Comparative risk of incident malignancies in rheumatoid arthritis patients treated with Janus kinase inhibitors or bDMARDs: observational data from the German RABBIT register.

Abstract

To estimate the effects of Janus kinase inhibitors (JAKis) vs biologic disease-modifying antirheumatic drugs (bDMARDs) on the risk of incident malignancies (excluding nonmelanoma skin cancer) in patients and patient subgroups with rheumatoid arthritis. Episodes of disease-modifying antirheumatic drug (DMARD) treatment initiated between January 2017 and December 2020 and followed up to June 2024 in RABBIT, the German register for the long-term observation of therapy with biologics and targeted disease-modifying antirheumatic drugs in adult patients with rheumatoid arthritis, were analysed. Incidence rates (IRs) per 1000 patient-years with 95% CIs were calculated, and incident malignancy risk was estimated as hazard ratios (HRs) by inverse probability weighted adjusted Cox models. Among 2285 JAKi and 4259 bDMARD treatment episodes, 88 and 135 malignancies occurred, respectively. JAKi treatments were dominated by baricitinib and tofacitinib, while most bDMARD treatments comprised tumour necrosis factor inhibitors. IRs were 11.6 (95% CI: 9.3, 14.3) in JAKi- and 8.9 (95% CI: 7.4, 10.5) in bDMARD-treated groups. The adjusted HR comparing JAKis with bDMARDs was 1.40 (95% CI: 1.09, 1.80). An increase in the malignancy risk for JAKi vs bDMARD treatment could only be observed in treatment episodes lasting longer than 16 months. The risk appeared higher in some subgroups of patients, including those who started treatment aged ≥60 years, patients with ≥3 prior conventional synthetic DMARD treatments, and patients with high disease activity. In this German observational cohort study, an overall small increase in malignancy risk for JAKi vs bDMARD treatment was observed, with more pronounced risks in some subgroups of patients. The observed risk should be carefully counterbalanced to the known malignancy risk associated with insufficient disease control.