A Comparison of Tumor-Associated and Non-Tumor-Associated Gastric Microbiota in Gastric Cancer Patients

Background How gastric cancer (GC) incidence is associated with changes in the gastric microbiome has not been firmly established. The present study therefore aims to investigate the microbial communities present within the gastric mucosa of patients with superficial gastritis (SG) or GC. Methods Paired tumor and paracancerous samples of the gastric mucosa were collected from 18 patients being surgically treated for GC and from 32 patients with SG being treated via gastroscopy. The gastric microbiome in these samples was then profiled via 16S rRNA sequencing, with a linear discriminant analysis effect size (LEfSe) approach used to identify and compare different bacteria, and with PICRUSt used for predictive functional analyses. Results GC patients exhibited a distinct gastric microbiota profile from that observed in SG patients. These changes were evident in both tumor and paracancerous tissues from GC patients. Specifically, we found that 6 bacterial genera were specifically enriched in GC tissue samples relative to SG samples, while 18 genera were depleted in these same samples. Based on the differential abundance of these bacteria, we were able to calculate microbial dysbiosis index (MDI) values, which were significantly higher in GC patients than in SG patients. In addition, MDI values were negatively correlated with gastric Shannon index and were positively correlated with relativeHelicobacterspp. abundance. Importantly, these MDI values were readily able to discriminate between GC and SG patient samples. Functional analysis suggested that GC patients were more likely to harbor a nitrosating microbial community. Conclusions GC patients exhibited a gastric microbiome profile distinct from that observed in SG patients, with these differences being evident in both tumor and paracancerous tissues. Differences in the relative abundance ofHelicobacterspp. may be the primary driver of gastric dysbiosis in GC patients.