Background Magnetic resonance imaging (MRI) is potentially a useful tool in clinical studies in JIA, since it allows the simultaneous evaluation of soft tissue and bony changes. The Outcome Measures in Rheumatology Clinical Trials (OMERACT) Rheumatoid Arthritis MRI score, developed to ensure standardization and reproducibility of MRI readings in rheumatoid arthritis clinical trials, was proved a valid and reliable score for the assessment of wrist synovitis in JIA.1 The responsiveness to synovial change, however, was moderate and lower compared to that obtained in RA, thus limiting its potential role in evaluating drug efficacy. Since many JIA patients with wrist arthritis show involvement of the finger joints as well, the question arises if inclusion of the metacarpophalangeal (MCP) joints improves the wrist MRI score.
Objectives To develop a modified wrist MRI score by including the MCP joints, and to validate it by comparing its metric properties with those of the original score.
Methods We identified 59 JIA patients with active wrist involvement who had participated in a prospective study during which they performed a wrist MRI. Two readers (PvD and FV) independently read the MRIs according to the original and the modified wrist MRI score, after a training and calibration session by an experienced reader (CM). The original score included the radioulnar (RU), radiocarpal (RC) and midcarpal and carpometacarpal (MCCM) joints with a total score ranging from 0-9, whereas the modified score comprised the RU, RC, mid-carpal (MC), first carpometacarpal (CM1), 2nd to 5th carpometacarpal (CM2-5) and the five MCP joints separately, together with the pisiform-pyramidal (PP) recess, with a range of 0-33. A follow-up MRI performed at least one year after the baseline MRI was available in 27 patients. Of both scores, the concordance between the readers (intra-class correlation coefficient [ICC] and Cohen's kappa), the sensitivity to change (standardized response mean [SRM]) and correlations with clinical variables (Spearman's rho) were calculated.
Results The sample included 20 (33.9%) polyarticular, 20 (33.9%) systemic, 16 (27.1%) extended oligoarticular and 3 (5.1%) persistent oligoarticular patients. MRI showed synovitis of one or more MCP joints in 48 (81.4%) patients. The inter-reader agreement was excellent for both scores with an ICC of 0.81 (95% confidence interval [CI] 0.71-0.89) of the original score and ICC of 0.85 (95% CI 0.75-0.91) of the modified score. Cohen's Kappa of the individual joints ranged from 0.49 for the PP recess to 0.70 for the CM2-5 joints. The SRM increased from 0.93 for the original score to 1.21 for the modified score. Correlations with clinical variables reflecting disease activity improved for the modified score (Table).
Conclusions MCPs were frequently involved in patients with wrist arthritis. The modified score showed excellent reliability, while the sensitivity to change and correlation with clinical variables improved. Modification of the score should be considered.
Malattia C, et al. Ann Rheum Dis 2011:70;440-6.
Disclosure of Interest None declared