Efficacy and Safety of Tofacitinib in Ankylosing Spondylitis by Baseline C-Reactive Protein Level: Post Hoc Analysis of Phase II and Phase III Clinical Trials

Objective. This post hoc analysis assessed the effect of baseline C-reactive protein (CRP) on the efficacy and safety of tofacitinib (TOF) use in ankylosing spondylitis (AS), as well as patient-reported outcomes (PROs). Methods. Phase II (ClinicalTrials.gov: NCT01786668) and phase III (ClinicalTrials.gov: NCT03502616) data from patients with active AS were used. Endpoints (weeks 12, 16, and 48), including 20% and 40% improvement in Assessment of SpondyloArthritis international Society (ASAS), AS Disease Activity Score with CRP low disease activity, 50% improvement in Bath AS Disease Activity Index (BASDAI50), and PROs (pain and fatigue), were stratified by baseline CRP (mg/L) as follows: < 5 (normal), >= 5 (elevated), < 10, and >= 10. Safety outcomes were evaluated between < 5 and >= 5 mg/L subgroups. Results. Overall, 372 patients were included (69.6% >= 5mg/L; 50.8% >= 10 mg/L). At baseline in the < 5mg/L group, more placebo-treated than TOF-treated patients received concomitant nonsteroidal antiinflammatory drugs (NSAIDs) or sulfasalazine (SSZ). Week 12 efficacy and PRO responses were generally higher for TOF vs placebo, regardless of baseline CRP. The treatment effect (placebo-adjusted response) at week 12 was generally numerically higher in >= 5 mg/L and >= 10 mg/L vs < 5 mg/L and < 10 mg/L groups. Incidence rates for treatment-emergent adverse events (TEAEs) and "all infections" were numerically higher for TOF vs placebo in patients in the < 5 mg/L group, but similar for TOF vs placebo in patients in the >= 5 mg/L group. Conclusion. Regardless of baseline CRP, TOF was more efficacious vs placebo at week 12. The placebo-adjusted efficacy and PRO responses were generally numerically higher in patients with CRP >= 5 mg/L and >= 10 mg/L vs < 5 mg/L and < 10 mg/L. The higher concomitant NSAID/SSZ exposure may have improved efficacy responses in the baseline < 5 mg/L placebo group, and ultimately affected the TOF treatment effect. Safety was consistent with previous studies of TOF use in AS, with numerically higher incidence rates for TEAEs and "all infections" for TOF vs placebo in patients with CRP < 5 mg/L.