Search for a command to run...
ABSTRACT Background and Aims Highly aggressive hepatobiliary tumours include gallbladder cancer (GBC), hepatocellular carcinoma (HCC), intrahepatic and extrahepatic cholangiocarcinoma (iCCA, eCCA) and ampulla of Vater cancer (AoV). We aimed to identify plasma biomarkers for the early diagnosis of hepatobiliary cancer by leveraging the metabolomic signatures of established clinical risk factors. Method Based on 273,190 participants from the UK Biobank, we (1) identified metabolites associated with gallstone-related conditions (e.g. cholecystitis), primary sclerosing cholangitis (PSC) and metabolic liver diseases (e.g. cirrhosis), and (2) evaluated the relationship between the identified metabolites and the risk of GBC, HCC, iCCA, eCCA and AoV. Findings were validated in an independent group of 227,809 participants from the UK Biobank. We also derived metabolomic scores summarizing the three risk-factor signatures and evaluated their ability to stratify cancer risk. Results We identified 27 metabolites associated with gallstone-related conditions, 11 with PSC, and 34 with metabolic liver diseases, some of which showed associations with inconsistent directions across risk factors, suggesting distinct pathogenic processes. Several metabolites were associated with cancer risk in both the discovery and validation datasets, independently of established risk factors, predominantly for HCC (16 signals) and for iCCA (4), with one for GBC and none for eCCA and AoV. Metabolomic scores clearly distinguished individuals at high risk for HCC and iCCA. Conclusion The preselection of plasma metabolites associated with established risk factors facilitated the subsequent identification and validation of biomarkers for early cancer detection. The identified metabolites suggest specific pathogenic pathways for each type of hepatobiliary cancer. Wider replication is urgently needed to advance toward clinical implementation. What you need to know BACKGROUND AND CONTEXT Clinical risk factors for hepatobiliary cancers often progress silently, making early identification of high-risk individuals difficult and highlighting the need for biological markers detectable before clinical diagnosis. NEW FINDINGS Risk-factor–based serum metabolomic profiling identified circulating metabolites that predict specific hepatobiliary cancers years before diagnosis, with strongest and most consistent signals for hepatocellular and intrahepatic cholangiocarcinoma. LIMITATIONS Clinical risk factors were assumed to be frequently underdiagnosed in UK Biobank, and event numbers were relatively small for some cancers, which may have reduced power and attenuated associations for less common endpoints. CLINICAL RESEARCH RELEVANCE This study shows that serum metabolic profiles can identify individuals at increased risk for hepatobiliary cancers long before symptoms appear, particularly for hepatocellular and intrahepatic cholangiocarcinoma. These findings support the development of precision risk-stratification strategies that may ultimately enable earlier surveillance. BASIC RESEARCH RELEVANCE By first identifying metabolites linked to specific liver and biliary clinical conditions, the study clarifies which metabolites are indirectly associated with hepatobiliary cancers through known disease pathways. Testing these metabolites again while adjusting for diagnoses of those conditions then reveals which ones also show direct, pathway-independent associations with individual hepatobiliary cancers, providing clearer insight into cancer-specific metabolic mechanisms.