Prevalence and predictors of long-term remission in rheumatoid arthritis in real-world practice: a longitudinal study

Objectives The aim of the present study was to provide real-world evidence for factors predicting long-term remission in a longitudinal study of rheumatoid arthritis (RA) patients.Methods Long-term remission was defined by meeting American Rheumatism Association (ARA) criteria for remission and prednisolone dose <= 5 mg/d for at least 5 years. Patients in this cohort were treated by tight control strategy using step-up combination therapy with conventional synthetic DMARDs (csDMARDs), biologic DMARDs. The parameters associated with long-term remission were subjected to univariate analysis, and parameters with P-values of < 0.1 in univariate analysis were included in a multivariate regression analysis.Results One thousand two hundred and eighty-six RA subjects were considered for eligibility, and finally, 499 patients were included in the study. Median duration of follow-up was 108 months. Long-term remission occurred in 157 (31.5%) patients. Median time to long-term remission was 8 (5, 41) months. Predictors of long-term remission were absence of flare during the course of disease, occurrence of sustained remission during 6 months after starting therapy, age at the disease onset > 60, being anti-citrullinated protein antibodies (ACPA) negative, and Disease Activity Score-28 (DAS28) at cohort entry <= 5.1.Conclusion In real-world practice, long-term remission occurs in 31.5% of patients treated with a tight control strategy. Absence of flare during the course of disease, occurrence of sustained remission during 6 months after starting therapy, age at the disease onset > 60, being ACPA negative, and DAS28 at baseline <= 5.1 are independent predictors of long-term remission.