#4-13E. Predicting the Unpredictable: AI and Colchicine Resistance in FMF (Rheumatology, 2026)

Can we foresee which patients with Familial Mediterranean Fever (FMF) will fail standard colchicine therapy before they suffer recurrent attacks and complications? This episode dives into a recent study that utilized machine learning and deep learning algorithms to predict colchicine resistance using data from 965 adult FMF patients. By analyzing clinical and genetic features, the researchers developed predictive models that achieved an area under the curve of 0.79, identifying factors like homozygous mutations, chronic arthralgia, and recurrent arthritis as significant clinical predictors. The most crucial clinical takeaway is that deep learning can effectively forecast this resistance, offering a sophisticated alternative to traditional clinical scoring systems. However, when critically appraising the paper, we must note its reliance on a retrospectively collected dataset from a single center, highlighting the need for external validation in larger, independent populations. For practicing physicians, this research is highly relevant because AI-assisted probabilistic modeling enables early, personalized treatment decisions, avoiding unnecessary delays in transitioning resistant patients to crucial biologic therapies.

Citation: Ozturk A, Ugurlu S. Machine-learning algorithms for predicting colchicine resistance in Familial Mediterranean Fever. Rheumatology. 2026. DOI: 10.1093/rheumatology/keag096

Disclaimer: This audio summary is based on personal interpretation and does not guarantee the exact content of the original paper. Please refer to the original article for details.