James Hamrick, Chair of Precision Oncology Alliance at Caris Life Sciences, shared on LinkedIn:
“For women with hormone receptor-positive breast cancer, the five-year mark on endocrine therapy brings a difficult decision. Recurrence risk can persist beyond that point, but extending treatment also means many years of additional side effects. For years, it has been difficult to know which patients are most likely to benefit.
A study from Caris Life Sciences developed with the NSABP Foundation Inc., NRG Oncology, and the ECOG-ACRIN Cancer Research Group and published in Cancer Research Communications, tested whether AI could help answer that question.
The model combines standard pathology slides alongside routine clinical data to estimate the risk of late recurrence years after treatment ends. Developed using 2,271 specimens from the NSABP B-42 trial and validated in 4,300 from TAILORx, it predicted late recurrence risk independently of established clinical factors and the Oncotype DX Recurrence Score. In an exploratory analysis, patients identified as higher risk appeared more likely to benefit from extended therapy.
What is especially promising is that the signal came from information already collected during routine care – a pathology slide and clinical data – suggesting AI could surface information current tools may not fully capture. This is early research, and it is not a test physicians can order today, but it points toward a future where one of the harder decisions in breast cancer care can be informed by a more precise understanding of each patient’s risk.”