Mucosal microbiome associates with progression to gastric cancer.

Dysbiosis is associated with gastric cancer (GC) development. However, no longitudinal study was carried out to identify key bacteria that could predict for GC progression. Here, we aimed to investigate changes in bacterial metagenome prior to GC and develop a microbiome-based predictive model to accurately classify patients at risk of GC. Bacterial 16S rDNA was sequenced from 89 gastric antral biopsies obtained from 43 participants. This study was nested in a prospective, longitudinal study, whereby study participants underwent screening gastroscopy, with further 1-2 yearly surveillance gastroscopies for at least 5 years. Putative bacterial taxonomic and functional features associated with GC carcinogenesis were identified by comparing between controls, patients with gastric intestinal metaplasia (IM) and patients with early gastric neoplasia (EGN). Patients with EGN had enrichment of (in particular genus) and depletion of (in particular family) in their gastric mucosa. Sequencing identified more patients with compared to histopathological assessment, while was also significantly enriched in EGN Furthermore, a total of 261 functional features, attributing to 97 KEGG pathways were differentially abundant at baseline between patients who subsequent developed EGN (n = 13/39) and those who did not. At the same time, a constellation of six microbial taxonomic features present at baseline, provided the highest classifying power for subsequent EGN (AUC = 0.82). Our study highlights early microbial changes associated with GC carcinogenesis, suggesting a potential role for prospective microbiome surveillance for GC.