Bile molecular landscape provides pathological insight and classifies signatures predictive of carcinoma of the gall bladder.

Abstract

Carcinoma of the gall bladder (CAGB) has a poor prognosis. Molecular analysis of bile could classify indicators of CAGB. Bile samples (n = 87; training cohort) were screened for proteomics and metabolomics signatures of cancer detection. In bile, CAGB showed distinct proteomic (217 upregulated, 258 downregulated) and metabolomic phenotypes (111 upregulated, 505 downregulated, p < 0.05, fold change > 1.5, false discovery rate <0.01) linked to significantly increased inflammation (coagulation, arachidonic acid, bile acid) and alternate energy pathways (pentose-phosphate pathway, amino acids, lipid metabolism); and decreased glycolysis, cholesterol metabolism, PPAR, RAS, and RAP1 signaling, oxidative phosphorylation, and others compared to gallstone or healthy controls (p < 0.05). Bile proteins/metabolites signatures showed significant correlation (r ²  > 0.5, p < 0.05) with clinical parameters. Metabolite/protein signature-based probability of detection for CAGB (cancer) was >90% (p < 0.05), with area under the receiver operating characteristic curve >0.94. Validation of the top four metabolites-toluene, 5,6-DHET, creatine, and phenylacetaldehyde-in separate cohorts (n = 80; bile [T1] and paired plasma [T2]) showed accuracy (99%) and sensitivity/specificity (>98%) for CAGB detection. Tissue validation showed bile 5,6-DHET positively correlated with tissue PCNA (proliferation), and caspase-3 linked to cancer development (r ² >0.5, p < 0.05). In conclusion, the bile molecular landscape provides critical molecular understanding and outlines metabolomic indicator panels for early CAGB detection.