Predictive Value of Magnetic Resonance Imaging-detected Tenosynovitis of the Metacarpophalangeal and Wrist Joints for the Development of Rheumatoid Arthritis among Patients with Undifferentiated Arthritis

OBJECTIVE: The early diagnosis of rheumatoid arthritis (RA) improves disease outcomes. Using bilateral magnetic resonance imaging (MRI), we investigated whether or not tenosynovitis at the level of the metacarpophalangeal (MCP) and wrist joints, as well as non-symmetrical versus symmetrical involvement, predicts RA development in undifferentiated arthritis (UA) patients. METHODS: We collected the clinical and serological findings as well as bilateral gadolinium-enhanced 1.5-T MRI data of UA patients after 1 year. A multivariate logistic regression analysis was used to determine the association of tenosynovitis in UA with RA development. Ninety-one UA patients from the Nagasaki Early Arthritis Clinic who did not meet the 2010 European League Against Rheumatism/American College of Rheumatology classification criteria for RA were selected. Tenosynovitis at the MCP and wrist joints was scored according to the RA MRI scoring system. RESULTS: Of these 91 UA patients, 29 (31.9%) progressed to RA, with a median disease duration of 3 months, despite only 10.9% being positive for anti-cyclic citrullinated peptide antibody (ACPA). A univariate analysis showed higher MCP tenosynovitis scores, MCP flexor tenosynovitis, and symmetrical MCP tenosynovitis in the RA development group than in the non-development group (p＜0.05). A multivariate analysis showed that symmetrical MCP tenosynovitis was independently associated with RA development after adjusting for age, gender, swollen joint count, C-reactive protein level, and ACPA positivity (odds ratio: 4.96). The presence of symmetrical MCP tenosynovitis had low sensitivity (35%) but high specificity (87%) for RA development. CONCLUSION: MRI-detected tenosynovitis, especially symmetrical findings at the MCP joint, is predictive of RA development in a UA population with low ACPA positivity.