The impact of glucocorticoids on the efficacy of JAK inhibitors or non-TNF-targeted biologics in RA

Objectives To evaluate the impact of oral low-dose glucocorticoids (GCs) on the effectiveness of Janus kinase inhibitors (JAKis) compared with non-TNF-targeted bDMARDs (other mechanisms of action [OMA] including tocilizumab, abatacept and rituximab) in RA patients with inadequate response to TNF inhibitors (TNFi-IRs). Methods Single-centre, retrospective cohort study. The primary outcome was the good EULAR response at 6, 12 and 24months. Secondary outcomes included treatment discontinuation and remission rates. Statistical analyses included structural equation modeling (SEM), competing risks analysis and linear mixed-effects models. Results The study included 299 patients (JAKi=161, OMA=138). Baseline GC use was similar between the groups (JAKi=66.5%, OMA=76%; P=0.089; mean (SD) daily dose 4.4 (5.5) vs 4.9 (4.5) mg/d; P=0.355). The proportion of GC users treated with JAKi or OMA at 6, 12 and 24months was 49.7% vs 63.8% (P=0.036), 37.3% vs 40.6% (P=0.161) and 27.3% vs 34.1% (P=0.430). No statistically significant differences in GC dose reduction were observed between the two groups across any time interval. The EULAR response improved at 24months, with JAKi outperforming OMA (SEM; P=0.040). GC dosage at 6months negatively affected the 24-month outcome (β=−0.456, P=0.020). DAS28-CRP remission was more frequent in the JAKi group (P<0.05), while discontinuation rates were similar (HR=1.49, P=0.34). Conclusion JAKi did not show superiority over OMA regarding steroid-sparing effects; however, JAKi yielded higher remission rates. Exposure to GC at 6months predicted poorer 24-month outcomes.