For the roughly 1.3 million Americans living with rheumatoid arthritis, choosing between newer JAK inhibitors and established biologic therapies isn't just a clinical question — it's a practical one measured in years of sustained relief. Real-world data now offers a clearer picture of how these options actually perform outside of controlled trial conditions, and the gap is larger than many clinicians may expect.
The ROSA study pooled registry data from 990 RA patients across Canada and Switzerland who initiated either upadacitinib — a selective JAK1 inhibitor — or a TNF inhibitor (TNFi) between January 2020 and June 2023. After propensity score adjustment for age, disease duration, prior treatment exposure, and comorbidities, upadacitinib users were 40% less likely to discontinue treatment than TNFi users, with a hazard ratio of 0.602 (95% CI: 0.469–0.773). Notably, this persistence advantage held regardless of whether patients had previously been exposed to advanced therapies, suggesting upadacitinib may retain its durability even in treatment-experienced populations. Combination with any conventional synthetic DMARD dramatically improved persistence for upadacitinib users, with a hazard ratio of 0.225 when paired with csDMARDs versus 0.516 for TNFi combinations.
This finding enters a landscape already shaped by safety debates surrounding JAK inhibitors, particularly the ORAL Surveillance trial's cardiovascular and malignancy signals in tofacitinib. Upadacitinib's JAK1 selectivity was designed partly to improve that profile, though long-term real-world safety comparisons remain incomplete. The observational design here cannot fully eliminate confounding — sicker or more refractory patients may have been systematically channeled toward one therapy — even with propensity adjustment. The cohort size is modest for subgroup analyses, and the 12-month follow-up window captures only early durability. Still, this multi-registry approach across distinct healthcare systems lends meaningful external validity. The strong synergy between upadacitinib and csDMARDs, particularly methotrexate, is a clinically actionable signal worth factoring into treatment sequencing decisions.