Magnetic resonance imaging evidence of tendinopathy in early rheumatoid arthritis predicts tendon rupture at six years.код для вставкиСкачать
ARTHRITIS & RHEUMATISM Vol. 52, No. 3, March 2005, pp 744–751 DOI 10.1002/art.20947 © 2005, American College of Rheumatology Magnetic Resonance Imaging Evidence of Tendinopathy in Early Rheumatoid Arthritis Predicts Tendon Rupture at Six Years Fiona McQueen,1 Vaughan Beckley,2 Jeff Crabbe,3 Elizabeth Robinson,2 Sue Yeoman,4 and Neal Stewart5 at 6 years (r ⴝ 0.11, P ⴝ 0.5). The strongest predictor of the 6-year tendinopathy score was the 1-year tendinopathy score (R2 ⴝ 0.36, P ⴝ 0.0003 [␤ ⴝ 1.28, SE ⴝ 0.31]). In 4 patients, extensor tendon rupture had occurred by 6 years. Their baseline and 1-year tendinopathy scores were higher than in the nonrupture group (medians 2.8 versus 1.0 [P ⴝ 0.04] and 4.3 versus 0.8 [P ⴝ 0.03], respectively), as were their Health Assessment Questionnaire scores (1.33 versus 0.54 [P ⴝ 0.02], 1.18 versus 0.25 [P ⴝ 0.03], and 0.98 versus 0.37 [P ⴝ 0.01] at 0, 1, and 6 years, respectively). For the group as a whole, the baseline tendinopathy score was predictive of rupture at 6 years (odds ratio [OR] 1.52, 95% confidence interval [95% CI] 1.02–2.32, P ⴝ 0.03), as was the 1-year score (OR 1.57, 95% CI 1.03–2.04, P ⴝ 0.02). Conclusion. MRI can be used to quantify tendinopathy at the wrist in RA patients. High scores in early disease were predictive of tendon rupture in a small group of patients, but further studies are required to determine whether this has clinical relevance. Objective. To determine whether magnetic resonance imaging (MRI) evidence of tendinopathy in early rheumatoid arthritis (RA) could be used to predict the course of tendon involvement in later disease and specifically the risk of tendon rupture. Methods. The occurrence, pattern, and progression of tendinopathy were studied prospectively over 6 years in a cohort of patients who had presented with RA. Of 42 patients enrolled, full MRI and clinical data were available for 31 at 6 years. MRI scans of the dominant wrist were scored for tendinopathy by 2 radiologists using a validated system. These data were compared with MRI synovitis, erosion scores, and disease activity measures. Prognostic factors for tendon inflammation and rupture were sought. Results. Thirty-four patients (81%) had MRI evidence of tendinopathy at baseline, falling to 59% at 1 year and 68% at 6 years. The most commonly affected site was the extensor carpi ulnaris. MRI tendinopathy and synovitis scores were correlated at baseline (r ⴝ 0.37, P ⴝ 0.01) and 1 year (r ⴝ 0.45, P ⴝ 0.003) but not Tendinopathy involving the extensor tendons of the wrist is observed in 50–64% of patients with rheumatoid arthritis (RA) (1). An important complication is tendon rupture, which has been related both to invasion of the tendon by tenosynovial pannus and fraying of the tendon against eroded bone margins (2,3). Studies of invasive tenosynovium have revealed high levels of matrix metalloproteinases, proinflammatory cytokines, and angiogenic factors, indicating a propensity for damage to neighboring tissues (4). The importance of ongoing tenosynovial inflammation in promoting tendon rupture has been confirmed by observations that prophylactic dorsal tenosynovectomy can prevent this complication in many patients (5). However, the decision to perform surgery is usually made in established cases of RA once Supported by grants from the Health Research Council of New Zealand, the Arthritis Foundation of New Zealand, the Auckland Medical Research Foundation, Lotteries Health, New Zealand, the Auckland Radiology Group, Sanofi-Winthrop, and the Auckland Regional Rheumatology Research Trust. 