Multi-Omics Analysis Reveals Intratumor Microbes as Immunomodulators in Colorectal Cancer.

Recent studies indicated that intratumor microbes are an essential part of the tumor microenvironment. Here, we performed an integrated analysis of genetic, epigenetic, and intratumor microbial factors to unravel the potential remodeling mechanisms of immune-cell infiltration (ICI) and tumorigenesis of colorectal cancer (CRC). We identified the components and structure of the intratumor microbiome as primary contributors to the difference in survival between ICI subtypes. Multiple tumor-infiltrating immune cells (TIICs) and immune-related genes were associated with intratumor microbial abundance. Additionally, we found that was enriched in CRC patients who were nonsensitive to immune checkpoint blockade (ICB) therapy. We further provided clues that the intratumor microbes might influence the response to ICB therapy by mediating TIICs, especially MAIT (mucosa-associated invariant T) cells. Finally, three ICB-related TIICs and 22 of their associated microbes showed the potential to predict the response to ICB therapy (area under the receiver operating characteristic curve [AUC] = 89%). Our findings highlight the crucial role of intratumor microbes in affecting immune-cell infiltration patterns, prognosis, and therapy response of CRC and provide insights for improving current immunotherapeutic treatment strategies and prognosis for CRC patients. Using the multi-omics data from The Cancer Genome Atlas (TCGA) colorectal cancer (CRC) cohort, we estimated the tumor microenvironment (TME) infiltration patterns of patients and unraveled the interplay of gene expression, epigenetic modification, and the intratumor microbiome. This study suggests the impact of intratumor microbes on maintaining the tumor immune microenvironment in the pathogenesis of CRC and modulating the response to immune checkpoint blockade (ICB) therapy. We identified a set of combined features, including 3 ICB-related tumor-infiltrating immune cells (TIICs) and 22 of their associated microbes, that are predictive of ICB responses.