Risk of ILD and safety outcomes across DMARD classes in rheumatoid arthritis and RA-ILD

To evaluate the risk of interstitial lung disease (ILD) and safety outcomes of disease-modifying antirheumatic drug (DMARD) classes in rheumatoid arthritis (RA), focusing on patients with RA-associated ILD (RA-ILD).

We conducted a nationwide cohort study using Korean National Health Insurance Service data (2011-2020). Adults with newly diagnosed RA prescribed at least one csDMARD were included and followed until 2022. Four treatment groups were evaluated: TNFi, non-TNFi biologics (abatacept, rituximab, tocilizumab), JAKi, and csDMARDs. RA-ILD was defined as ≥2 ILD claims plus chest computed tomography confirmation. Outcomes included ILD incidence, hospitalisation, infection, MACE, and all-cause mortality. Time-varying Cox models with inverse probability of treatment weighting were applied for first-event analyses.

We analysed 26,120 patients with RA (209,852 treatment episodes), including 641 RA-ILD patients (1688 episodes). Adjusted ILD incidence was similar across DMARD classes. In RA and RA-ILD, non-TNFi biologics were associated with higher risks of hospitalisation (RA aHR 1.30, 95% CI 1.21-1.40; RA-ILD 1.35, 1.02-1.80), infection (RA 1.16, 1.05-1.28; RA-ILD 1.51, 1.06-2.15), and mortality (RA 4.09, 3.04-5.50; RA-ILD 4.31, 2.21-8.41) compared with csDMARDs. ILD-related admissions were similar, but pneumonia-related hospitalisations and recurrences were higher with non-TNFi biologics. TNFi and JAKi showed no significant associations with hospitalisation or infection. MACE risk did not differ across groups.

DMARD class was not associated with ILD incidence or ILD-related hospitalisation. Non-TNFi biologics were consistently linked to increased pneumonia-related hospitalisation, recurrence, and mortality, highlighting the need for close infection monitoring in patients receiving these agents.