A 1-year follow-up study of dynamic magnetic resonance imaging in early rheumatoid arthritis reveals synovitis to be increased in shared epitope-positive patients and predictive of erosions at 1 year

Objectives. Dynamic magnetic resonance imaging (MRI) allows visualization o f the synovial membrane and measurement of synovitis within the joint. A co hort of patients with early rheumatoid arthritis (RA) were studied using MR I of the dominant wrist and clinical assessments. Associations between syno vitis and the shared epitope genotype (SE) were looked for and synovitis as a predictor of joint erosion was examined. Methods. Gadolinium-enhanced MRI scans of the dominant wrist were performed in 42 early RA patients at baseline (median disease duration = 4 months) a nd after 1 yr. Images were obtained at 42-s intervals over the first 6 min after gadolinium-diethylenetriamine pentaacetic acid injection using six cu ts in the coronal plane, 2 mm apart. The site of maximal synovial enhanceme nt was selected as the region of interest (ROI). The rate of enhancement (E -rate) was calculated and compared with synovitis scores from static MRI sc ans, clinical disease activity scores and HLA-DRB1*04/01 genotyping [sequen ce-specific primer polymerase chain reaction (SSP-PCR) and DNA sequencing]. Results. Reproducibility of the E-rate measurement was assessed by re-evalu ating 10 randomly selected scans in a blinded fashion. Intra-observer relia bility was high with an intraclass correlation coefficient of 0.91, 95% con fidence interval (CI) 0.65-0.97. The E-rate correlated strongly at baseline with the maximum level of synovial enhancement (E-max) (r = 0.88, P < 0.00 01) and the static MRI synovitis score (r = 0.52, P = 0.0004). There was al so a weaker but significant correlation between E-rate and the pain score ( r = 0.29, P = 0.04). The E-rate fell from baseline to 1 yr (P = 0.02) conco rdant with clinical improvement after treatment with standard therapies. E- rate scores were higher in SE + than SE - patients (F-1,F-25 = 5.19, P = 0. 03) and were predictive of MRI erosions at 1 yr [chi-square = 5.0 (1 d.f.), P = 0.03], The baseline C-reactive protein (CRP) was also predictive of MR I erosions at 1 yr to a similar degree [chi-square = 4.7 (1 d.f.), P = 0.03 ] but the mean static synovitis score at baseline was the strongest predict or [chi-square = 9.2 (1 d.f.), P = 0.003]. Conclusions. These results show that dynamic MRI can be used to score synov itis objectively in early RA patients. Synovitis was greater in SE + patien ts, suggesting an early genetic influence on joint inflammation, and was pr edictive for the development of erosions at 1 yr.