The study, published in Nature, uses a unique combination of liquid biopsy - a test that examines blood samples to determine if there are any traces of cancer DNA circulating in the blood - epigenetic alterations and machine learning. By profiling epigenetic alterations, rather than mutations, the team was able to identify thousands of modifications that are unique to different types of cancer. Then, using a machine learning approach they could create classifiers that recognize cancer cells in blood samples and determine the kind of cancer a patient is suffering from.The scientists compared the type and origin of 300 tumour samples from seven disease sites (lung, pancreatic, colorectal, breast, leukemia, bladder and kidney) with samples from healthy donors by analyzing cell-free DNA circulating in the blood plasma. In every test the circulating DNA matched the tumour DNA.
Dr. De Carvalho, Senior Scientist at Princess Margaret Cancer Centre and principal investigator of the study, said: "We are very excited at this stage. A major problem in cancer is how to detect it early. It has been a 'needle in the haystack' problem of how to find that one-in-a-billion cancer-specific mutation in the blood, especially at earlier stages, where the amount of tumour DNA in the blood is minimal."
The team is now testing their method in samples collected from healthy people that ended up developing cancer over time, including on samples from 200,000 people profiled in The Canadian Partnership of Tomorrow Project. “Over time, many of these healthy people went on to develop cancer," Dr. De Carvalho said. "The project has a blood sample from these people before they were diagnosed with cancer. So now we can test our approach on patients from this project six months to three years before the cancer develops. If we can already discover the cancer at this stage we know we have the sensitivity to perform very early-stage detection.”As a next step the researchers will bring this test to the general population.
