Engineered organoid platforms for mechanistic studies of gut microbiota-cancer interactions

Associations between the gut microbiota and cancer are increasingly reproducible at the cohort level, but association alone cannot establish whether a microorganism, microbial product, or metabolite directly alters tumor-relevant host biology. Engineered organoid systems address this gap by placing defined microbial exposures against human or genetically specified epithelia under experimentally controlled conditions. The strength of the resulting inference depends on platform architecture. Three-dimensional organoids preserve epithelial organization but restrict luminal access; microinjection and apical-out configurations enable direct microbial exposure; organoid-derived monolayers permit directional sampling and barrier measurements; and microfluidic systems add flow, oxygen gradients, and longer-term host-microbe co-culture. Tumor organoids and defined immune or stromal co-cultures can be added when the biological question requires cancer-specific or multicellular readouts. This review evaluates these platforms according to the mechanistic questions they can answer and distinguishes evidence from live microorganisms, complex communities, cell-free microbial products, purified metabolites, and clinical associations. The strongest organoid-based evidence currently concerns microbial genotoxicity, epithelial transformation, barrier and repair responses, selected tumor-growth phenotypes, and intestinal treatment toxicity. Direct organoid evidence linking the gut microbiota to distant, non-gastrointestinal tumors remains limited despite substantial clinical and animal data. We further discuss how physicochemical control, omics-based benchmarking, and standardized reporting can improve reproducibility without adding unnecessary model complexity. The central requirement is explicit matching of experimental architecture to the causal claim being tested.