1 Fiona McQueen, MBChB, MD, FRACP: University of Auckland and Auckland District Health Board, Auckland, New Zealand; 2 Vaughan Beckley, Elizabeth Robinson, MSc: University of Auckland, Auckland, New Zealand; 3Jeff Crabbe, MBChB, FRANZCR: Auckland Radiology Group, Auckland, New Zealand; 4Sue Yeoman: Auckland District Health Board, Auckland, New Zealand; 5Neal Stewart, MBChB, FRANZCR: Auckland District Health Board and Auckland Radiology Group, Auckland, New Zealand. Address correspondence and reprint requests to Fiona McQueen, MBChB, MD, FRACP, Department of Rheumatology, Building 7, Auckland Hospital, Private Bag 92024, Auckland, New Zealand. E-mail: email@example.com. Submitted for publication July 6, 2004; accepted in revised form December 2, 2004. 744 PREDICTING TENDON RUPTURE THROUGH MRI EVIDENCE clinical or radiologic features of joint damage have developed (6). Magnetic resonance imaging (MRI) has been shown to be a powerful tool in identifying early soft tissue and bony change at sites in the hand and wrist in early RA (7,8). At 4 months, as many as 90% of patients show MRI evidence of wrist synovitis (9). MRI is an excellent modality for imaging tendons at the wrist in view of its ability to identify characteristic signal changes in tissues which have become inflamed or damaged (7). Hypertrophy of the tenosynovium within the tendon sheath is revealed as bright signal in thickened tissue surrounding a tendon on T1-weighted postcontrast and T2-weighted sequences, while an accompanying effusion within the sheath is low signal on early T1-weighted postcontrast scans. Changes within the tendon itself can be identified and include increased signal on T2-weighted images implying inflammatory change, as well as thinning and attenuation, which are likely to precede rupture (7,10). MRI identification of tendon rupture is highly correlated with surgical confirmation and has been shown to be more accurate than clinical evaluation (3). In this report we describe the progression and complications of tendinopathy, as revealed by MRI of the dominant wrist, in a cohort of RA patients who have been studied prospectively over a 6-year period. We investigated associations between tendinitis/ tenosynovitis, synovitis, and erosion, and sought prognostic indicators of tendon rupture. PATIENTS AND METHODS Patient population and clinical assessments. Patients for this study have been studied longitudinally over a 6-year period and have been previously described (9,11). Patients were referred to an early arthritis clinic by primary care physicians and specialist rheumatologists after a media campaign. Approval for the study was granted by regional ethics committees. Enrollment proceeded from 1994 through 1996, and then patients were followed with clinical assessments and MRI scans of the dominant wrist at 0, 1, and 6 years. Disease activity was assessed using tender and swollen joint counts, erythrocyte sedimentation rate (ESR), and C-reactive protein (CRP) levels. The Disease Activity Score (DAS) was derived according to the method described by van der Heijde et al (12). At the onset of the study the now-validated DAS28 measure had not been developed, so the original DAS calculation was used at each time point. Joint subluxation and malalignment were not scored, but function was assessed using the Health Assessment Questionnaire (HAQ) score (13). In 31 patients, full clinical and MRI data were available at 6 years. In an additional 3 patients, clinical assessments were performed but MRI scans could not be obtained because of 745 Table 1. Medication profile of the study cohort at baseline and at 1 and 6 years* NSAIDs DMARDs Sulfasalazine (1–3 gm/day) Methotrexate (7.5–25 mg/week) Prednisone (1–13 mg/day) Baseline (n ⫽ 42) 1 year (n ⫽ 42) 6 years (n ⫽ 34) 38 (90) 21 (50) 18 (43) 1 (2) 6 (14) 24 (57) 30 (71) 15 (36) 11 (26) 10 (24) 17 (50) 30 (88) 10 (29) 15 (44) 9 (28) * Values are the number (%). Treatments received by the 4 patients in whom extensor tendon rupture occurred during the study were as follows: at baseline, patient 1 piroxicam (pirox.), patient 2 diclofenac, sulfasalazine (SSZ), prednisone (pred.), patient 3 naproxen (naprox.), methotrexate (MTX), pred., patient 4 pirox., SSZ, pred.; at 1 year, patient 1 SSZ, patient 