Multicohort Validation of Gut Microbiome Signatures for Cholangiocarcinoma Diagnosis and Functional Characterization of Bifidobacterium Pseudocatenulatum

ABSTRACT Growing evidence suggests a role for the gut microbiome in progression of cholangiocarcinoma (CCA), however, its diagnostic and therapeutic potential remains incompletely characterized. Here, metagenomic sequencing was performed on fecal samples ( n = 785) from individuals across East, Central, and Northwestern China. Gut microbial dysbiosis in CCA was characterized by depletion of short‐chain fatty acids‐producing species and enrichment of potential pathobionts ( Klebsiella aerogenes , Clostridium symbiosum ). Diagnostic models built using species‐level markers demonstrated superior performance, compared to pathway‐based models, achieving area under the curve (AUC) values of 98.63% and 99.42% in the discovery cohort, with robust cross‐regional validation (AUC = 80.89% and 80.43%). The model effectively distinguished CCA from hepatocellular carcinoma (AUC = 97.86%) and liver fibrosis (AUC = 98.73%) and nonalcoholic fatty liver disease (mean AUC = 96.86%). Analysis of public datasets encompassing 6847 samples across 31 studies and 11 disease states revealed moderate disease specificity influenced by biomarker overlap across conditions. Mechanistically, depleted Bifidobacterium pseudocatenulatum suppressed CCA progression, associated with inhibition of the PI3K‐AKT‐mTOR pathway. Collectively, this study supports the potential of fecal metagenomic signatures as a complementary noninvasive aid for CCA detection, and provides functional evidence for a candidate protective microbe.