Abstract
Microbial communities residing in tumors constitute a critical component of the tumor microenvironment, particularly in gastrointestinal cancers arising from mucosal sites. However, the relationship between microbiota and oncogenesis, as well as its clinical applications, remains underexplored. Here, we performed a comprehensive analysis of the tumor microbiome across six gastrointestinal cancer types and identified a core microbiota composed of 15 bacterial genera associated with patient prognosis. Using bacterial abundance data, we developed a microbiota-based prognostic model to calculate a "risk score." Patients with high-risk scores exhibited poorer prognosis and increased metastatic potential, driven by the activation of tumor metastasis-related signaling pathways, including epithelial-mesenchymal transition, angiogenesis, KRAS, and TGFβ signaling. Furthermore, the model predicted sensitivity to anti-cancer drugs, identifying XL999 and tandutinib as potential targeted therapies for high-risk patients. The core microbiota is also linked to host immunity; patients with high-risk scores were less likely to benefit from immunotherapy. Taken together, these findings highlight the potential of a microbiota-based prognostic model to enhance cancer diagnostics, inform therapeutic decision-making, and advance personalized medicine for gastrointestinal cancers.IMPORTANCEIntratumoral microbiota influence cancer progression, yet their prognostic potential remains underutilized. This study identifies a core microbiota of 15 bacterial genera associated with survival in gastrointestinal cancers and develops a microbiota-based prognostic model. Unlike traditional gene-based models, this approach stratifies patients by microbial signatures, linking high-risk scores to enhanced metastasis via epithelial-mesenchymal transition, angiogenesis, and KRAS signaling. Additionally, we identify XL999 and tandutinib as potential therapies for high-risk patients and reveal that microbiota composition correlates with immunotherapy response. By integrating microbiome profiling into cancer prognosis and treatment selection, this study offers a novel strategy for precision oncology, advancing microbial biomarkers for risk assessment, drug selection, and personalized immunotherapy.