Herpes zoster (HZ) is a common adverse event associated with Janus kinase inhibitors (JAKi); however, direct comparisons using real-world data are lacking, which we aimed to clarify.

We retrospectively analyzed data from the multicenter ANSWER cohort, including 1,096 treatment courses from 765 patients with RA (84.1% female; mean age, 63.9 years; mean follow-up, 18.2 months) who initiated treatment with tofacitinib, baricitinib, peficitinib, upadacitinib, or filgotinib. Incidence rates (IRs) of HZ were calculated per 100 patient-years using exact Poisson method. Time-to-event and treatment retention analysis were performed using Kaplan-Meier methods, and hazard ratios (HRs) were estimated using Cox proportional hazards models adjusted for patient characteristics.

During follow-up, HZ occurred in 88 treatment courses. The IRs per 100 person-years were 5.84 for tofacitinib, 6.21 for baricitinib, 3.79 for peficitinib, 7.38 for upadacitinib, and 1.73 for filgotinib. In the adjusted Cox model, filgotinib (reference) was associated with a significantly lower hazard of HZ compared with tofacitinib (HR, 7.68; 95% CI, 1.71–34.6; P = 0.008), baricitinib (HR, 6.35; 95% CI, 1.48–27.1; P = 0.013), and upadacitinib (HR, 4.75; 95% CI, 1.11–20.3; P = 0.036), whereas no significant differences were observed in treatment retention. Patient-related variables, including age, sex, prior HZ, and vaccination history, were not significantly associated with HZ risk.

Filgotinib showed a lower observed risk of HZ compared with tofacitinib, baricitinib, and upadacitinib. Given the limited number of events and observational design, further large-scale studies with longer follow-up are warranted to confirm these findings.