A new AI-powered tool may take the guesswork out of prostate cancer treatment decisions. Researchers developed a biomarker using histology images alone that predicted survival in men with metastatic hormone-sensitive prostate cancer — and flagged those most likely to benefit from triplet therapy. The tool showed strong results across two large clinical trials, offering a promising step toward more personalized cancer care.
Choosing the right treatment intensity for metastatic prostate cancer has long been a challenge — but a new AI-powered biomarker could change that. Researchers developed a computational histology tool that analyzes whole-slide diagnostic images to generate a risk score, sorting patients into favorable-risk (90%) and unfavorable-risk (10%) groups. The biomarker was built using data from 507 patients in the CHAARTED trial and validated in 684 patients from the ENZAMET trial.
The results were striking: unfavorable-risk patients had a 5-year overall survival rate of just 39%, compared to 69% for favorable-risk patients. Crucially, unfavorable-risk patients who received triplet therapy — androgen deprivation therapy (ADT) plus an androgen receptor pathway inhibitor plus docetaxel — saw significantly better survival than those on doublet therapy. Favorable-risk patients, meanwhile, did equally well on either regimen.
By the Numbers:
Why it matters: This tool could help oncologists move beyond one-size-fits-all treatment decisions, sparing low-risk patients from unnecessary chemotherapy while ensuring high-risk patients get the aggressive treatment they need — all from a standard tissue sample.
