Richard M. Goldberg, MD: Stool DNA Test Technology Has Promise
« RELATED FEATURE »
A multitarget stool DNA test detected more colorectal cancers (CRCs) than other testing but also resulted in more false-positives, according to a new study. Researchers and clinicians are now examining the test's future with the FDA as well as its potential to change screening practices. Read more
Richard M. Goldberg, MD
I have been following the development of this multitargeted stool DNA test for about 15 years. Preliminary studies have been published previously, but researchers have made some substantial advances in this area.
Identifying a way to properly test the sample, for instance, is on important improvement. Most of the DNA in stool is either bacterial DNA or DNA from healthy parts of the gastrointestinal tract, so one could say that looking for human DNA, particularly DNA related to a polyp or cancer, is very difficult. With this test, however, researchers have found a way to isolate the DNA so that it can be tested.
Researchers have also refined and narrowed down the genes that are most likely to show evidence of a polyp that could potentially become malignant. Further, they have sorted out how to analyze the information to get the most power from the available data.
Finally, one other major advance includes combining this gene test with a hemoglobin assay. This makes for a more powerful test.
In this particular study published in The New England Journal of Medicine, Imperiale et al collected about 10,000 patient samples before colonoscopy. Stool samples were tested using a standard fecal immunochemical test (FIT) or the DNA test.
What they found was that the DNA test, coupled with the hemoglobin test, was more sensitive and more specific compared with the FIT. However, the DNA test also yielded more false-positives. Specifically, the data showed that about 732 of 1,611 positive tests for DNA testing turned out to be false-positives, which is a substantial amount compared with the 248 of 696 with FIT.
Ideally, clinicians want a test that is highly sensitive, highly specific, and has a very low false-positive rate. In this case, however, a false-positive means that a clinician will tell a patient that he or she ought to have a colonoscopy, so it does not mean that that patient has cancer and it goes unaddressed.
Clearly, neither test is perfect, but a clinician would end up performing more colonoscopies when using the FIT in people who do not have malignancies. Moreover, the purpose of this type of test is to get individuals who won't undergo colonoscopy as an initial test to be screened because it does not entail what a colonoscopy does. By doing so, we would still eliminate about 8,500 people of 10,000 from having a colonoscopy at all. That equates to significant savings both in terms of cost as well as inconvenience for the patient.