Background: In Japan, five JAK inhibitors (JAKinibs) are available for rheumatoid arthritis (RA) patients. As with biologic drugs (bDMARDs), there are no specific criteria for drug selection and they have been used in the order of approval, with or without MTX. In this context, the second-generation selective JAKinibs which was developed with the aim of achieving higher selectivity and reducing the risk of side effects, is also expected to improve efficacy.

Objectives: Upadacitinib (UPA) has been reported to be a more selective inhibitor of JAK1 than other JAK isoforms, and head-to-head studies comparing efficacy between JAKinibs are difficult to conduct in clinical practice. The present study evaluated the efficacy of UPA in RA patients who had discontinued first-generation JAKinibs due to inadequate efficacy.

Methods: At our institution, RA patients are assessed for disease activity at least once every three months. Patients with a positive rheumatoid factor (RF) and/or anti-cyclic citrullinated peptide antibody and a history of discontinuation of Tofacitinib (TOF) and/or Baricitinib (BAR) due to inadequate response, and who had started UPA by November 2022 were retrospectively studied. For each discontinued JAKi, one-year continuation rates were assessed by the Kaplan-Meier method, disease activity by the DAS28-ESR and SDAI, PRO by the HAQ-DI, and changes in RF as a biomarker were investigated. In addition to the UPA dose, the dose of methotrexate (MTX) and prednisolone (PSL) concomitant was also investigated. After achieving the treatment goal, MTX was tapered or discontinued following PSL, and then UPA was allowed to be tapered.

Results: Forty-three patients (5 males and 38 females) were included in the study, with age at the start of UPA 67.8 ± 11.7 years, disease duration 19.0 ± 10.3 years, DAS28-ESR 4.12 ± 0.73, SDAI 8.83 ± 5.01 and HAQ-DI 1.00 ± 0.89. The two-drug JAKinibs ineffectiveness group (two-drug group) was older (mean 71.6 years), had more active disease (DAS28-ESR mean 4.43) and used more bDMARDs (mean 5.4) than the one-drug JAKinibs ineffectiveness group. Continuation rates were 81.3% for 16 patients in the TOF ineffectiveness group, 95.5% for 22 patients in the BAR ineffectiveness group and 100% for 5 patients in the two-drug group, with no significant differences (P=0.25). Overall, DAS28-ESR fell by a mean of 1.20 (P<0.01) in the continuing cases, with 28 cases achieving 3.2 or less, SDAI fell by a mean of 3.82 (P<0.01) and HAQ-DI by a mean of 0.08 (P=0.09). RF decreased from pre-commencement in 32 patients. Reasons for discontinuation were two adverse events (pneumonia and myocardial infarction), primary invalidity and patient convenience in one case each. Overall, including three Herpes Zoster (HZ) Vaccine recipients, only one patient developed HZ. UPA doses at 1 year were less than 15 mg/day in 22 patients and less than 7.5 mg/day in 17 patients.

Conclusion: UPA was effective in patients with inadequate response to first-generation JAKinibs in this single-center observational study. The retention rate was high, suggesting the possibility of tapering UPA doses, mainly in patients who achieved treatment targets. Limitations of the study include a short observation period and small sample size to assess safety, and some patients with a history of multiple drug use extended the use of JAKinibs even if their response was somewhat inadequate. Further studies are needed to understand the impact on safety and efficacy, including head-to-head trials of JAKinibs for JAKinibs naïve and refractory cases and switching to UPA for patients who discontinued due to adverse events during remission or LDA with first-generation JAKinibs.

REFERENCES: [1] Bonelli M, Kerschbaumer A, Kastrati K, et al. Selectivity, efficacy and safety of JAKinibs: new evidence for a still evolving story. Ann Rheum Dis 2023; 0 :1–22. doi:10.1136/ard-2023-223850.

Acknowledgements: NIL.

Disclosure of Interests: None declared.