2 MTX, pred., patient 3 naprox., SSZ, MTX, pred., patient 4 pirox., MTX, pred.; at 6 years, patient 1 ibuprofen, SSZ, patient 2 MTX, patient 3 naprox., SSZ, MTX, azathioprine, patient 4 MTX, pred., cyclosporine. NSAIDs ⫽ nonsteroidal antiinflammatory drugs; DMARDs ⫽ disease-modifying antirheumatic drugs. surgical placement of pins across the carpus (2 patients) or a pacemaker in situ (1 patient). At surgery, 2 of these patients were found to have tendon rupture and their data are included in the analysis. Demographic information and disease activity measures for the total group at baseline and at 1, 2, and 6 years have been published (11). At 6 years, 65% of the cohort was female with a mean age of 46 years. The median (range) for the DAS at 0 and 6 years were 4.2 (1.3–7.6) and 2.7 (0.9–4.6). Table 1 gives details of antirheumatic medication for the group as a whole and for those in whom tendon rupture occurred. MRI scans. An MRI scan of the dominant wrist was obtained using a 1.5T MRI scanner (General Electric Signa, Milwaukee, WI) with a dedicated wrist coil (Medical Devices, Waukesha, WI). The hand was placed in the wrist coil where it fitted snugly at the patient’s side with the palm facing the body, thumb anteriorly. The field of view was 8 cm and included the distal radioulnar, radiocarpal, and midcarpal joints as well as the metacarpal bases. The small field of view was chosen to optimize resolution and did not include metacarpophalangeal joints. Coronal and axial T1 sequences were performed, followed by axial fat-suppressed fast spin-echo T2, then coronal fat-suppressed T1 sequences after injection of gadolinium diethylenetriaminepentaacetic acid (Gd-DTPA; Nicomed Omniscan, Amersham Health, Lindesnes, Norway), which acts as a contrast agent. An axial fat-suppressed T1 post–Gd-DTPA sequence was also included at 6 years. Slice thickness of 3 mm was used throughout. Scoring MRI scans. The system used to score MRI scans has been previously described (9). Briefly, 9 tendon groups were scored including extensor pollicis brevis and abductor pollicis, extensor carpi radialis (ECR), extensor pollicis longus (EPL), extensor digitorum and indicis (ED), extensor digiti minimi, extensor carpi ulnaris (ECU), flexor carpi radialis (FCR), flexor digitorum superficialis and profundus, and flexor carpi ulnaris. They were evaluated on axial scans using T1-weighted and T2-weighted sequences. A score of 0 (if absent) or 1 (if present) was given for inflammation in the tendon sheath (increased bright signal around the tendon 746 McQUEEN ET AL on T2-weighted images and ⬎1 mm thickness of sheath on 2 consecutive slices). Inflammation within the tendon itself was also scored (0 ⫽ normal signal, 1 ⫽ increased signal on T1-weighted and/or T2-weighted sequences). Tendon size was scored as 0 (normal), 1 (thickened), or 2 (attenuated). The definition of “normal” tendon size was determined by the individual radiologists on the basis of their specialist experience. Scores were added for a total tendon/tendon sheath score for the carpus (maximum possible score ⫽ 36). Seven individual sites within the carpus were assessed for synovitis: the distal radioulnar joint, radiocarpal joint (ulnar aspect), radiocarpal joint (radial aspect), intercarpal joint (between the proximal and the distal carpal row), carpometacarpal joints (second through fifth assessed together), first carpometacarpal joint, and the pisotriquetral joint. Synovitis was scored according to synovial thickening and postgadolinium enhancement. Erosions were defined as focal areas of loss of low signal cortex, with sharply defined margins, identified on both T1-weighted and T2-weighted sequences in 2 planes, with a cortical break seen in at least 1 plane. Erosions were scored at 15 sites according to number and size and summed for a total carpal score (maximum possible score ⫽ 90). Bone marrow edema was identified as poorly defined regions of low signal within bone on T1-weighted images with