Microbial signatures with diagnostic potential in early gastric cancer: insights beyond H. pylori.

BACKGROUND

The involvement of the intratumoral microbiome in gastric cancer (GC) pathogenesis remains insufficiently characterized, particularly in early gastric cancer (EGC). Here, we explored the relationship between the microbial community and EGC, with a specific focus on Helicobacter pylori (Hp)-negative microbiota. METHOD: Gastric mucosal microbiota from patients with EGC (n = 76) and healthy controls (HC, n = 39) were analyzed using 16 S rRNA V3-V4 amplicon sequencing. Linear discriminant analysis effect size (LEfSe) and random forest modeling were employed to identify microbial taxa that could serve as potential biomarkers distinguishing EGC from HC. In addition, the functional potential of the microbial communities was predicted using PICRUSt2, with annotation based on KEGG pathways.

RESULTS

A total of 115 participants contributed 306 gastric mucosal samples. Significant microbial differences were observed across groups stratified by EGC and Hp status. In EGC-HP-Negative (EHN) group, Firmicutes and Actinobacteria decreased and Proteobacteria increased, accompanied by a reduced Firmicutes/Proteobacteria ratio. LEfSe analysis indicated that Prevotella and Enterococcus were markedly enriched, each presenting LDA scores above 4. In the random forest model, Prevotella was identified as the most influential genus for separating EHN from HC-HP-Negative (HHN), with the classifier achieving an AUC of 0.90. Microbial network complexity was highest in general EGC and lowest in EHN, whereas Hp infection enhanced network connectivity. Functional prediction indicated distinct microbial functional shifts among groups, characterized by enrichment of bile secretion and metabolic pathways in EHN and increased RNA-related pathways in EHP compared with HC.

CONCLUSION

Gastric microbiota in EGC shows distinct alterations both with and without Hp. Prevotella may serve as a promising non-Hp microbial biomarker for early detection, while Hp further shapes microbial composition, network interactions, and pro-inflammatory pathways. These findings highlight the potential of targeting gastric microbiota for early diagnosis and therapeutic strategies in EGC.