PREDICTION OF LOW DISEASE ACTIVITY IN PATIENTS WITH ANKYLOSING SPONDYLITIS TREATED WITH SECUKINUMAB IN REAL WORLD – DATA FROM THE GERMAN AQUILA STUDY
Abstract: OP0058
Authors: A. Vodencarevic et al.
Key content:
Treatment with biologics is effective in a majority of patients. However, there are no predictors for the response to an individual biologic to guide the personalized choice of drugs in axSpA. This study included patients with ankylosing spondylitis (AS) treated with Secukinumab in a German real world non-interventional study (Aquila). Predictors for achievement of low disease activity (LDA), as defined by BASDAI ≤2 at week 16, were analyzed using machine-learning methods. 10 different prediction models were compared and the importance of each prediction factor in individual prediction was estimated by explainable artificial intelligence. BASDAI at baseline, the number of pretreatments with biologics, CRP, the ASAS health index (ASAS-HI) and the height of the patient were the most important predictors. Lower BASDAI, number of pretreatments and ASAS-HI and higher CRP and height predicted LDA at 16 wks.
Relevance:
This study reports predictive factors for a response to therapy with Secukinumab in patients with AS. The predictors found match with existing data. However, the innovative approach with artificial intelligence additionally allowed to calculate patient-individual prediction of therapy response. Such an approach may allow in the future to guide treatment decisions in AS and possibly also in other rheumatic diseases.
