A simple mouth rinse may carry molecular clues about cancers hidden deep in the digestive system. In a study published in Cell Host & Microbe on August 20, researchers at Yonsei University in Seoul report that patterns linking oral and intestinal microbes can distinguish people with gastric or colorectal cancer from healthy individuals. The findings raise the possibility of a non-invasive screening approach based on the oral microbiome, potentially offering an easier alternative to stool-based testing for identifying people who should undergo more definitive procedures such as colonoscopy or upper endoscopy.
The study is built on a growing recognition that the microbiome is not confined to separate anatomical compartments. Microbes that colonize the mouth can be swallowed continuously, pass through the stomach, and reach the intestines. Under normal conditions, many of these organisms may fail to persist in the lower gastrointestinal tract. Cancer, however, can alter local physiology, immune activity, nutrient availability, tissue structure, and transit through the digestive system. These changes may create environments in which oral-associated bacteria survive or expand in the gut. The researchers therefore asked whether the degree of microbial overlap between a person’s mouth and feces might reflect the presence of gastrointestinal disease.
To investigate the question, the team recruited 507 volunteers and collected both oral and fecal specimens using a highly standardized sampling process. The cohort included 129 healthy participants, 215 people with metabolic disorders, 77 patients with gastric cancer, and 86 patients with colorectal cancer. The metabolic-disorder group included individuals with conditions such as metabolic syndrome, hypertension, hyperlipidemia, and type 2 diabetes, allowing the investigators to test whether microbial overlap was simply a general marker of poor health. Samples from the cancer groups were obtained before treatment, reducing the likelihood that surgery, chemotherapy, or other interventions had reshaped their microbial communities.
The researchers used microbial gene sequencing to identify sequence variants in each sample. These variants act as high-resolution signatures of bacterial populations, allowing the investigators to determine whether closely matching microbial features appeared in both the mouth and feces of the same individual. They combined this information into a measure called the mouth-to-feces, or MF, index. Rather than asking only which organisms were present, the index estimated how extensively oral-associated microbial signatures were detectable in the intestinal sample. A high MF index therefore indicated a greater degree of person-specific microbial sharing between the two body sites.
The resulting pattern was strikingly selective. The MF index was significantly elevated in people with gastric cancer and in those with colorectal cancer, while no comparable increase was observed among participants with metabolic disorders. This distinction suggests that the signal was not merely a reflection of obesity, diabetes, cardiovascular risk, or generalized metabolic dysfunction. The association also remained robust after the researchers accounted for alcohol consumption, regular exercise, and body-mass index. These adjustments are important because lifestyle and body composition can strongly influence both oral and gut microbial communities, potentially creating misleading correlations if they are not considered.
The cancer-associated signal was not limited to fecal material. When the investigators examined oral samples alone, they found that microbial features associated with cancer could still be detected. Only a small fraction of oral bacterial variants were also identified in the gut, but those shared features appeared to contain disproportionately useful information. Models based exclusively on oral microbial signatures were able to distinguish cancer patients from healthy individuals across several independent cohorts. This result suggests that a mouth rinse may preserve a measurable record of biological changes occurring farther down the digestive tract, even when the organisms themselves are present in the mouth at relatively low abundance.
The team also compared the oral-microbiome approach with fecal occult blood testing, a standard colorectal cancer screening method that searches stool for hidden blood. In the study, oral samples showed higher sensitivity than the occult-blood test. Sensitivity refers to the ability of a test to correctly identify people who have disease, although a highly sensitive result does not necessarily mean that the test is specific enough for diagnosis. The investigators emphasize that their microbial models are not yet ready to replace established screening or diagnostic procedures. Instead, an oral-rinse assay could eventually serve as a triage tool, helping prioritize individuals for colonoscopy, upper endoscopy, imaging, or tissue-based evaluation.
The biological explanation remains unresolved. One possibility is that oral microbes actively contribute to cancer development after reaching the gastrointestinal tract. Certain bacteria can influence inflammation, produce metabolites, interact with epithelial cells, or affect immune surveillance, mechanisms that could theoretically promote tumor formation or progression. Another possibility is that an existing tumor changes the gastrointestinal environment in ways that allow swallowed oral bacteria to persist. Altered mucus, acidity, oxygen levels, intestinal motility, barrier function, and immune responses could all affect microbial colonization. The relationship may also work in both directions, with cancer-related changes encouraging microbial persistence while microbial activity further modifies the diseased tissue.
Because the investigation was observational, it cannot establish whether the shared microbial signatures cause cancer, result from cancer, or simply accompany other biological changes. The study also needs to be tested prospectively in people who have not already received a cancer diagnosis, since screening populations differ from research cohorts that include known patients. The researchers plan such validation studies and intend to use higher-resolution metagenome sequencing, which can provide more detailed information about microbial species, genes, metabolic pathways, and potentially strain-level differences. If the findings remain reliable in large, diverse screening groups, the long-term goal would be a convenient oral test that combines microbial patterns with genetic, lifestyle, and clinical data to improve individualized risk prediction for gastrointestinal cancers.
Subject of Research: People
Article Title: Mouth-to-gut microbial transmission signatures enable robust, non-invasive diagnosis of gastrointestinal cancers
News Publication Date: 20-Aug-2026
Web References: https://doi.org/10.1016/j.chom.2026.07.007; https://www.cell.com/cell-host-microbe
References: Kim et al., “Mouth-to-gut microbial transmission signatures enable robust, non-invasive diagnosis of gastrointestinal cancers,” Cell Host & Microbe, DOI: 10.1016/j.chom.2026.07.007
Keywords: oral microbiome, gut microbiome, mouth-to-feces index, gastric cancer, colorectal cancer, gastrointestinal cancer, microbial gene sequencing, non-invasive cancer screening, oral-rinse test, cancer biomarkers
