Nonsteroidal Antiinflammatory Drug-Sparing Effect of Tumor Necrosis Factor Inhibitors in Early Axial Spondyloarthritis: Results From the DESIR Cohort

Objective. To evaluate the effect of tumor necrosis factor (TNF) inhibitors on nonsteroidal antiinflammatory drug (NSAID) intake in a cohort of patients with early axial spondyloarthritis (SpA) over the first 2 years of followup. Methods. The Devenir des Spondylarthropathies Indifferenciees Recentes (DESIR) cohort is a prospective, multicenter, observational study cohort of patients with early inflammatory back pain. The management and treatment of these patients were decided by their treating rheumatologists. Data regarding NSAID intake (yes/no) and the Assessment of SpondyloArthritis international Society NSAID score were collected at each visit over 2 years of followup. Patients receiving a TNF inhibitor were matched with those receiving usual care, based on a propensity score. The NSAID-sparing effect of TNF inhibitors was estimated by comparing the percentage of patients reaching several end points (e.g., a decrease in the NSAID score to <10 over 2 years) and by modeling NSAID intake using mixed models. Results. Among the 627 patients who were followed up, 181 (28.9%) received a TNF inhibitor, and these patients were matched to 181 patients who received usual care. The baseline characteristics of the patients in the 2 groups were comparable (similar to 40% of the patients were male, and the mean age was 34 years). Initially, 90.2% of patients receiving TNF inhibitors and 90.0% of those receiving usual care had been treated with NSAIDs during the previous 6 months. The number of patients who received an NSAID decreased over time in both groups, but the decrease was greater in the group receiving TNF inhibitors (P=0.04). The decrease in the median NSAID score was significantly greater in the TNF inhibitor group (54.9 versus 41.9), and the percentage of patients in whom the NSAID score decreased by >50% or to <10 or in whom NSAID treatment was discontinued was greater in the TNF inhibitor group (67.6% versus 46.2%). Conclusion. Treatment with TNF inhibitors was associated with a decrease in the proportion of patients taking NSAIDs and with a rapid and sustained decrease in NSAID intake. This study is the first to confirm the NSAID-sparing effect of TNF inhibitors in patients with early axial SpA in a real-life clinical setting.