In a groundbreaking study, researchers from Weill Cornell Medicine, NewYork-Presbyterian, the New York Genome Center, and Memorial Sloan Kettering Cancer Center have developed an artificial intelligence-powered blood test that can predict cancer recurrence with remarkable sensitivity.
Published in Nature Medicine on June 14, the study reveals how a machine learning model was trained to detect circulating tumour DNA (ctDNA) with high accuracy. This method has shown success in lung cancer, melanoma, breast cancer, colorectal cancer, and even precancerous colorectal polyps.
Dr. Dan Landau of Weill Cornell Medicine stated, “We’ve achieved significant signal-to-noise enhancement, allowing us to detect cancer recurrence months or years before traditional methods.”
The technology represents a leap forward in liquid biopsy techniques. Traditional methods often miss sparse cancer-associated mutations in blood, but Dr. Landau’s team used whole-genome-sequencing for more sensitive detection.
Their system, MRD-EDGE, employs advanced machine learning to identify cancer patterns in DNA sequencing data. In tests, MRD-EDGE accurately predicted residual cancer post-surgery and chemotherapy in colorectal cancer patients without any false negatives.
MRD-EDGE’s sensitivity extends to early-stage lung and triple-negative breast cancers and can even detect ctDNA from precancerous colorectal adenomas—a significant advance for early detection strategies.
Dr. Landau emphasised the potential of this technology to transform cancer care by detecting recurrences and monitoring tumour response during therapy at unprecedented early stages.
