Stool DNA Test Accurately Detects Colorectal Cancers

March 26, 2014

A new study reports that a multitarget stool DNA test is sensitive and accurate and is an improvement over previous DNA stool-based detection tests.

The American Cancer Society’s recent analysis shows that colorectal cancer incidence rates are decreasing among older adults due to colonoscopy screenings. Early screening can find precancerous polyps and early-stage tumors that could potentially be cured with treatment. Still, there is room for improvement in colorectal cancer screening rates. The US Preventive Services Task Force recommends screening by fecal occult blood testing, flexible sigmoidoscopy, or colonoscopy for adults between the ages of 50 and 75.

Now, in an article in the New England Journal of Medicine, Thomas F. Imperiale, MD, a gastroenterologist at Indiana University in Indianapolis, and colleagues report that a multitarget stool DNA test is sensitive and accurate-an improvement over previous DNA stool-based detection tests. This test detected significantly more colon cancers compared with the commercially available fecal immunochemical test (FIT) in average-risk patients between the ages of 50 and 84 in a large prospective clinical trial funded by Exact Sciences, manufacturer of the test. However, the specificity of the new test is lower compared with FIT.

The prospective trial enrolled 12,776 participants throughout 90 sites in the United States and Canada. Of these participants, 78.2% (9,989) had DNA test results that could be fully evaluated. All patients were evaluated by colonoscopy.

Sixty-five patients (0.7%) had colorectal cancer and 757 patients (7.6%) had advanced precancerous lesions detected by colonoscopy. Of the 65 patients with colorectal cancer, 60 had screening-relevant cancer that was stage I to III disease.

The DNA test detected 92.3% of colorectal cancers compared with 73.8% with FIT (P = .002), demonstrating the overall sensitivity of the test. The sensitivity with the DNA test was 93.3% for screening-relevant lesions compared with 73.3% with FIT (P = .002). Of the 60 participants with screening-relevant cancer, 56 were identified by the DNA test (60 of the 65 total).

The DNA test could more accurately detect distal colon cancers compared with proximal colon cancers.

The DNA test was able to detect 42.4% of advanced precancerous lesions, while FIT detected 23.8% (P < .001). The detection rate of polyps with high-grade dysplasia was 69.2% with the DNA test compared with 46.2% with FIT (P = .004). Rates of detection of sessile serrated polyps measuring 1 cm or more were 42.4% with the DNA test compared with 5.1% with FIT (P < .001).

FIT had fewer false positive results compared with the DNA test. The specificity of the DNA test was 86.6% compared with 94.9% with FIT among those who had nonadvanced tumors or negative colonoscopy results (P < .001), and 89.8% and 96.4%, respectively, among those who only had a negative colonoscopy result (P < .001).

Still, the DNA test resulted in fewer people that would need to be screened to detect a single colorectal cancer: 166 people needed to be screened with the DNA test compared with 208 with FIT. The number of people who needed to be screened by colonoscopy to detect a single cancer was 154.

The DNA test is a molecular assay that detects promoter regions of the BMP3 and NDRG4 genes for aberrant methylation, as well as for mutations in the KRAS gene. The β-actin gene is used as a reference gene to quantify the amount of human DNA in a given sample. The test also includes an immunochemical assay for human hemoglobin.

In an editorial that accompanied the study, Douglas J. Robertson, MD, of the Veterans Affairs Medical Center in White River Junction, New Hampshire, and Jason A. Dominitz, MD, of the Veterans Affairs Puget Sound Healthcare System in Seattle, pointed out that while the sensitivity of the DNA test is improved, there are caveats to the current results. First, 6.3% of the participants were excluded from DNA stool sample analysis due to problems with sample collection or with getting results of the assays included in the test. This is compared with only 0.3% of participants who were not able to get results with their samples using FIT. “Given that colorectal cancer was detected in nearly one of 154 participants on colonoscopy, it is possible that four cancers would have been missed simply because of the complexity of the test,” stated the authors.

Second, the current study only compared a one-time sensitivity of the DNA and FIT tests, while previous studies have shown that repeated FIT improves detection, suggesting that a longer-term study of a series of tests would better assess the specificity and sensitivity of these two tests. Third, the study showed that 10% of participants had a positive DNA test but a negative colonoscopy. According to the editorial authors, decision models are needed to “guide how frequently this test should be used as a screening tool to be cost-effective.”

Robertson and Dominitz pointed out that compliance may be in favor of the DNA test compared with colonoscopy and FIT, but further studies or real-world data are needed to determine if this is indeed the case. They called for comparative-effectiveness studies to better understand where stool DNA testing fits in with other colorectal cancer screening regimens.