Researchers at the University of Oxford have unveiled TriOx, a groundbreaking blood test that leverages machine learning to detect cancer in its earliest and most challenging stages.
Published in Nature Communications, the study revealed that TriOx accurately identified cancer across six types, including early-stage cases, with 94.9% sensitivity and 88.8% specificity. This precision ensures patients without cancer avoid unnecessary procedures while enabling earlier treatment for those diagnosed.
Unlike most liquid biopsy tests that focus on one or two DNA features, TriOx analyses multiple features of cancer DNA using a cutting-edge technique called TAPS, combined with machine learning. This comprehensive approach enhances sensitivity by detecting the small fractions of cancer DNA circulating in the bloodstream.
“Our test combines cutting-edge science and machine learning to assess the entire cancer genome,” explained Professor Anna Schuh, lead researcher at the University of Oxford. “While still under development, we believe TriOx could significantly improve survival rates globally by enabling routine blood tests to catch cancer earlier, when treatment is most effective.”
The test was evaluated on blood samples from symptomatic patients and asymptomatic individuals, demonstrating its ability to reliably distinguish between those with and without cancer. Cancer types studied included colorectal, oesophageal, pancreatic, renal, ovarian, and breast.
Funded by Innovate UK and the NIHR Oxford Biomedical Research Centre, the project reflects years of collaboration between Oxford researchers and Exact Sciences. TriOx is licensed to SerenOx, a company co-founded by Professor Schuh.
Further research is underway to validate TriOx across additional cancer types and larger patient groups. The ultimate goal is to integrate this test into routine healthcare, making early cancer detection as accessible as cholesterol or blood sugar testing.
With its potential to revolutionise cancer diagnostics, TriOx offers a promising leap forward in improving early detection and survival outcomes.
Read the full story here.
