AI Model Predicts Late Recurrence Risk in Breast Cancer Patients

The research, a collaboration between Caris Life Sciences, the NSABP Foundation/NRG Oncology, and the ECOG-ACRIN Cancer Research Group, addresses a persistent clinical hurdle: identifying which patients truly benefit from extended endocrine therapy. While current standard care involves five years of treatment, the risk of recurrence often continues long after, forcing clinicians to weigh the necessity of additional therapy against its significant side effects.

Developed using 2,271 tumor specimens from the NSABP B-42 trial, the AI model demonstrated its predictive power by identifying a nearly 8% difference in 10-year absolute distant recurrence risk between high- and low-risk groups. External validation using 4,300 specimens from the TAILORx study confirmed that the model effectively predicts recurrence risk independently of existing clinical factors and the Oncotype DX Recurrence Score. Dr. George W. Sledge, Chief Medical Officer at Caris, noted that the tool transforms routinely collected pathology data into actionable insights for personalized treatment plans. By providing a scalable alternative to costly genomic assays, this AI-driven approach offers a practical path for oncologists to refine decisions regarding long-term therapy.