high signal on fat-suppressed T2-weighted images and was scored at the same sites (maximum possible carpal score ⫽ 30). MRI scans were scored independently by 2 experienced musculoskeletal radiologists (NS and JC) who were blinded to clinical data. Their mean scores were used for data analysis. Statistical analysis. Spearman’s correlation was used to test for associations between tendinitis, other MRI scores, and clinical measures. Mann-Whitney U tests were used to investigate whether tendon scores differed between rupture and nonrupture groups. Linear and exact logistic regressions were used to determine predictors of tendinitis at 6 years and rupture at 6 years, respectively. Generalized linear mixed models, which allow for the correlations between data from the same patients, were used to investigate changes over time in tendinitis scores. Site-specific analysis was performed to determine whether tendon involvement was associated for each case, first with synovitis in a neighboring joint and second with erosion at an adjacent bone. For example, in the case of the EPL tendon, relevant joints were the radiocarpal joint (radial aspect) and first carpometacarpal joint, and adjacent bones were the radial head, scaphoid, and trapezium (data available on request). RESULTS Interobserver reliability for scoring tendinopathy. Interobserver reliability for tendinopathy scoring was high with intraclass correlation coefficients (ICCs) of 0.77 (95% confidence interval [95% CI] 0.60–0.87) at baseline, 0.73 (95% CI 0.54–0.84) at 1 year, and 0.85 (95% CI 0.71–0.92) at 6 years. Reliability data for erosion and synovitis scoring have been published elsewhere (11). Figure 1. Box plots of the mean observer score for magnetic resonance imaging tendinopathy at baseline and at 1 and 6 years. Each box represents the 25th to 75th percentiles. Lines outside the boxes represent the 10th and 90th percentiles. Lines inside the boxes represent the median. Circles indicate outliers and ⴱ indicates an extreme data point. Progression of tendinopathy over 6 years. The total MRI tendon score comprised 3 measures; tendon signal change, tendon size, and tendon sheath signal change. The profile of tendon scores within the patient group was examined at baseline and at 1 and 6 years (Figure 1). Thirty-four patients (81%) had MRI evidence of tendinopathy at baseline, compared with 59% at 1 year and 68% at 6 years. There was an indication that the median MRI tendinitis score increased over the 3 time points (P ⫽ 0.08), and by 6 years there was a greater spread of scores with 3 patients having very high scores, indicating severe involvement. Most common sites of tendon involvement. The ECU tendon was most frequently involved at baseline and at 1 and 6 years. At all time points, the highest scores were obtained for the ECU tendon. In order of decreasing total score, the following tendons were affected: at baseline ECU, FCR, ECR, at 1 year ECU, FCR, ED, and at 6 years ECU, ED, and ECR. Correlations between MRI tendinopathy scores and clinical measures. Correlations between MRI tendinopathy scores and clinical measures were investigated at baseline and at 1 and 6 years (Table 2). At baseline, the total tendon score correlated with the swollen joint PREDICTING TENDON RUPTURE THROUGH MRI EVIDENCE 747 Table 2. Correlations of magnetic resonance imaging (MRI) tendon scores with clinical measures and other MRI scores* Correlation, r (P) Baseline Clinical measures CRP level ESR Pain score Ritchie articular index Swollen joint count DAS HAQ score Other MRI scores Synovitis Bone edema Bone erosion 1 year 6 years 0.190 (0.2) 0.272 (0.08) 0.047 (0.7) 0.121 (0.4) 0.386 (0.01) 0.334 (0.03) ⫺0.149 (0.3) 0.501 (0.001) 0.313 (0.05) 0.274 (0.08) 0.225 (0.1) 0.352 (0.02) 0.387 (0.01) 0.219 (0.1) 0.148 (0.4) 0.372 (0.03) 0.401 (0.02) 0.369 (0.04) 0.459 (0.009) 0.458 (0.009) 0.344 (0.05) 0.368 (0.01) 0.224 (0.1) 0.308 (0.04) 0.450 (0.003) 0.138 (0.3) 0.349 (0.02) 0.114 (0.5) 0.113 (0.5) 0.118 (0.5) * CRP ⫽ C-reactive protein; ESR ⫽ erythrocyte sedimentation rate; DAS ⫽ Disease Activity Score; HAQ ⫽ Health Assessment Questionnaire. count and DAS (r ⫽ 0.39, P ⫽ 0.01 and r ⫽ 0.33, P ⫽ 0.03, respectively). At 1 year, these associations remained significant and additional clinical factors were also associated, including the CRP level and ESR (r ⫽ 0.50, P ⫽ 0.001 and r ⫽ 0.31, P ⫽ 0.05, respectively). By 6 years, all scores correlated with MRI tendinopathy scores except the CRP level. The strongest correlations were with the DAS and swollen joint count (r ⫽ 0.46, P ⫽ 0.009 for both). Correlations between tendinopathy scores and synovitis and erosion scores. Correlations between tendinitis and other available MRI scores at each time point are shown in Table 2. Interestingly, the associations with synovitis and erosion scores were similar at baseline and 1 year but were no longer apparent at 6 years. The bone edema score did not correlate with tendinopathy at any time. Synovitis and erosion adjacent to tendinitis. Sitespecific analysis was undertaken to investigate for any relationship between tendinitis and synovitis in the tissues adjacent to each of the 9 carpal tendon groups scored on MRI. At baseline, there were a total of 71 of 378 “sites” (tendon groups) in the cohort with a positive tendinitis score, compared with 48 of 369 sites at 1 year and 60 of 306 sites at 6 years. At baseline, 97% of affected tendons were adjacent to inflamed joints, while 1- and 6-year images showed a slightly lower percentage (91%). A similar analysis was performed investigating Figure 2. Scatterplots showing relationship of 6-year magnetic resonance imaging tendinitis scores to baseline scores (A) and 1-year scores (B). 748 whether there were erosions near involved tendons. At baseline, 37 of the 71 tendinitis sites (52%) had positive erosion scores at adjacent carpal bones. By 6 years, this number had increased to 92%, and in only 6 sites (8.3%) were there no erosions present in adjacent bones. Early MRI tendinitis scores are predictive of 6-year scores. Linear regression was used to measure the ability of the baseline and 1-year MRI tendinitis scores to predict the 6-year score. Baseline scores were moderately predictive of tendinitis (R2 ⫽ 0.12, P ⫽ 0.05 [␤ ⫽ 0.69, SE ⫽ 0.34]), but less so than 1-year scores, which were highly predictive of 6-year tendinitis (R2 ⫽ 0.36, P ⫽ 0.0003 [␤ ⫽ 1.28, SE ⫽ 0.31]) (Figure 2). Tendon rupture in patients with high tendinitis scores. At 6 years, full-thickness tendon rupture at the ECU was diagnosed from wrist MRI scans in 2 patients (1 man and 1 woman, ages 36 and 59 years, respectively, at baseline). Rupture of extensor tendons to the index and little fingers was also diagnosed in 1 male patient (age 54 at baseline) on surgical exploration of the wrist prior to fusion at 5 years. Another female patient (age 42 years at baseline) needed wrist surgery 1 year after enrollment and was found to have a ruptured EPL. Neither of these patients was able to have MRI scanning performed at 6 years, due to metal pins at the carpus. Figure 3 shows the baseline and 1-year scans from 1 of the tendon rupture patients, revealing early tenosynovitis followed by swelling and partial rupture of the tendon at 1 year, probably associated with invasion by pannus. Scans at 6 years did not reveal the ECU tendon at all, and it was assumed to have ruptured. The median tendon score in the rupture group was higher than that in the nonrupture group at baseline (2.8 versus 1.0; P ⫽ 0.04) and 1 year (4.3 versus 0.8; P ⫽ 0.03). Exact conditional logistic regression analysis revealed that both baseline and 1-year MRI tendinitis scores were predictors of tendon rupture by 6 years (odds ratio [OR] 1.52, 95% CI 1.02–2.32, P ⫽ 0.03 and OR 1.57, 95% CI 1.03–2.04, P ⫽ 0.02, respectively). Patients with tendon rupture had higher DAS and HAQ scores than those without rupture. The DAS in the tendon rupture group was higher than in the nonrupture group