AI Breakthrough Predicts Bowel Cancer Drug Efficacy for NHS Patients
A groundbreaking artificial intelligence tool has been developed to forecast how patients with advanced bowel cancer will respond to a recently introduced NHS medication. This innovation aims to prevent thousands of individuals from undergoing treatments that prove ineffective against their specific cancers, thereby avoiding unnecessary side-effects.
Study Details and Patient Impact
The research, conducted by scientists at London's Institute of Cancer Research and the RCSI University of Medicine and Health Sciences in Dublin, monitored 117 European bowel cancer patients. All participants had previously received chemotherapy alongside bevacizumab, a drug approved by the NHS in December. Bevacizumab functions by inhibiting tumour growth through protein deprivation, yet it is only beneficial for a limited subset of patients and can induce serious adverse effects such as blood clots and gastrointestinal complications.
In the United Kingdom, approximately 10,000 new cases of advanced bowel cancer are diagnosed annually, with a notable increase among young adults. Bowel cancer ranks as the second deadliest cancer, trailing only lung cancer in mortality rates. While early detection can yield survival rates as high as 98%, the five-year survival rate for advanced stages plummets to as low as 10%.
How PhenMap AI Technology Works
The AI tool, named PhenMap—a blend of "phenotype" (referring to an organism's observable characteristics) and "mapping"—enables researchers to amalgamate intricate data regarding the genetic composition of tumours. This capability allowed the team to discern patterns in patient responses to bevacizumab and identify a specific group with a common gene mutation who faced elevated risks of negative reactions.
Professor Anguraj Sadanandam, a specialist in stratification and precision medicine at the ICR, emphasised the significance of this development. "Once bowel cancer metastasises, treatment options become severely limited," he stated. "It is encouraging that bevacizumab is now accessible via the NHS, but the reality is that most patients will not derive benefit, subjecting thousands to needless side-effects. Previously, we lacked the means to pinpoint these individuals."
Future Applications and Validation
Professor Sadanandam highlighted that the AI methodology processes vast amounts of complex data to reveal patterns imperceptible to human analysis, uncovering critical insights within tumour biology. The research demonstrated an ability to flag patients least likely to respond to bevacizumab therapy.
However, he cautioned that while the findings are promising, further validation through larger patient cohorts is essential. "I aspire for this approach to eventually yield a clinical test, enabling personalised care tailored to maximise therapeutic success against cancer," Sadanandam added. The research team now plans to expand sample sizes and explore the tool's applicability to other cancer types, potentially revolutionising oncology treatment strategies.
