For two decades, tests like Oncotype DX have helped oncologists decide which early-stage breast cancer patients should receive chemotherapy. But those tools were designed mainly to predict early recurrence, leaving physicians with far less clarity about the risk that cancer might return years later.
For today’s program, George Sledge, Chief Medical Officer at Caris Life Sciences, discusses new findings from the TAILORx trial showing how multimodal AI—combining molecular sequencing, digital pathology, and clinical data—can improve long-term prediction of breast cancer recurrence.
Sledge explains that breast cancer recurrence may actually reflect two different biological processes unfolding over time. Molecular signals captured through RNA analysis appear most informative for predicting recurrence in the first five years, while computational analysis of digital pathology images becomes especially powerful for predicting recurrence later in the disease course.
“The best results come from looking at multiple omic levels,” Sledge says, describing a shift away from single biomarker tests toward integrated biological analysis.
