Prof Eleftherios Mamounas talks to ecancer at ESMO 2026 about a multimodal artificial intelligence (AI) tool designed to personalise extended endocrine therapy (EET) in patients with early hormone receptor-positive breast cancer.

Using data from 2,271 patients in the NSABP B-42 trial, researchers developed two complementary AI models to predict late distant recurrence risk and treatment benefit.

The models successfully identified distinct patient subgroups, with those classified as high-risk and high-benefit experiencing a significant 10% absolute benefit from EET (p=0.001).

Patients in the low-benefit group showed no significant treatment benefit.

The findings suggest that integrating clinicopathological data with AI-based digital pathology analysis could help identify patients most likely to benefit from extended endocrine therapy, supporting more personalised treatment decisions.