Efficacy and safety of continuing versus withdrawing adalimumab therapy in maintaining remission in patients with non-radiographic axial spondyloarthritis (ABILITY-3): a multicentre, randomised, double-blind study

Background: Success of treatment withdrawal in patients with non-radiographic axial spondyloarthritis who are in remission remains unknown. The ABILITY-3 study explored the ability to withdraw adalimumab treatment in patients with non-radiographic axial spondyloarthritis who achieved sustained clinical remission after open-label treatment with adalimumab. Methods: ABILITY-3 was a multicentre, two-period study done in 107 sites in 20 countries. We enrolled adult patients (≥18 years) diagnosed with non-radiographic axial spondyloarthritis, fulfilling Assessment of SpondyloArthritis international Society classification criteria but not the modified New York radiologic criterion, who had objective evidence of active inflammation, active disease, and inadequate response to at least two non-steroidal anti-inflammatory drugs. Patients who achieved Ankylosing Spondylitis Disease Activity Score (ASDAS) inactive disease (<1·3) with open-label adalimumab (40 mg subcutaneously every other week for 28 weeks) at weeks 16, 20, 24, and 28 were randomly assigned (1:1) using an interactive voice or web response system to 40-week, double-blind treatment with adalimumab (continuation) or placebo (withdrawal). The primary efficacy endpoint was the proportion of patients who did not experience a flare (defined as ASDAS ≥2·1 at two consecutive visits) during the double-blind period. Patients who flared were rescued with open-label adalimumab. This study is registered with ClinicalTrials.gov, number NCT01808118. Findings: Between June 27, 2013, and October 22, 2015, 673 patients were enrolled to the study. The trial completed on April 14, 2017. Of 673 enrolled patients, 305 (45%) achieved sustained remission and were randomly assigned to double-blind treatment (152 patients to adalimumab and 153 to placebo). A greater proportion of patients continuing adalimumab than those receiving placebo did not experience a flare (107 [70%] of 152 patients vs 72 [47%] of 153 patients; p<0·0001) up to and including week 68. Among 673 patients receiving adalimumab at any time, 516 (77%) patients reported an adverse event and 28 (4%) experienced a serious adverse event. The most common adverse events in both the adalimumab and placebo groups were nasopharyngitis (25 [16%] vs 20 [13%]), upper respiratory tract infection (20 [13%] vs 12 [8%]), and worsening of axial spondyloarthritis (ten [7%] vs 21 [14%]). Interpretation: In patients with active non-radiographic axial spondyloarthritis who achieved sustained remission with adalimumab, continued therapy was associated with significantly fewer patients flaring than was treatment withdrawal. Funding: AbbVie.