Reasons-stratified treatment retention after JAK inhibitor discontinuation in rheumatoid arthritis: the ANSWER cohort study

Objective To compare the retention and disease activity trajectories of subsequent biologic and targeted synthetic disease-modifying antirheumatic drugs (b/tsDMARDs) after Janus kinase inhibitor (JAKi) discontinuation in patients with rheumatoid arthritis (RA). Methods We analysed 500 treatment courses (TCs) from the ANSWER cohort in which subsequent b/tsDMARDs were initiated after JAKi discontinuation. Subsequent treatments were classified as TNF inhibitors (TNFi), anti-IL-6 receptor antibodies (aIL-6R), CTLA4-Ig and JAKi. Drug retention was stratified by the reason for prior JAKi discontinuation and evaluated using Kaplan-Meier and multivariable Cox analyses. Changes in Clinical Disease Activity Index (CDAI) over 12 months were analysed using mixed-effects models for repeated measures. Results After prior JAKi discontinuation due to ineffectiveness, CTLA4-Ig was associated with a significantly higher risk of discontinuation due to ineffectiveness than subsequent JAKi (HR 2.83, 95% CI 1.20–6.70). Among TCs in which the prior JAKi was discontinued due to toxic adverse events (AEs), TNFi and aIL-6R were associated with lower risks of toxic AE-related discontinuation than JAKi. Overall CTLA4-Ig was associated with higher risks of discontinuation due to ineffectiveness (HR 2.57, 95% CI 1.46–4.54) and total discontinuation excluding non-toxic reasons and remission (HR 1.77, 95% CI 1.02–3.07). CDAI trajectories over 12 months did not differ significantly between treatment groups. Conclusion The optimal treatment strategy after JAKi discontinuation may depend on the reason for prior discontinuation. JAKi cycling, TNFi and aIL-6R may be reasonable after ineffectiveness, whereas after toxic AEs, TNFi and aIL-6R were associated with lower risks of subsequent toxic AE-related discontinuation than JAKi cycling.