Do newly approved drugs have a worse observed safety profile than once established? A study on time trends in risks of key safety outcomes with immunomodulatory drugs against rheumatoid arthritis.

Objectives: To investigate rates of key safety outcomes in patients with rheumatoid arthritis (RA) initiating biologic/targeted synthetic disease-modifying antirheumatic drugs (b/tsDMARDs) and in reference cohorts, presented over time since the market entry of each b/tsDMARD class and over calendar period at treatment start.

Methods: This was a nationwide register-based cohort study conducted from 2006 to 2022. From the Swedish Rheumatology Quality Register and national registers, we identified treatment initiators of b/tsDMARDs (n = 33,550 initiations), an early bionaive RA cohort (n = 16,011), and a matched general population cohort (n = 111,074). The main outcome was first of either major adverse cardiovascular event, venous thromboembolism, cancer, or serious infection. We stratified rates by time since market entry of each b/tsDMARD class at treatment start, and by calendar year of treatment start. We calculated incidence rates (IRs) and hazard ratios (HRs) using Cox regression and adjusted for patient characteristics.

Results: Overall, 5862 events were observed in the b/tsDMARD initiator cohort. b/tsDMARD treatments initiated >5 (vs <2) years since market entry of that class were associated with lower outcome rates (unadjusted HR = 0.74; 95% CI = 0.67-0.81). This association was attenuated once adjusting for patient characteristics (adjusted HR = 0.93; 95% CI = 0.84-1.03). By contrast, during our study period, adjusted rates declined (adjusted HR = 0.74 and 95% CI = 0.69-0.80 for b/tsDMARDs initiated 2016-2021 vs 2006-2010), despite a constant rate in the background population.

Conclusions: Modest channelling makes the safety profile of b/tsDMARDs appear worse when new on the market. Declining incidences of typical RA comorbidities in b/tsDMARD initiators during recent years suggest that the bar defining an "acceptable" safety profile for new b/tsDMARDs for use in RA should be lower(ed).