Efficacy and Safety of Upadacitinib in a Randomized, Double-Blind, Placebo-Controlled, Multicenter Phase 2/3 Clinical Study of Patients with Active Ankylosing Spondylitis

Abstract

Background/Purpose: Patients (pts) with ankylosing spondylitis (AS) who have an inadequate response/contraindication to NSAIDs have limited treatment options other than biologics. The Janus kinase (JAK) pathway is a potential therapeutic target in AS. This study assessed the efficacy and safety of upadacitinib (UPA), a selective JAK1 inhibitor, in pts with active AS.

Methods: In this double-blind, placebo (PBO)-controlled, phase 2/3 study (NCT03178487; SELECT-AXIS 1) pts with AS were randomized 1:1 to UPA 15 mg or matching placebo. Enrolled pts (≥18 y) met modified New York Criteria for AS based on central reading of radiographs, had a BASDAI ≥4 and pt assessment of total back pain ≥4 (numeric rating scale, 0–10) at screening and baseline (BL), were biologic DMARD naive, and had inadequate response to ≥2 NSAIDs or intolerance to/contraindication for NSAIDs. The primary efficacy endpoint was Assessment of SpondyloArthritis international Society (ASAS) 40 response at wk 14. Multiplicity-adjusted key secondary endpoints included change from BL to wk 14 (∆) in Ankylosing Spondylitis Disease Activity Scores (ASDAS), ∆ Spondyloarthritis Research Consortium of Canada (SPARCC) MRI spine, BASDAI50 at wk 14, ∆ AS quality of life (QoL), ASAS partial remission (PR) at wk 14, ∆ BASFI, ∆ BASMI, ∆ Maastricht Ankylosing Spondylitis Enthesitis Score (MASES), ∆ Work Productivity and Activity Impairment (WPAI), and ∆ ASAS health index (HI). Adverse events (AEs) were monitored throughout the study.

Results: All randomized pts (N=187) received assigned treatment (PBO, n=94; UPA, n=93); 95.7% of pts completed the study through wk 14 (PBO, 90/94; UPA, 89/93). Most pts were male (70.6%) and HLA-B27 positive (76.5%). Mean symptom duration was 14.4 y, and mean age was 45.4 y; BL disease characteristics were balanced between arms. Significantly more pts treated with UPA vs PBO achieved the primary endpoint of ASAS 40 response at wk 14 (51.6% vs 25.5%; P< 0.001; Figure). Accounting for multiplicity-adjustment, the following endpoints were statistically significant for UPA vs PBO at week 14: ∆ ASDAS, ∆ SPARCC MRI spine, BASDAI50, ASAS PR, and ∆ BASFI (Figure). Other ranked secondary endpoints, except WPAI, were significant based on nominal P values (Figure). The proportions of pts with AEs leading to discontinuation (2.2% vs 3.2%), serious AEs (1.1% vs 1.1%), and infections (20.4% and 27.7%) were similar for UPA and PBO groups, respectively. No serious infections, herpes zoster, malignancy, venous thromboembolic events, or deaths were reported.

Conclusion: UPA 15 mg QD was significantly more efficacious than PBO at wk 14 in pts with active AS for improvement in signs and symptoms, function, and imaging. The proportion of patients with AEs was similar in the UPA and PBO arms, and no new safety findings were observed compared with previous UPA studies in other diseases.