Background MRI had higher sensitivity of detecting inflammation than physical examination and higher sensitivity of detecting bone damage than X-rays. Physical examination and X-rays are usually performed on bilateral hands of RA patients, however, MRI evaluation of unilateral hand was recommended by OMERACT at the very beginning when RA MRI score (RAMRIS) was validated in the databases only consisting of images of dominant wrists and/or metacarpal phalangeal joints (MCPJs). Recently, high performance of whole-body MRI and multi-channel synergic coil enable simultaneous scan of bilateral hands.
Objectives To explore the advantages of bilateral hands MRI on RA.
Methods Consecutive hospitalized RA patients were recruited from April 2014 to April 2016 at Sun Yat-Sen Memorial Hospital, Sun Yat-Sen University. Together with assessing swelling and tenderness of 28 joints essential for DAS28, bilateral wrists and MCPJ2∼5 of each patient were scanned simultaneously on 3.0T MRI system. MRI synovitis, osteitis and bone erosion were scored referred to the definitions and atlas of RAMRIS system.
Results (1) Among 120 RA patients were included, 79% were female, age (median and IQR, similarly hereinafter) was 52 (44∼61) years, disease duration was 48 (12∼120) months and DAS28-crp was 5.9 (4.7∼6.9). The mean imaging time for the entire MRI examinations including the time of patient positioning and contrast agent injection was 23 minutes. The mean scoring time was 10 minutes for RAMRISbilateral hands and 7 minutes for RAMRISunilateral hand. Interreader ICC for RAMRISbilateral hands was 0.852 in synovitis, 0.739 in osteitis and 0.815 in bone erosion.
(2) The frequencies of MRI synovitis, osteitis and bone erosion per joint or bone were shown in figure1A. For MCPJ2∼5, 27%∼42% of RA patients showed unilateral synovitis, 45%∼53% showed unilateral osteitis and 51%∼69% showed unilateral bone erosion (Figure1B).
(3) Of wrists with MRI synovitis, 73% were not swollen and 48% were not tender at physical examination. Of wrists with MRI osteitis, 74% were not swollen and 49% were not tender. Similar trend was seen in MCPJ2∼5 (Figure1C).
(4) The most clinically involved hands (or dominant hand in case of equally severe involved) were usually chosen for MRI of unilateral hand. Here 70 patients showed more severe involvement in unilateral hand according to patients' complaint and physical examination. However, MRI verified 42%, 40% and 42% of them respectively had more synovits, osteitis and bone erosion in the opposite hand. The other 50 patients had equally severe involvement at physical examination and the dominant hands were chosen. MRI verified 30%, 49% and 33% of them respectively had more synovits, osteitis and bone erosion in the non-dominant hand.
(5) RAMRISbilateral hands was correlated positively with RA disease activity indices, X-ray Sharp/vdH total score and subscores (all p<0.05). Moderate correlation was found between RAMRISbilateral hands in synovitis and ESR, CRP, TJC28, SJC28, DAS28, SDAI or CDAI (all r>0.4, p<0.001). Relatively high correlation was found between RAMRISbilateral hands in bone erosion and X-ray Sharp/vdH total score or erosion subscore (all r>0.7, p<0.001).
Conclusions Performing MRI of bilateral hands yields additive information compared with imaging only unilateral hand.
Disclosure of Interest None declared