178P Interpretable AI to predict anal cancer local failure at 3 years with the FFCD Anabase prospective multicentric cohort

Machine learning (ML) has transformed oncological risk prediction by enabling personalized therapeutic strategies. Local tumor control remains a critical endpoint in anal cancer management. This study aimed to develop and validate an explainable ML model for predicting local recurrence at 3 years in patients with anal cancer.