Abstract
BACKGROUND: MRI is the most sensitive imaging modality in juvenile idiopathic arthritis (JIA), but has practical limitations. Optimizing the scanning protocol is, therefore, necessary to increase feasibility and patient comfort. OBJECTIVE: To determine the feasibility of bilateral non-contrast-enhanced open-bore MRI of knees and to assess the presence of literature-based MRI features in unsedated children with JIA. MATERIALS AND METHODS: Children were classified into two clinical subgroups: active arthritis (group 1; n = 29) and inactive disease (group 2; n = 18). MRI features were evaluated using a literature-based score, comprising synovial hypertrophy, cartilage lesions, bone erosions, bone marrow changes, infrapatellar fat pad heterogeneity, effusion, tendinopathy and popliteal lymphadenopathy. RESULTS: The MRI examination was successfully completed in all 47 children. No scan was excluded due to poor image quality. Synovial hypertrophy was more frequent in group 1 (36.2%), but was also seen in 19.4% of the knees in group 2. Infrapatellar fat pad heterogeneity was more prevalent in group 2 (86.1%; P = 0.008). Reproducibility of the score was good (Cohen kappa, 0.49-0.96). CONCLUSION: Bilateral non-contrast-enhanced open-bore knee MRI is feasible in the assessment of disease activity in unsedated children with JIA. Signs differing among children with active and inactive disease include infrapatellar fat pad heterogeneity and synovial hypertrophy.