Overview
- Researchers at the Institute of Cancer Research in London and RCSI in Dublin reported in Scientific Reports an AI method called PhenMap that links tumor DNA and clinical factors to predict response to bevacizumab.
- The study examined 117 European patients who received bevacizumab with chemotherapy and then used a second AI tool to sort patients into high, moderate, or low risk of death after treatment.
- No patients in the high‑risk group responded to therapy, and patterns that included a BRAF gene mutation were tied to poor outcomes.
- The approach could spare patients an ineffective drug that the NHS approved in December and that can cause high blood pressure, gut problems, and blood clots.
- The team plans to test the model in larger and more diverse groups and to build a trial‑ready test, with the hope that the method could also help match other targeted drugs to the right patients.