only at the 1-year examinations (3.59 versus 2.32; P ⫽ 0.04), but the HAQ score was higher in the tendon rupture group at all 3 time points (1.33 versus 0.54 [P ⫽ 0.02], 1.18 versus 0.25 [P ⫽ 0.03], and 0.98 versus 0.37 [P ⫽ 0.01] at baseline, 1 year, and 6 years, respectively). Medical management of tendon rupture was relatively aggressive (but did not include anti–tumor necrosis factor ␣ [anti-TNF␣] therapy) and is detailed in Table 1. McQUEEN ET AL DISCUSSION This is the first prospective MRI study to investigate the progression of tendinopathy at the wrist in RA. Tendon rupture is a serious complication of RA with an estimated incidence of 1 in 100 patients (14). It occurs more frequently in those with aggressive disease (15), who may have been overrepresented in our group (fulfilled American College of Rheumatology [formerly, the American Rheumatism Association] criteria for RA at presentation) (16), and is frequently insidious in onset and painless (15). Rubens et al clinically identified extensor tendon rupture in only half of their patients who had tendon rupture as confirmed by surgical inspection, whereas tendon rupture was identified by MRI scanning in 9 of 10 (3). Thus, MRI evaluation of tendinopathy may be superior to clinical examination, and could be useful in planning management of disease. In our study, those with tendon rupture had higher MRI tendinopathy scores at baseline and 1 year than the nonrupture group, and higher HAQ scores at baseline, 1 year, and 6 years. However, specific tendon changes prior to rupture were variable, and included examples of florid tenosynovitis, tendon enlargement, and signal change as well as tendon attenuation. We developed a scoring system for quantifying MRI findings of soft tissue and bony pathology in 1996, and we have applied this score to this cohort consistently over a 6-year period (9). Our system was developed before the Outcome Measures in Rheumatology Clinical Trials (OMERACT) RA-MRI scoring system (RAMRIS) was published (17). The 2 scores are very similar, but in our system erosions were scored by number and size, rather than percent of involved bone as in RAMRIS. A score for tendinopathy was omitted from the RAMRIS because of concerns about feasibility (18), but we have shown that in the hands of experienced radiologists, scoring reliability was high (ICCs ranging from 0.73 to 0.85 at 3 time points). Whether a 3-point scale is sufficiently sensitive to describe the range of severity of tendinopathy in RA has not been determined, and others may develop the system further. Ostendorf et al recently described their own tendinopathy scoring system as an add-on to the OMERACT score when describing pathology at the metatarsophalangeal joints in patients with very early RA (8). MRI features of tenosynovitis have also been described in other conditions including spondylarthropathies (19) and systemic lupus erythematosus (20). An interesting point of difference between lupus-associated Jaccoud’s arthropathy and RA is that the flexor tendons are more often involved in the PREDICTING TENDON RUPTURE THROUGH MRI EVIDENCE 749 Figure 3. A, Baseline axial T1-weighted magnetic resonance imaging scan of the wrist, showing thickening of the extensor carpi ulnaris (ECU) tendon sheath (arrow). B, Increased signal on the equivalent T2-weighted image (arrow). C, Axial T1-weighted scan at 1 year, showing extensive tenosynovitis within the ECU tendon sheath (thick arrow) and a partial tear within the substance of the ECU tendon with likely pannus invasion (thin arrow). D, T2-weighted image corresponding to C, confirming the extensive ECU tenosynovitis (thick arrow) and the partial ECU tear with likely pannus invasion (thin arrow). 