Bile metabolic fingerprints distinguish biliary tract cancer from benign biliary diseases

Abstract

Background and Aims: Biliary tract cancers (BTCs) are aggressive gastrointestinal malignancies characterized by a dismal 5-year overall survival rate less than 20%. Current diagnostic modalities suffer from limitations regarding sensitivity and specificity. This study aimed to develop a bile metabolite-based platform for precise discrimination between malignant and benign biliary diseases. Approach and Results: Samples were collected from 336 patients with BTC or benign biliary diseases across three independent cohorts. Untargeted metabolic fingerprinting was performed on 300 bile samples using novel nanoparticle-enhanced laser desorption/ionization mass spectrometry (NPELDI MS). Subsequently, a diagnostic assay was developed based on the exploratory cohort using a selected bile metabolic biomarker panel, with performance evaluated in the validation cohort. Further external validation of disease-specific metabolites from bile samples was conducted in a prospective cohort (n=36) using quantitative analysis. As a result, we established a novel bile-based assay, BileMet, for the rapid and precise detection of malignancies in the biliary tract system with an area under the curve of 0.891. We identified 6 metabolite biomarker candidates and discovered the critical role of the chenodeoxycholic acid glycine conjugate as a protective metabolite associated with BTC. Conclusions: Our findings confirmed the improved diagnostic capabilities of BileMet assay in a clinical setting. If applied, the BileMet assay enables intraoperative testing and fast medical decision-making for cases with suspected malignancy where brush cytology detection fails to support malignancy, ultimately reducing the economic burden by over 90%.