A Combination Model to Predict Relapse and Successful Conventional DMARDs De-Escalation in Rheumatoid Arthritis Patients with Sustained Clinical Remission.

L. Wang,Y. Geng,J. Han,X. Sun,Z. Zhang
2018-01-01
Abstract:Objective To determine the long-term outcomes of RA patients in sustained clinical remission under different therapeutic strategies and explore the risk factors to relapse. Methods RA patients in sustained clinical remission (DAS28(CRP) <= 2.6 for at least 6 months) were enrolled. Their baseline clinical features, ultrasonography and x-ray of hands were collected. The usage qf conventional synthetic disease-modified anti-rheumatic drugs (csDMARDs) at baseline and every follow-up visits were recorded. Patients were divided into maintain-therapy group or de-escalate-therapy group according to their treatment during follow-up. The time-point of follow-up visits reaching 2 years or flare (DAS28(CRP)>2.6) was defined as the endpoint of the study. The risk factors to predict flare was analysed by logistic regression model. Results 94 patients were enrolled in the study, with 59 in de-escalate-therapy group and 35 in maintain-therapy group. During an average of 20.8 months of follow-up, 40 (42.6%) patients relapsed, with 31 (52.5%) from de-escalate-therapy group and 9 (25.7%) from maintain-therapy group. De-escalate-therapy increased the risk of fiare by 2.3 times (OR=3.38, p=0.044). Baseline DAS28(CRP) (OR=6.97, p=0.038), presence of subclinical synovitis (OR=3.67, p=0.024), combination of 2 csDMARDs (OR=3.72, p=0.030) were the risk factors for relapse, and the best cut-off value of DAS28(CRP) for relapse prediction through ROC curve was 1.82. Taking the three parameters into the model for a combined prediction probability, the area under the ROC curve was 0.722 (95% CI 0.61, 0.82, p=0.000). Conclusion De-escalation therapy was associated with higher risk of relapse in RA patients with sustained clinical remission. A combination model of DAS28(CRP)<1.82 and no subclinical synovitis may help to predict successful csDMARDs reduction in RA patients with sustained clinical remission receiving csDMARDs monotherapy.
What problem does this paper attempt to address?