Is a False-Positive mt-sDNA Test (Cologuard) Actually Not False?

An Unexpected Signal from Colorectal Cancer Screening: Aerodigestive Cancers — DDW 2026 Wisconsin Cohort

Cologuard's three biomarker layers

The multi-target stool DNA test (mt-sDNA), approved by the FDA in 2014, is now one of the most widely used noninvasive colorectal cancer screening tests worldwide. Its commercial name is Cologuard (Exact Sciences, Madison, Wisconsin). The test measures three parallel biomarkers in a stool sample:

• Occult blood (FIT component): Detection of hemoglobin in stool — a classic indicator of any bleeding GI lesion. This is the conventional FIT (fecal immunochemical test) component.
• KRAS somatic mutation panel: PCR detection of seven specific mutations in codons 12 and 13 of the KRAS gene. KRAS mutations are common in the early stages of the colorectal adenoma-carcinoma sequence — but they are also detected in lung adenocarcinoma, pancreatic ductal adenocarcinoma, biliary tract tumors and gastric adenocarcinoma and many other aerodigestive cancers.
• Abnormal DNA methylation panel: Detection of hypermethylation in the promoter regions of two tumor suppressor/signaling genes — BMP3 (Bone Morphogenetic Protein 3) and NDRG4 (N-Myc Downstream Regulated Gene 4). This methylation pattern is far from specific to colorectal neoplasia; it is also observed in gastric, pancreatic and lung cancers.

This is where the fundamental biological rationale of the DDW 2026 study lies: The mt-sDNA test is called a "colorectal cancer test," but the molecular signature it detects — KRAS mutation, BMP3/NDRG4 methylation, blood in stool — does not have to originate in the colon. These signatures may also come from tumor DNA descending from the upper GI tract into stool, tumor DNA in swallowed sputum from the lungs, or KRAS-mutant DNA flowing from the pancreatic duct into the duodenum. Thus, the term "false positive" may be correct in the colorectal context but misleading in the aerodigestive context.

A 10-year single-center cohort at the University of Wisconsin

Kelly and colleagues retrospectively analyzed patients screened with mt-sDNA in the University of Wisconsin health system between 2014-2024, numbering 54.361 patients . The data sources were the institution's electronic health record system (EPIC) and billing codes.

The demographic characteristics of the false-positive and negative groups were as follows:

The lung, upper GI and head-and-neck axis

The distribution of cancers over 3 years of follow-up was as follows:

Three possible explanations

First mechanism — swallowed DNA. A tumor in the upper aerodigestive system — for example, a squamous cell carcinoma at the base of the tongue, expectorated material from a lung adenocarcinoma, or an esophageal/gastric tumor — either bleeds or sheds cells that descend into the GI tract. Swallowed tumor DNA makes a long journey to the stool; although most is broken down by digestive enzymes, a small amount may pass into the stool. This DNA may be detected by the mt-sDNA KRAS mutation panel or BMP3/NDRG4 methylation analysis.

Second mechanism — a shared molecular signature. KRAS mutations are not specific to colorectal cancer; they occur in 25-30% of lung adenocarcinomas, 90-95% of pancreatic cancers, 5-15% of biliary tract tumors and a small proportion of gastric cancers. BMP3 and NDRG4 methylation can likewise be detected across a broad tumor spectrum. Thus, the test may actually behave as a pan-aerodigestive molecular screening test.

Third mechanism — shared risk factors. Colorectal cancer, lung cancer, upper GI cancer and head-and-neck cancer share risk factors: aging, smoking, alcohol, obesity, chronic inflammation and immunosuppression. mt-sDNA positivity may actually be a proxy marker indicating that the individual belongs to a cohort characterized by smoking, exposure to tobacco smoke or other carcinogens. Important note: the Kelly team could not include smoking history in their study; this is an important confounder.

Important note: This approach is not yet part of any international guideline. The Kelly study is based on single-center retrospective data; it cannot become a standard of practice without prospective validation, multicenter studies and cost-benefit analyses. However, when decisions on additional investigations are based on individual risk factors — especially in false-positive patients aged 60+ with a heavy smoking history — low-dose CT lung screening and upper endoscopy are worth considering.

This study has additional importance for cancer epidemiology in Turkey: lung cancer is the country's third most common cancer, with approximately 35.000 new cases annually; gastric cancer accounts for 12.000-15.000 new cases annually (above the OECD average), and head-and-neck cancers for 10.000+ new cases. All three cancer categories identified by the Wisconsin cohort therefore have substantial epidemiological importance for Turkey. If mt-sDNA testing becomes widespread in Turkey at some point — especially through international patient centers — this literature will need to be followed when interpreting "false-positive" results.

Three practical points for clinical practice:

First, updating the language of physician-patient communication. The statement "It was a false positive, there is nothing wrong, don't worry" should now be a little more nuanced. The appropriate statement is: "Screening for colon cancer was negative, which is very good news. However, the signal from our test may be associated with other aerodigestive cancers — so it would be useful to assess your smoking history, family risk profile and other symptoms."

Second, selective additional investigations based on risk. If there is a high-risk profile (for example: age 60+ + 30+ pack-years of smoking + false-positive mt-sDNA), low-dose CT lung screeningis already indicated; in these patients, a false-positive mt-sDNA result may encourage the use of LDCT. The indication for upper endoscopy should be assessed together with conventional findings such as dyspepsia, GERD, B12/iron deficiency and a family history of gastric cancer — not solely on the basis of "false-positive mt-sDNA."

Third, future research. These findings need to be confirmed in prospective multicenter studies. As Kelly suggests, mt-sDNA may actually be a test with multi-cancer early detection (MCED) potential — a screening signal not only for colorectal cancer, but also for upper GI and lung cancers. Alongside blood-based MCED tests such as Galleri, Adela and GRAIL, the concept of a stool-based MCED platform is very interesting. However, this direction is still at the hypothesis stage.

• Single-center retrospective design. Wisconsin's predominantly White ethnic distribution differs even from that of the United States overall; extrapolation to Turkey or other countries is limited.
• Smoking history was excluded. The most important confounder; it could explain a large proportion of the findings.
• Identification of cancer based on billing codes. Positive cases were confirmed, but negatives were not; some cancers may have been missed.
• No distinction was made between asymptomatic and symptomatic individuals. The symptom status of the tested individuals was unknown at the time of the presentation.
• The FIT component could not be distinguished from the DNA component. It was not determined which component produced the positive result (occult blood, KRAS mutation or methylation).
• No difference in overall survival. Although cancer-independent survival was significantly lower, there was no difference in overall survival — the long-term clinical impact is therefore uncertain.

This article is intended for medical news/information and does not provide individual treatment guidance. The DDW 2026 presentation is a conference abstract that has not undergone peer review; results may be revised when it becomes a full publication. The mt-sDNA test (Cologuard) is not covered by SGK in Turkey; the national colorectal cancer screening program is based on FIT and colonoscopy through the KETEM network. Individuals with a family history, a smoking history and conventional symptoms of aerodigestive cancer should be assessed by a multidisciplinary oncology board. Editor's note: The article is based on the DDW 2026 presentation by Kelly and colleagues, the Medscape conference report and the pivotal Cologuard NEJM 2014 reference.