Microbial Signatures of Early Gastric Cancer

Abstract

Background: Gastric microbiome has been well characterized gastric cancer. Alterations associated with gastric cancer have been identified. However, little is known regarding the dysbiotic features in early gastric cancer. Methods: Gastric microbiome was assessed in early gastric cancer (n=30), advanced gastric cancer (n=30) and chronic gastritis (n=60). Random forest machine learning algorithm was used to identify signatures capable of predicting early gastric cancer. In addition, influence of Helicobacter pylori virulence and cancer-associated single nucleotide polymorphisms in PRKAA1, PTGER4, PSCA, MUC1 and UNC5CL on gastric microbiome was determined. Findings: There were significant differences in the microbial profile and composition between early and advanced gastric cancer, suggesting alterations associated with cancer progression. Compared with chronic gastritis, 32 bacteria genera were found to be enriched or depleted in early gastric cancer. Production of urease and synthesis of bacteria flagellum were decreased, while glycolysis of fructose and hydrolysis of glycosides were enhanced. The correlation network became simplified and fragmented. A microbial signature enclosing 24 bacteria was identified, which could distinguish early gastric cancer from chronic gastritis with an AUC value of 0.97. H. pylori had impacts on gastric microbiome. The minor allele of MUC1 SNP rs4072037 was associated with increased abundance of Ochrobactrum. Interpretation: Gastric microbiome showed distinct features in early gastric cancer. The identified microbial signature could be used as biomarkers for predicting early gastric cancer. Funding: National natural Science Foundation of China (31870777, 81602144). Declaration of Interest: The authors declare no conflicts of interest. Ethical Approval: This study was approved by the Research Ethnics Committee of Qingdao Municipal Hospital, China. Informed written consent was obtained from all participants.