Artificial intelligence (AI) could soon help rheumatologists tailor treatment decisions for patients with psoriatic arthritis (PsA) and axial spondyloarthritis (axSpA), according to a new study.
Researchers from Germany developed machine learning (ML) models capable of predicting which patients are most likely to achieve low disease activity (LDA) and high health-related quality of life (HRQoL) after 16 weeks of treatment with secukinumab, a biologic therapy targeting interleukin-17A.
Data-driven predictions
The findings come from the AQUILA study, a large, ongoing, multicentre, real-world investigation involving 1,961 patients with active PsA or axSpA. The AI-driven approach analysed baseline clinical, demographic, and laboratory data to forecast outcomes before treatment began.
The models used binary machine learning algorithms combined with explainable artificial intelligence (XAI) tools to not only predict outcomes but also interpret how individual factors influenced the results.
Key predictors identified
For PsA, the strongest predictors of achieving LDA included patient and physician global assessments, prior biologic treatment, tender joint count, and age. Predictors for high HRQoL included disease impact scores, depression levels, height, tender joint count, and body mass index (BMI).
In patients with axSpA, the most influential factors for LDA were disease activity index (BASDAI), prior biologic use, C-reactive protein (CRP) levels, ASAS Health Index, and height. High HRQoL was linked to better baseline functional scores, lower depression, and lower BMI.
Towards personalised care
The researchers emphasised that XAI added transparency to model predictions, allowing clinicians to understand why certain patients are more or less likely to respond to treatment.
The study represents a step toward AI-assisted clinical decision support systems that could refine therapy choices in chronic inflammatory diseases.
Reference
Vodenčarević A et al. Predicting treatment outcomes in patients with psoriatic arthritis or axial spondyloarthritis: An artificial intelligence–driven approach. J Rheumatol. 2025;DOI: 10.3899/jrheum.2025-0327.
