Prognostic implications of the interaction between intratumoral microbiome and immune response in gastric cancer.

Gastric cancer (GC) prognosis is significantly influenced by intratumoral microbiomes, which modulate host-immune interactions. This study analyzed data from the The Cancer Genome Atlas (TCGA) and Gene Expression Omnibus (GEO) databases to identify immune genes associated with GC prognosis and conducted prognostic immune subtypes. GC patients were classified into two distinct prognostic immune phenotypes C1 and C2 based on the non-negative matrix factorization consensus clusters. Phenotype C2 exhibited a better prognosis and distinct immune characteristics, including enhanced presence of Th2 and Th17 cells and improved response to chemotherapy. In contrast, phenotype C1 showed higher expression levels of PDCD1LG2 and TLR9, which were critical immune factors involved in immune regulation. Both phenotypes were linked to immune genes influencing intratumoral microbiomes and GC immunotherapy responses. A prediction risk model was constructed using the LASSO regression analysis and showed great prognostic value for GC patients. The key genes were correlated with immune cells and suppressed the function of the host immune system. The intratumoral microbiomes were strongly associated with the hosts' immune infiltration and significantly interacted with host immune genes to influence GC outcomes. Candidatus Nitrosotenuis plays a significant role in predicting the prognosis of GC patients. This research underscores the pivotal role of intratumoral microbiomes in GC prognosis and supports the development of future personalized therapeutic approaches.IMPORTANCEIncreasing evidence confirms the presence of intratumoral microbiomes. However, the role of the intratumoral microbiomes in the progression of gastric cancer and their relationship with the immune microenvironment remain unclear. Our study classified gastric cancer patients into two immune prognostic subtypes, C1 and C2, using non-negative matrix factorization consensus clusters. The C2 subtype exhibited a better prognosis and more pronounced immune characteristics. Microbiome analyses revealed associations between both subtypes and immune genes that affect intratumoral microbiomes and their responses to immunotherapy. The intratumoral microbiomes were closely linked with host immune infiltration and significantly interacted with immune genes, which influence the prognosis of gastric cancer. Notably, Candidatus Nitrosotenuis showed a significant prognostic value in gastric cancer patients. Our findings highlight the critical role of the intratumoral microbiomes in affecting gastric cancer prognosis and its interaction with the immune microenvironment, supporting future personalized therapeutic approaches.