Analysis of gut and oral shotgun metagenomic data from 317 individuals identifies 20 oral-gut shared species enriched in the gut of gastric cancer, predominantly lactic acid bacteria, with microbiome-based classifiers achieving high predictive accuracy (AUROC = 0.85 for stool, 0.87 for saliva) for GC diagnosis.
Key Findings
Results
Twenty oral-gut shared species were enriched in the gut microbiome of gastric cancer patients, predominantly lactic acid bacteria (LAB).
Analysis was performed on shotgun metagenomic data from 317 individuals across two independent cohorts, with validation in a Harbin cohort.
The 20 shared species were identified as enriched in the gut of GC patients.
Lactic acid bacteria represented the predominant group among these shared species.
While most gut microbial markers were abundant in saliva, none were significantly altered in GC at the oral level.
Results
Strain-level analysis confirmed oral-gut transmission of Streptococcus species in gastric cancer.
Strain-level analysis was performed on 87 matched saliva-stool metagenomes.
The analysis confirmed oral-to-gut transmission specifically for Streptococcus species.
This finding supports the role of the oral-gut microbiome axis in GC pathogenesis.
Results
GC-enriched lactic acid bacteria form robust co-abundance networks in both oral and gut microbiomes, suggesting synergistic interactions.
Co-abundance network analysis was applied to both oral and gut microbiome data.
LAB enriched in GC demonstrated robust co-abundance patterns in both microbiome compartments.
The finding suggests synergistic interactions among these organisms across body sites.
Results
Functional analysis revealed enriched lactate fermentation pathways in GC stool, consistent with lactic acid bacteria dominance.
Metagenomic functional pathway analysis was conducted on stool samples.
Lactate fermentation pathways were specifically enriched in GC stool.
This finding aligns with LAB dominance in the gut microbiome of GC patients and with previous findings on gastric microbiota.
Results
Microbiome-based classifiers achieved high predictive accuracy for gastric cancer diagnosis using both stool and saliva samples.
The stool-based classifier achieved an AUROC of 0.85.
The saliva-based classifier achieved an AUROC of 0.87.
Classifiers were developed using shotgun metagenomic data from 317 individuals across two independent cohorts.
Performance was validated in an independent Harbin cohort.
The authors highlight the 'translational potential' of these microbiome-based diagnostic tools.
Results
Most gut microbial markers in gastric cancer are also detectable in saliva but are not significantly altered in GC saliva samples.
Comparative analysis was conducted between gut and oral microbiome compositions in GC and non-GC individuals.
The majority of gut microbial markers associated with GC were found to be abundant in saliva.
Despite their presence in saliva, none of these markers showed significant differential abundance in GC saliva.
This dissociation suggests differential regulation or colonization dynamics between the oral cavity and gut in GC.
What This Means
This research suggests that gastric cancer (stomach cancer) is associated with specific changes in the microbial communities living in both the gut and the mouth. The researchers analyzed genetic material from microbes in stool and saliva samples from 317 people across multiple groups, including some with gastric cancer and some without. They found 20 species of bacteria that were elevated in the guts of gastric cancer patients, and most of these belonged to a group called lactic acid bacteria — the same types of bacteria found in fermented foods like yogurt. These bacteria were also detectable in saliva, but interestingly, their levels in saliva were not notably different between cancer and non-cancer individuals.
The study also found evidence that bacteria from the mouth, particularly Streptococcus species, are being transmitted to the gut in gastric cancer patients. These cancer-associated bacteria appeared to work together in networks in both the mouth and gut, and their presence corresponded with increased activity of lactate fermentation — a metabolic process that produces lactic acid. Using these microbial patterns as a basis for diagnostic tools, the researchers developed classifiers that could predict gastric cancer with high accuracy from stool samples (AUROC = 0.85) and even slightly higher accuracy from saliva samples (AUROC = 0.87).
This research suggests that the connection between the oral and gut microbiomes plays an important role in gastric cancer, and that analyzing microbes in saliva or stool could one day serve as a non-invasive way to help detect gastric cancer. This is particularly significant because gastric cancer is often diagnosed at late stages, and better early detection tools could improve patient outcomes. The findings also raise questions about whether oral bacteria traveling to the gut might contribute to cancer development, opening potential avenues for prevention research.
Qin Y, Zhang Y, Liu L, Xie Y, Ma X, Hao Y, et al.. (2026). Distinct signatures in the human gut and oral microbiomes of gastric cancer.. Cell reports. Medicine. https://doi.org/10.1016/j.xcrm.2026.102761