750 former condition, as opposed to extensor tendons in the latter (20). Although we did not investigate healthy subjects in this study, MRI scans from normal subjects have been studied by others (3,10). Small amounts of fluid in 1 or more dorsal wrist compartments have been described, with rare examples of increased signal intensity within the substance of tendons, but pannus has not been identified. Thus, it is possible that some of the lowpositive tendon scores from our cohort could have occurred without significant tendinopathy. However, the major conclusions of this study have been drawn from patients with high tendinopathy scores, which are most unlikely to occur in the absence of true tendon involvement. Tendonitis was frequently found adjacent to regions of synovitis at the wrist, especially within the first year of disease. This is consistent with an immunopathologic picture of early RA, in which inflammatory change within the tenosynovium mirrors that of the synovial membrane, as shown by Jain et al (4). In our cohort, erosions were identified adjacent to tendinitis/ tenosynovitis in only half of patients at baseline, but in 92% at 6 years. Friction of tendons against sharp, eroded bony margins could promote inflammatory change and eventually lead to tendon attrition and rupture. However, in many instances the erosions, although involving adjacent bones, were not directly under the path of the tendons in question, and a causal relationship between erosion and tendon rupture could not be established with certainty. Swen et al compared MRI with the gold standard of surgical inspection for the diagnosis of partial tears of the finger extensor tendons in RA (14). They found MRI to have high specificity for partial tears (0.83), but low sensitivity and positive predictive value. It is difficult to compare our study with theirs, as our MRI imaging protocol was different, using a more powerful 1.5T scanner and the addition of postcontrast T1-weighted coronal and axial sequences to T2-weighted axial sequences for better resolution of tendon pathology. Our MRI assessment of tendinopathy also differed, in that it was graded on a continuous scale. Higher grades would have included partial tears, but these were not specifically documented. However, complete tears, which were more easily verifiable, were recorded and used as an end point for our analysis. In conclusion, MRI appears to be a useful imaging modality in the evaluation of tendon pathology at the wrist in patients with RA. Our data suggest that while tenosynovitis parallels synovitis in early disease, other McQUEEN ET AL factors, including bone erosion, may contribute to tendon attrition and rupture in more advanced RA. If MRI evidence of high-grade tendinopathy is identified in patients within a year of presentation, they are likely to have ongoing tendon disease, which in some cases may result in tendon rupture. Prophylactic surgical intervention might be planned for these patients, or local steroid injections to the dorsal sheath (21,22) may be useful in controlling tenosynovitis. TNF␣ inhibitors can reduce collagenase concentrations in tenosynovial tissue (23), and could also play a role in the early treatment of these at-risk patients. Further prospective studies are required to explore these possibilities and clarify the clinical relevance of these MRI findings. ACKNOWLEDGMENTS The authors wish to acknowledge the assistance of the Auckland rheumatologists who referred patients for this study. We are also most grateful to the technical staff at the Auckland Radiology Group (in particular Ms Rika Nel), who supervised the performance of the MRI scans. REFERENCES 1. Brown FE, Brown ML, Hanover NH. Long-term results after tenosynovectomy to treat the rheumatoid hand. J Hand Surg [Am] 1988;13:704–8. 2. Moore JR, Weiland A, Valdata L. Tendon ruptures in the rheumatoid hand: analysis of treatment and functional results in 60 patients. J Hand Surg [Am] 1987;12:9–14. 3. Rubens DJ, Blebea JS, Totterman SM, Hooper MM. Rheumatoid arthritis: evaluation of wrist extensor tendons with clinical examination versus MR imaging, a preliminary report. Radiology 1993; 187:831–8. 