A new AI-powered test offers more accurate prediction of breast cancer recurrence risk

A new artificial intelligence (AI) model called IICM+ has been developed to more accurately predict the risk of breast cancer recurrence in patients with early-stage, hormone receptor-positive (HR+) and HER2-negative disease, according to a study published in the journal npj Breast Cancer. Researchers from the ECOG-ACRIN Cancer Research Group and Caris Life Sciences found that IICM+ outperformed the widely used 21-gene Recurrence Score in distinguishing patients at different risk levels, particularly for late recurrence that may occur more than five years after initial surgery.

The study, which analyzed data from 4,429 participants in the TAILORx breast cancer trial, combined digitized pathology images of the tumor with patient age, tumor size, grade, and molecular information from a 42-gene panel. While the Recurrence Score is effective at predicting recurrence within the first five years, the IICM+ model, created using multimodal data and modern AI techniques, is better at predicting distant recurrence later in the disease course.

The study's lead author, Joseph A. Sparano, MD, emphasized the potential of IICM+ to create more individualized breast cancer care, although further studies are needed to validate its clinical utility for treatment decisions.