A next-generation mt-sDNA test had 94% sensitivity for colorectal cancer and 91% specificity for advanced neoplasia

An industry-funded study found that a next-generation multitarget stool DNA (mt-sDNA) test with assessments for DNA molecular markers and hemoglobin level had approximately 94% sensitivity for colorectal cancer and approximately 91% specificity for advanced neoplasia versus colonoscopy. Fecal immunochemical testing (FIT), meanwhile, had a sensitivity of approximately 67% for colorectal cancer and approximately 95% for advanced neoplasia.

The study involved 20,176 asymptomatic U.S. adults undergoing colonoscopy at 186 sites. It was published March 13 by the New England Journal of Medicine and summarized in the March ACP Gastroenterology Monthly.

The following commentary by Sander Veldhuyzen van Zanten, MD, MSc, MPH, PhD, and Ernst J. Kuipers, MD, PhD, was published in the ACP Journal Club section of Annals of Internal Medicine on Sept. 3.

Stool tests for colorectal cancer screening provide an a priori likelihood for advanced neoplasia and can help select patients for colonoscopy. Sensitive and specific stool tests can identify a high proportion of patients with advanced neoplasia while reducing colonoscopy demand. They also often increase screening participation.

Test sensitivity and specificity depend on the selected cut point. FIT cut points are set at selected Hb levels (Hb mcg/feces g). mt-sDNA testing quantitatively measures DNA markers and fecal Hb. Using a logistic regression algorithm, the results are translated into a number, which is used to select a cut point. Similar to their previous study, Imperiale and colleagues found that a mt-sDNA test was more sensitive than FIT for advanced neoplasia detection. The new mt-sDNA test also had higher specificity than the previous test, with results close to the FIT. However, both studies were imbalanced in cut point selection for the stool tests, which would have led to a colonoscopy referral rate in everyday practice about 2-fold higher after mt-sDNA testing than FIT (13.4% vs. 7.4% in the current study [calculated from data in Table S10 of the article's supplement] and 16.1% vs 7.0% in the previous study). The question is whether mt-sDNA is a better test than FIT or outperformed because of a cut point that would lead to a lower threshold for colonoscopy referral. To address this question, the performance of mt-sDNA testing and FIT should be compared at the same colonoscopy referral rate, as done in other studies. The design of BLUE-C allows for this analysis to be done. We expect that there will still be a difference in favor of mt-sDNA testing, but the magnitude and clinical relevance remain to be determined.

Imperiale and colleagues' study is important and underlines the relevance of mt-sDNA testing for colorectal cancer screening. We await results of the comparison between mt-sDNA testing and FIT for a range of the same positivity rates. This will determine the true value of mt-sDNA testing over FIT.