4. Jain A, Nanchahal J, Troeberg L, Green P, Brennan F. Production of cytokines, vascular endothelial growth factor, matrix metalloproteinases, and tissue inhibitor of metalloproteinases 1 by tenosynovium demonstrates its potential for tendon destruction in rheumatoid arthritis. Arthritis Rheum 2001;44:1754–60. 5. Thirupathi RG, Ferlic DC, Clayton ML. Dorsal wrist synovectomy in rheumatoid arthritis: a longterm study. J Hand Surg [Am] 1983;8:848–56. 6. Ryu J, Saito S, Honda T, Yamamoto K. Risk factors and prophylactic tenosynovectomy for extensor tendon ruptures in the rheumatoid hand. J Hand Surg [Br] 1998:23:658–61. 7. Peterfy CG. Magnetic resonance imaging in rheumatoid arthritis: current status and future directions. J Rheumatol 2001;28: 1134–42. 8. Ostendorf B, Scherer A, Modder U, Schneider M. Diagnostic value of magnetic resonance imaging of the forefeet in early rheumatoid arthritis when findings on imaging of the metacarpophalangeal joints of the hands remain normal. Arthritis Rheum 2004;50:2094–102. 9. McQueen FM, Stewart N, Crabbe J, Robinson E, Yeoman S, Tan P, et al. Magnetic resonance imaging of the wrist in early rheumatoid arthritis reveals a high prevalence of erosions at 4 months after symptom-onset. Ann Rheum Dis 1998;57:350–6. PREDICTING TENDON RUPTURE THROUGH MRI EVIDENCE 10. Valeri G, Ferrara C, Ercolani P, de Nigris E, Giovagnoni A. Tendon involvement in rheumatoid arthritis of the wrist: MRI findings. Skeletal Radiol 2001;30:138–43. 11. McQueen FM, Benton N, Perry D, Crabbe J, Robinson E, Yeoman S, et al. Bone edema scored on magnetic resonance scans of the dominant carpus at presentation predicts radiographic joint damage of the hands and feet six years later in patients with rheumatoid arthritis. Arthritis Rheum 2003;48:1814–27. 12. Van der Heijde DM, van’t Hof MA, van Riel PL, Theunisse LA, Lubberts EW, van Leeuwen MA, et al. Judging disease activity in clinical practice in rheumatoid arthritis: first step in the development of a disease activity score. Ann Rheum Dis 1990;49:916–20. 13. Fries JF, Spitz P, Kraines RG, Holman HR. Measurement of patient outcome in arthritis. Arthritis Rheum 1980;23:137–45. 14. Swen WA, Jacobs JW, Hubach PC, Klasens JH, Algra PR, Bijlsma JW. Comparison of sonography and magnetic resonance imaging for the diagnosis of partial tears of finger extensor tendons in rheumatoid arthritis. Rheumatology (Oxford) 2000;39: 55–62. 15. Flatt AE. Soft tissues: operative treatment. In: Flatt AE, editor. The care of the arthritic hand. St. Louis: Quality Medical Publishing; 1995. p. 218–32. 16. Arnett FC, Edworthy SM, Bloch DA, McShane DJ, Fries JF, Cooper NS, et al. The American Rheumatism Association 1987 751 17. 18. 19. 20. 21. 22. 23. revised criteria for the classification of rheumatoid arthritis. Arthritis Rheum 1988;31:315–24. Ostergaard M, Peterfy C, Conaghan P, McQueen F, Bird P, Ejbjerg B, et al. OMERACT rheumatoid arthritis magnetic resonance imaging studies: core set of MRI acquisitions, joint pathology definitions, and the OMERACT RA-MRI scoring system. J Rheumatol 2003;30:1385–6. McQueen FM, Ostergaard M, Conaghan P, Peterfy C, Lassere M, Bird P, et al. OMERACT rheumatoid arthritis magnetic resonance imaging studies: summary of OMERACT 6 MR imaging module. J Rheumatol 2003;30:1387–92. Olivieri I, Barozzi L, Pierro A. Toe dactylitis in patients with spondyloarthropathy: assessment by magnetic resonance imaging. J Rheumatol 1997;24:926–30. Ostendorf B, Scherer A, Specker C, Modder U, Schneider M. Jaccoud’s arthropathy in systemic lupus erythematosus: differentiation of deforming and erosive patterns by magnetic resonance imaging. Arthritis Rheum 2003;48:157–65. Nalebuff EA, Feldon PG, Millendor LH. Rheumatoid arthritis in the hand and wrist. In: Green DP, editor. Operative hand surgery. 2nd ed. New York: Churchill Livingstone; 1988. p. 1658–69. Williamson SC, Feldon P. Extensor tendon ruptures in rheumatoid arthritis. Hand Clin 1995;11:449–59. Jain A, Brennan F, Nanchahal J. Treatment of rheumatoid tenosynovitis with cytokine inhibitors. Lancet 2002;360:1565–6.