Background: Ankylosing spondylitis (AS) is a progressive, inflammatory arthritis associated with pain and physical impairment. Standard treatment for patients with active AS and an inadequate response to conventional therapy comprises biologic agents, for which assessment of response is important to determine treatment continuation. Previous health economic models presented to the National Institute for Health and Care Excellence (NICE) have defined treatment response as a 50% reduction from baseline in the Bath Ankylosing Spondylitis Disease Activity Index (BASDAI 50), a readily reported clinical trial outcome. However, British Society for Rheumatology (BSR) and NICE guidelines define treatment response as BASDAI 50 or a ≥2 unit drop in BASDAI total score from baseline, AND a ≥2 cm reduction in spinal pain visual analogue scale (VAS). This research explored the impact of different treatment response definitions on response rates and long-term outcomes using data from two randomised controlled trials of secukinumab in active AS.
Methods: Data for tumour necrosis factor alpha (TNFα) inhibitor-naïve patients treated with secukinumab 150 mg in the MEASURE 1 (n=92) and MEASURE 2 (n=44) trials were analysed to determine response rates at week 12 according to three definitions (Table 1). Mean BASDAI and Bath Ankylosing Spondylitis Functional Index (BASFI) scores up to week 104 were evaluated by response definition.
Results: Use of BASDAI 50 alone to define treatment response resulted in fewer classified responders, but also the greatest reductions in mean BASDAI and BASFI scores from baseline to week 104 across both trials (Table 1). Broadening the response criteria to also include a ≥2 unit reduction in BASDAI resulted in an expanded pool of responders, but reduced mean BASDAI and BASFI changes from baseline over time. The BSR/NICE definition (inclusion of a ≥2 cm reduction in spinal pain VAS) resulted in mean BASDAI and BASFI changes from baseline between those observed for the other two definitions.
Conclusion: Use of different treatment response criteria has significant implications for response rates and responder clinical outcomes. Standardisation of outcome measures and study design is needed to minimise bias in interpretation. The exploration of further response criteria is an area for potential future research.
