JAK-inhibitors and risk on serious viral infection, venous thromboembolism and cardiac events in patients with rheumatoid arthritis: A protocol for a prevalent new-user cohort study using the Danish nationwide DANBIO register.

Janus Kinase inhibitors (JAKis) are targeted synthetic disease-modifying antirheumatic drugs and represent an important alternative to treat patients with moderate to high rheumatoid arthritis (RA) disease activity. Safety concerns associated with increased risk for venous thromboembolism (VTE), serious viral infection, and, more recently, major adverse cardiovascular events (MACE) in JAKi users have emerged worldwide. However, as the exact mechanisms to explain these safety concerns remain unclear, the increased risk of VTE, MACE, and serious viral infection in JAKi users is heavily debated. In light of the need to enrich the safety profile of JAKis in real-world data, we aim to quantify the incidence and risk of MACE, VTE, and serious viral infections in RA patients registered in the Danish DANBIO registry, a nationwide registry of biological therapies used in rheumatology. Therefore, we will conduct a population-based cohort study using a prevalent new-user design. We will identify all RA patients in the DANBIO, ≥ 18 years old, receiving a JAKi or a tumor necrosis factor α inhibitor (TNF-αi) from January 2017 to December 2022. Prevalent and new users of JAKis will be matched to TNF-αi comparators with similar exposure history using time-conditional propensity scores (TCPS). We will describe the cumulative incidence of the outcomes (VTE, MACE, serious viral infection) in each exposure group (JAKi users; TNF-αi users), stratified by outcome type. Additionally, the Aalen-Johansen method will be used to estimate the time-to-event survival function stratified by outcome type. We will also estimate the hazard ratio (HR) with 95% confidence interval (CI) of each outcome in both exposure groups using the time-dependent Cox proportional hazards model. Results will enrich the safety profile of JAKis in real-world data.