Abstract
OBJECTIVE: A North American registry of JDM patients was examined for frequency of and factors associated with corticosteroid discontinuation, complete clinical response and remission. METHODS: We evaluated probability of achieving final corticosteroid discontinuation, complete clinical response and remission in 307 JDM patients by Weibull time-to-event modelling; conditional probability of complete clinical response and remission using Bayesian network modelling; and significant predictors with multivariable Markov chain Monte-Carlo Weibull extension models. RESULTS: The probability of corticosteroid discontinuation was 56%, complete clinical response 38% and remission 30% by 60 months after initial treatment in 105 patients. The probability of remission was conditional on corticosteroid discontinuation and complete clinical response. Photosensitivity, contractures and a longer time to complete clinical response were predictive of the time to final corticosteroid discontinuation. Anti-MJ (NXP2) autoantibodies and a Northwest residential geoclimatic zone were predictive of shorter time to complete clinical response, while dysphonia, contractures, an increase in medications within 24 months and a longer time to corticosteroid discontinuation were associated with longer time to complete clinical response. Anti-p155/140 (TIF1) autoantibodies, an increase in medications within 12-24 months, or longer times to corticosteroid discontinuation and complete clinical response were associated with longer time to remission. CONCLUSION: JDM patients achieve favourable outcomes, including corticosteroid discontinuation, complete clinical response and remission, although timelines for these may be several years based on time-dependent analyses. These outcomes are inter-related and strong predictors of each other. Selected clinical features and myositis autoantibodies are additionally associated with these outcomes.