Scientists have unveiled an AI-based approach to determine how individuals with advanced bowel cancer might react to a medication newly adopted by the NHS. This innovation, created by experts from the Institute of Cancer Research in London and RCSI University of Medicine and Health Sciences in Dublin, aims to prevent unnecessary administration of ineffective treatments to many patients. Each year in the UK, approximately 10,000 people receive diagnoses of advanced bowel cancer, with increasing cases among younger adults. This form of cancer ranks second in lethality after lung cancer. Early detection yields survival rates up to 98%, but for advanced stages, the five-year survival drops to about 10%. The research examined 117 European patients treated with chemotherapy combined with bevacizumab, a drug the NHS approved in December. Bevacizumab inhibits tumor growth by blocking essential proteins, yet it benefits only a limited number of patients and can cause severe side effects like blood clots and digestive problems. Employing an AI system called PhenMap, which combines ‘phenotype’ and ‘mapping,’ the team integrated detailed genetic information from tumors. This enabled them to detect response patterns and pinpoint a subgroup with a shared gene mutation linked to poor outcomes. The researchers plan to analyze more patient data and explore applications for other cancers. Anguraj Sadanandam, a professor specializing in precision medicine at the ICR, noted that options are scarce once bowel cancer metastasizes. He welcomed NHS access to bevacizumab but highlighted that most patients gain no advantage, leading to avoidable side effects for thousands in England. Previously, identifying non-responders was challenging. The AI technique compiles vast datasets to reveal hidden patterns in tumors, identifying those unlikely to benefit from the drug. While promising, Sadanandam stressed the need for validation with larger groups. He envisions this leading to clinical tests for tailored treatments maximizing effectiveness against cancer.
