Objectives To compare the 2-year clinical effectiveness of the four globally approved Janus kinase inhibitors (JAKis; tofacitinib (TOF), baricitinib (BAR), upadacitinib (UPA) and filgotinib (FIL)) in patients with rheumatoid arthritis (RA) in real-world settings. Methods This retrospective cohort study used data from FIRST registry, a multicentre registry of patients with RA. The primary endpoint was the change in Clinical Disease Activity Index (CDAI) score at year 2. Secondary endpoints included changes in individual CDAI components, patient-reported outcomes (PROs) and reasons for JAKi discontinuation. Multivariable mixed-effects models adjusted for baseline characteristics were used to compare the four JAKis. Results A total of 607 treatment courses with JAKis (TOF: 159, BAR: 262, UPA: 122, FIL: 64) were included. Baseline characteristics differed notably among treatment groups: UPA and FIL were frequently used as the second-line JAKis for older patients with comorbidities. The 2-year overall retention rate was 78%. The most common reason for discontinuation was insufficient effectiveness, with 6.5/100 person-years (py), followed by adverse events of 4.2/100 py. As-observed analysis demonstrated the slower improvement in the UPA and FIL groups. However, multivariable analysis revealed no significant differences in CDAI or PROs. The UPA group demonstrated greater improvement in two CDAI components: tender joint count and evaluator’s global assessment. Conclusion This real-world study found no clinically meaningful differences in 2-year effectiveness among four JAKis, although the study was not powerful enough to detect differences in safety. Further long-term, real-world data are needed to evaluate the safety of these agents and refine their risk-benefit profiles.
Sonomoto et al. (2026) studied this question.