Cancer is difficult to detect and even more difficult to treat. Often, the disease is detected very late leaving little time for effective treatments. But that might be about to change. In a recently published study, researchers were able to detect cancer by using a blood test, even in people who exhibited no symptoms.
Diagnosis of cancer involves taking a sample of the cancerous tissue from the patient and analyzing it in the lab. In cases, where a sample cannot be obtained, clinicians rely on imaging techniques to make a diagnosis. However, either of these approaches is used only after the patient displays certain symptoms that point in the direction of the disease.
Blood-based cancer detection methods use a very different approach. It has been known that our body cells shed the genetic material - the DNA, into blood on a regular basis. This is called cell-free DNA or cfDNA. Cancerous cells are no exception, rather they do this more frequently. Interestingly, cfDNA of cancerous cells carries with it chemical changes, that can aid early cancer detection. Called methylation patterns, these chemical changes can also provide information on which organ in the body, the cfDNA has originated from, providing valuable information on the location of cancer. Companies like Grail and many others are leveraging this information to detect cancer, sometimes even years before its symptoms.
Typically, a test is designed to be specific - has a narrow purpose and sensitive - detects very low amounts of the disease. Out to solve healthcare's most important challenges, as it claims its website, Grail set itself an ambitious task of detecting over 50 types of cancers, thereby broadening its purpose while yet retaining the ability to detect very small amounts of disease. Last year, Grail published the preliminary data on how they had trained a 'classifier' that could take in the data from the tests and be able to identify which samples exhibited methylation patterns of early cancer.
The company tested the performance of this 'classifier' using over 2,823 known cancer samples and 1,254 samples that were known to not have cancer. The results of this performance study were published in Annals of Oncology.
The company was able to detect over 50 types of cancer, irrespective of the stage of cancer in 51.5 percent of the cases. While this might seem like a low number, the specificity of the test is remarkable. It wrongly identified cancer in only 0.5 percent of cases, meaning the likelihood of the test missing out on cancer was very low, making it a good screening tool.
The detection of cancer progressed with stages showing 16.1 percent for stage I while 90.1 percent detection of stage IV cancer. More importantly, the detection capacity of the test was twice as better for solid tumors such as those that occur in the esophagus, liver, and pancreas and that currently do not have any screening methods available than for tumors of the breast, bowel, and prostate that have screening methods available.
Dr. Eric Klein, the first author of the paper and who works at Cleveland Clinic, USA said, "Cancers that shed more cfDNA into the bloodstream are detected more easily and also more likely to be lethal. Cancers such as prostate shed less DNA than other tumors, which is why existing screening tests are still important for these cancers"
Grail has planned a pilot study of its test with the National Health Service (NHS) in the UK with 140,000 participants, results for which are expected by 2023, The Guardian reported