How the Microbiome Drives Cancer and Could Unlock Better Treatments

The microbiome — comprising bacteria, fungi, viruses, and archaea colonizing the gut, oral cavity, skin, and tumor tissue itself — has emerged as a critical but underappreciated axis in cancer biology. This 2025 review in iMeta, authored by a large international consortium, synthesizes the current state of evidence across the full spectrum of cancer types and microbiome compartments. The authors argue that the microbiome is not merely a bystander but an active participant in oncogenesis, immune modulation, and therapeutic response, making it a high-priority target for next-generation cancer medicine.

The review details multiple mechanistic pathways by which microbes promote carcinogenesis. Fusobacterium nucleatum, strongly implicated in colorectal cancer, activates Wnt/β-catenin signaling and suppresses NK cell and T cell activity, while producing the FadA adhesin that directly invades epithelial cells. Helicobacter pylori drives gastric cancer through CagA oncoprotein injection, inducing chronic inflammation and DNA double-strand breaks. Certain Escherichia coli strains harboring the pks genomic island produce colibactin, a genotoxin that creates characteristic mutational signatures (SBS88 and ID18) now detectable in colorectal cancer genomes. Oral bacteria including Porphyromonas gingivalis and Treponema denticola are enriched in pancreatic ductal adenocarcinoma tissue and correlate with worse prognosis. The intratumoral mycobiome — fungal communities within tumors — is also reviewed, with Malassezia species shown to activate complement pathways and promote pancreatic cancer progression in mouse models.

On the protective side, the review synthesizes robust evidence that microbiome diversity and specific taxa — notably Akkermansia muciniphila, Faecalibacterium prausnitzii, and Bifidobacterium species — are associated with improved responses to PD-1/PD-L1 checkpoint inhibitor therapy across melanoma, non-small cell lung cancer, renal cell carcinoma, and other tumor types. Mechanistically, these bacteria enhance dendritic cell maturation, increase CD8+ T cell infiltration into tumors, and promote production of short-chain fatty acids (SCFAs) that modulate regulatory T cell differentiation. Multiple clinical studies cited in the review show that patients with high Akkermansia abundance at baseline achieve objective response rates roughly 2–3 times higher with anti-PD-1 therapy compared to low-abundance patients.

The review dedicates substantial attention to therapeutic microbiome modulation. Fecal microbiota transplantation (FMT) from immunotherapy responders to non-responders has shown proof-of-concept in small clinical trials: two landmark studies (Davar et al. and Baruch et al., both 2021) demonstrated that FMT from melanoma responders converted some non-responders to responders, with response rates of approximately 20–30% in heavily pretreated patients. Engineered bacteria — such as E. coli Nissle programmed to produce IL-2 or anti-CD47 nanobodies selectively within the tumor microenvironment — represent a frontier approach with early preclinical promise. Phage therapy targeting oncogenic bacteria like F. nucleatum is also discussed as a precision strategy to deplete pro-tumorigenic species without broad dysbiosis.

The authors acknowledge significant methodological heterogeneity across the field, including inconsistent microbiome profiling methods (16S rRNA vs. shotgun metagenomics), lack of standardized biobanking, and the challenge of distinguishing causation from correlation in human studies. Confounders including diet, antibiotic use, geography, and host genetics complicate interpretation. The review calls for large, prospective, multi-omic cohort studies and standardized protocols. Despite these caveats, the authors conclude that microbiome-based diagnostics and therapeutics are approaching clinical readiness, and that integrating microbiome profiling into oncology trials should become standard practice.