Background Rheumatoid arthritis (RA) is treated with disease modifying antirheumatic drugs (DMARDs), including biological DMARDs (bDMARDs) and more recently, Janus kinase inhibitors (JAKi). Both JAKi and bDMARDs are safe and effective, however infection risk remains a concern. Recently, the ORAL Surveillance trial showed an increased risk of infections and especially herpes zoster (HZ) in JAKi compared to TNF-inhibitors (a type of bDMARDs).^[1] However, clinical trials do not fully reflect the real-world population as they usually include younger and healthier patients and have a short follow-up time. Therefore, larger long-term observational studies also including more frail and older adults as seen in clinical practice are needed.

Objectives The objective of this study was to compare infection risk between bDMARDs and JAKi in RA patients, with a special focus on HZ and older adults.

Methods From a nationally representative prescription database, all adult RA patients starting a new JAKi or bDMARD between August 1^st 2018 and January 31^st 2021 were included. This database covers ~63% of all outpatient prescriptions dispensed in the Netherlands and contains coded prescription records including patient characteristics, medication and prescriber information. Prescriptions of antibiotic, antiviral or antifungal medication were used as proxy for infections. Infection incidence rates (IR) were compared between JAKi and bDMARDs and infection risks using multilevel Poisson regression adjusted for follow-up time and potential confounders and stratified for age < 65 and ≥ 65 years.

Results In total, 14,989 patients were included, with 20,050 treatment episodes with either JAKi or bDMARDs. Most patients were female (72%) and median age was 61 years (IQR 52-70). Infection IRs were higher in JAKi (48/100 patient years) compared to bDMARDs (35/100 patient years, adjusted incidence rate ratio (IRR) 1.22, 95% CI 1.12-1.33). More HZ infections were seen in JAKi compared to bDMARDs (adjusted IRR 3.03, 95% CI 2.26-4.07) (Table 1). No significant differences in infection IRs were found between JAKi baricitinib and tofacitinib. In older patients, absolute infection IRs were higher (overall infection IR 42/100 patient years for age ≥ 65 versus 31/100 patient years for age < 65). However, comparing JAKi to bDMARDs, IRRs were similar for all ages (adjusted IRR for age ≥ 65 1.31 (95% CI 1.15-1.49), adjusted IRR for age < 65 1.17 (95% CI 1.05-1.30)).

Conclusion JAKi are - compared to bDMARDs - associated with a slightly higher infection risk and a higher risk of HZ specifically. In older patients, absolute infection IRs are higher but relative infection risks for JAKi versus bDMARDs are similar in all age groups. No differences in infection risk between tofacitinib and baricitinib were found.

Reference [1] Balanescu AR et al. Ann Rheum Dis. 2022;81(11):1491-1503. doi:10.1136/ard-2022-222405

* adjusted for age, sex, number of used DMARDs in the 2 years prior to the index date, concurrent use of csDMARDs and corticosteroids, comorbidities (diabetes mellitus and chronic obstructive respiratory diseases), dosage (as percentage of defined daily dose) and follow-up time

**adjusted for age, sex, dosage (as percentage of defined daily dose) and follow-up time

Acknowledgements This study was funded by Pfizer.

Disclosure of Interests Merel Opdam Grant/research support from: Received grants (to the institution) from Pfizer, Nathan den Broeder: None declared, Bart van den Bemt: None declared, Kelly Mulder Employee of: Works for IQVIA solutions., Kayleigh M. van de Wiel Employee of: Works for IQVIA solutions., Hanne van Ballegooijen Employee of: Works for IQVIA solutions., Reinout van Crevel: None declared, Alfons den Broeder Grant/research support from: Grants for quality of care projects and research on RA, PsA and axSpA from Abbvie, Galapagos, Pfizer, Novartis, Lilly, Sanofi, Gilead to the institution/department rheumatology.

Keywords: bDMARD, Rheumatoid arthritis, Targeted synthetic drugs

DOI: 10.1136/annrheumdis-2023-eular.1488