Integrative analysis of immune and microbial subtypes predicts immunotherapy response in stomach adenocarcinoma

ABSTRACT The tumor immune microenvironment and intratumoral microbiota play critical roles in cancer progression and immunotherapy response, yet their integrated functions in stomach adenocarcinoma (STAD) are not well understood. This study conducted a multi-omics analysis of transcriptomic and microbiome data from 348 patients with STAD. Using the ImmuCellAI algorithm, immune cell infiltration (ICI) was estimated, and non-negative matrix factorization classified samples into three immune subtypes (INC-1, INC-2, and INC-3). Differential expression analysis identified immune-related signature genes enriched in immune signaling pathways. Tumor mutational burden, microsatellite instability, immune checkpoint gene expression, and drug sensitivity were compared across subtypes. Microbiome clustering identified three subtypes (MC-1, MC-2, and MC-3), with associations to immune infiltration and microbial composition. The immune subtypes showed distinct patterns of ICI, clinical stage, and gene expression, with differentially expressed genes enriched in immune and tumor-related pathways. Microbiome subtypes exhibited unique diversity metrics and associations with the immune microenvironment. Integration of immune and microbial data improved immune checkpoint blockade (ICB) prediction, with genera like Staphylococcus and Ralstonia correlating with immune genes such as CD22, VIPR2, and FLT3. These findings provide insights into ICB response and support more precise immunotherapy strategies for STAD. IMPORTANCE Deciphering the interactions between the tumor immune microenvironment and the intratumoral microbiota is crucial for advancing precision immunotherapy in stomach adenocarcinoma (STAD). In this study, we present an integrative multi-omics framework that stratifies patients into distinct immune and microbial subtypes, uncovering their associations with immunogenomic profiles, immune cell infiltration patterns, and clinical features. Notably, we identify specific microbial genera correlated with immune-related gene expression and immune checkpoint blockade responsiveness. These findings provide novel insights into the immune–microbiome axis in STAD and underscore the potential of integrative multi-omics approaches to enhance patient stratification and guide more effective immunotherapeutic strategies.