthy, S.; Lange, R.; Lynch, K.; Jokl, P. Meniscal abnormalities in the asymp- tomatic population at MR imaging.
`
`465; 1990.
`
`Semin. Arthritis Rheum.
`
`23. 24. Ilardi, C.F.; Sokoloff, L. The pathology of osteoarthri- tis: Ten strategic questions for pharmacological manage- ment.
`1 l(Supp1 1):3-7; 1981, Radin, E.L.; Paul, I.L. Does cartilage compliance reduce skeletal impact loads? The relative force attenuating properties of articular cartilage, synovial fluid, periar- titular soft-tissues and bone.
`
`144; 1970.
`
`Arthritis Rheum. 13: 139-
`
`Acknowledgment-
`
`&and. J. Rheumatol. 77:29-33; 1989.
`
`Arthritis Rheum. 34:
`
`Lawrence, J.S. Rheumatism in Populations. London: Heinemann Medical Books; 1977. Brandt, K.D.; Fife, R.S.; Braunstein, E.M.; Katz, B. Ra- diographic grade of the severity of knee osteoarthritis: Relation of the Kellgren and Lawrence grade to a grade based on joint space narrowing, and correlation with ar- throscopic evidence of articular cartilage degeneration.
`138 1 - 1386; 1991. Myers, S.L. ; Brandt, K.D.; Ehrlich, J.W.; Braunstein, E.M.; Shelbourne, K.D.; Heck, D.A.; et al. Synovial in- flammation in patients with early osteoarthritis of the knee. .I.
`McDevitt, C.A.; Muir, H. Biochemical changes in the cartilage of the knee in experimental and natural osteo- arthritis in the dog. J. Bone
`
`Rheumatol. 17:1662-1669;
`
`1990.
`
`1976.
`
`Joint Surg. 58B:94-101;
`
`Braunstein, E.M.; Brandt, K.D.; Albrecht, M. Magnetic resonance imaging of osteoarthritis after cruciate ligament transection in the dog shows sustained hypertrophic re- pair of articular cartilage.
`
`Skeletal. Radiol. 19:335-339;
`
`1990.
`
`-708-
`
`Fairbank, T.J. knee joint changes after meniscectomy.
`25. Sokoloff, L. The pathology of osteoarthrosis and the role
`

`
`of ageing. In: G. Nuki (Ed). The Aetiopathogenesis of Osteoarthrosis. Turnbridge Wells: Pitman Medical Pub- lishing Co.; 1980: pp. I-15. 26. Sokoloff, L. The Biology of Degenerative Joint Disease. Chicago: University of Chicago Press; 1969. 27. Pelletier, J.P. Symposium on osteoarthritis - Proteases: Their involvement in osteoarthritis. J. Rheumatol. 14(Suppl):l-133; 1987. 28. Howell, D.S. Etiopathogenesis of Osteoarthritis. In: D.J. McCarthy (Ed). Arthritis and Allied Conditions. 10th Edition, Philadelphia: Lea and Febiger; 1985: pp. 1400-1407. 29. Checkley, D.; Johnstone, D.; Taylor, K.; Waterton, J.C. High-resolution NMR imaging of an antigen-induced ar- thritis in the rabbit knee. Mpg.
`
`31.
`
`Cushnaghan, J.; Dieppe, P. Study of 500 patients with limb joint osteoarthritis. I. Analysis by age, sex, and dis- tribution of symptomatic joint sites.
`
`Ann. Rheum. Dis.
`
`50:8-13: 1991.
`
`F. FERNANDEZ-MADRID ET AL.
`
`709
`
`Res. Med.
`
`11:221-235; 1989. 30. Jelinek, J.S.; Kransdorf, M. J.; Utz, J.A.; Berrey, B.H., Jr.; Thomson, J.D.; Heekin, R.D.; et al. Imaging of pig- mented villonodular synovitis with emphasis on MR im- aging.
`
`Am. J. Roentg. 152:337-342;
`
`1989.
`
`-709-
`
`MR features of osteoarthritis of the knee 0