Abstract There is growing evidence that microorganisms represent an important component of the tumor microenvironment. However, conflicting reports regarding the cancer microbiome have left the extent of microbial presence across cancer types unclear, highlighting the need for more robust methods for identifying tumor-associated microorganisms. Leveraging the completed human reference genome, we built a host-subtraction and microbial classification pipeline to accurately identify tumor-associated microorganisms in whole-genome sequencing data, which we validated and benchmarked on in silico and in vitro mixtures of human and microbial DNA. We used this pipeline to perform the largest pan-cancer microbiome analysis to date, spanning 16,369 high-depth tumor whole genomes from the UK 100,000 Genomes Project. After decontamination, microbial signatures were indistinguishable from background in most cancer types. However, in oral, esophageal, gastric, and colorectal cancers, we detected multi-kingdom polymicrobial communities, including bacteria, fungi, viruses, and archaea. In some oral and colorectal cancers, we also detected the parasite Trichomonas vaginalis. These microbial communities varied by tumor site and subtype, with increased microbial colonization of microsatellite-instable and POLE/POLD1-mutated tumors. This pattern was driven by a correlation between microbial load and tumor mutation burden across cancer types that was independent of genomic subtype. Additionally, we observed an over tenfold depletion of Akkermansia muciniphila in early-onset colorectal cancer. Together, our analysis helps resolve the microbial landscape of cancer and provides a new tumor microbiome atlas for future studies. Citation Format: Anders B. Dohlman, Robin Mjelle, Henry M. Wood, Alaina Shumate, Iris T.-h. Lee, Gianmarco Piccinno, Phil Quirke, Curtis Huttenhower, Nicola Segata, Matthew L. Meyerson. Biodiversity and biogeography of the multi-kingdom cancer microbiome abstract. In: Proceedings of the American Association for Cancer Research Annual Meeting 2026; Part 1 (Regular Abstracts); 2026 Apr 17-22; San Diego, CA. Philadelphia (PA): AACR; Cancer Res 2026;86(7 Suppl):Abstract nr 4901.
Dohlman et al. (Fri,) studied this question.