AI-powered CMBs provide a New Approach Methodology for precision oncology

AI-discovered cellular morphometric biomarkers in needle biopsy of prostate cancer predict neoadjuvant androgen deprivation therapy response and enable therapeutic targeting of mTOR in androgen deprivation therapy-resistant tumors

Researchers at Berkeley Lab have led the development of an artificial intelligence-based biomarker that predicts neoadjuvant androgen deprivation therapy response in prostate cancer using standard histopathology images. The approach, termed a 13-cellular morphometric biomarker (13-CMB), identifies a tumor state associated with resistance to androgen deprivation therapy.

This state is characterized by activation of the mTOR pathway and DNA damage response signaling: features not readily detectable by conventional pathology but revealed through AI-driven analysis, which enables therapeutic targeting of mTOR in androgen deprivation therapy-resistant tumors.

The study, recently published in Cancer Letters, demonstrates how AI-enabled digital pathology can serve as a New Approach Methodology (NAM), providing scalable and clinically relevant biological insights without additional assays.

The findings support a broader framework of Genesis AI, integrating tissue phenotypes, molecular mechanisms, and predictive modeling to enable mechanism-informed precision health.