Researchers have developed a blood test that may revolutionize cancer testing. It can identify over 50 types of cancer, often long before symptoms develop.
Imagine being a woman concerned about a lump in your breast, or a man worried about possible symptoms of prostate cancer. Or consider the plight of people who have no idea that what they and their doctors often assume is chronic indigestion is actually pancreatic cancer, a malignancy with a high fatality rate because it’s very hard to diagnose early.
Those and countless other scenarios might play out in the not too distant future with less worry and better outcomes, thanks to a breakthrough in cancer screening.
The results of a study, published in Annals of Oncology, showed for the first time a blood test can accurately detect more than 50 types of cancer. What’s more, the test can also be used to pinpoint the tissue where the malignancy started, often long before obvious signs or symptoms of the disease.
Another clinical trial of the technique, presented at the American Society of Clinical Oncology’s annual meeting in Chicago, showed similar results. The study of the Galleri test, with 140,000 participants, was conducted by the National Health Service in England.
Although advances in cancer treatment save many lives, cancer remains a top killer of Americans. The National Cancer Institute reports cancer is responsible for the death of about 600,000 people yearly in the U.S. Finding an easy way to screen for malignancies is important for better detection and early treatment.
“This is a landmark study and a first step toward the development of easy-to-perform screening tools,” said Fabrice André MD, PhD, director of research at the Gustave Roussy Cancer Institute in Villejuif, France, and editor-in-chief of Annals of Oncology. “Earlier detection of more than fifty percent of cancers could save millions of lives every year worldwide and could dramatically reduce morbidity induced by aggressive treatments. “
How a blood test can spot cancer using AI
The investigation into finding a blood test to diagnose cancer is part of the ongoing international Circulating Cell-free Genome Atlas (CCGA) study, involving 15,000 participants from 142 clinics in North America.
For this arm of the CCGA research, an international team (including scientists from Mayo Clinic, the Dana Farber Cancer Institute, Cleveland Clinic, and the University College London Cancer Institute) used high-tech DNA analyses to study the blood of more than 4,000 study participants, about equal numbers with and without cancer.
They searched for specific markers related to changes in DNA. Tumors are known to shed DNA into the blood, contributing to what’s called cell-free DNA (cfDNA). But cfDNA doesn’t all come from cancer cells, so it can’t be used alone to pinpoint if cfDNA is the result of a malignancy.
That’s why the scientists zeroed in on analyzing chemical changes to DNA called methylation, which mostly control gene expression. Abnormal methylation patterns that make changes in gene expression can contribute to the development of tumors. So, the research team looked for signals in cfDNA indicating abnormal methylation, to see if they could be used to detect cancer.
They used a blood test, targeting about one million of the 30 million methylation sites in the human genome, and artificial intelligence (AI) in the form of a machine-learning classifier (an algorithm) that analyzed and predicted the presence of cancer and the type of malignancy, based on the patterns of methylation in the cfDNA shed from tumors.
The AI classifier was trained using what’s believed to be the world’s largest methylation database of cancer and non-cancer signals in cfDNA. The database is owned by GRAIL, Inc., a subsidiary of illuminia, a provider of DNA testing involved in funding the research that is battling the Federal Trade Commission and the European Union over its acquisition of GRAIL in 2021.
Results are extremely accurate
The results of the test turned out to be remarkably accurate in detecting more than 50 types of cancer and identifying in which tissue the cancer originated. The false positives for cancer were only .7 percent — which means if the blood test was used on patients, less than one percent would be told, wrongly, they had cancer.
Current methods of diagnosing cancer can result in far higher false positives. For example, about 10 percent of women who are screened with mammograms are incorrectly identified as having cancer, the researchers noted.
While more research is needed, the results indicate screening for cancer with a blood test could become a reality. It would likely be a game changer in the diagnosis and treatment of the deadliest types of cancer that are often found late:
- Head and neck
- Liver and bile duct
- Pancreatic cancers
- Cancers of white blood cells, such as multiple myeloma
June 06, 2023
Janet O'Dell, RN