Incidence of serious infections in the working-age Japanese adult population with rheumatoid arthritis treated with tumor necrosis factor-α inhibitors and interleukin-6 inhibitors: A nationwide retrospective cohort study.

Abstract

This retrospective cohort study aimed to compare the risk of serious infections in patients with rheumatoid arthritis (RA) treated with tumor necrosis factor-α inhibitors (TNFαi) and interleukin-6 inhibitors (IL-6i), with no prior use of biological disease-modifying antirheumatic drugs (bDMARDs). We employed the nationwide insurance claims database encompassing the years 2005 to 2018 in Japan. The inclusion criteria specified patients who were prescribed any type of bDMARDs, including TNFαi and IL-6i. The following exclusion criteria were applied: missing prescription dates, RA not diagnosed, below 16 years of age, bDMARDs prescribed within 6 months of registration, RA diagnosed post-bDMARDs prescription, and incidence of serious infections within 2 weeks before bDMARDs therapy. We applied stabilized inverse probability weights and utilized a Cox regression model to estimate the risk of serious infections associated with TNFαi and IL-6i. The cohort of 2493 patients with RA was categorized into a TNFαi group and an IL-6i group of 2018 and 475 participants, respectively. The median follow-up duration (interquartile range) was 347 (147-820) days in the TNFαi group and 369 (149-838) days in the IL-6i group. In the inverse probability-weighted cohort, the incidence rates (95% confidence interval) of serious infections were 2.13 (1.65-2.71) and 3.25 (2.15-4.69) per 100 person-years for the TNFαi and IL-6i groups, respectively. The hazard ratio (95% confidence interval) comparing the TNFαi group to the IL-6i group was 0.66 (0.36-1.20, p = 0.168). The results underscore the lack of evidence to preferentially favor either TNFαi or IL-6i as later-line therapy in the management of bDMARDs-naive RA to mitigate the risk of serious infections.