Cross-population genome-wide meta-analysis identifies pancreatic cancer susceptibility loci and etiologic pathways

Abstract

Abstract Background Pancreatic cancer is a fatal malignancy influenced by genetic, metabolic, and lifestyle factors. We aimed to identify susceptibility loci and clarify associated genes, biological pathways, and potential etiologic factors. Methods We performed a cross-population genome-wide association meta-analysis of 11 GWAS summary-statistic datasets comprising 20,120 pancreatic cancer cases and 1,446,192 controls genetically similar to European or East Asian reference populations. We integrated six gene-prioritization approaches, bulk-tissue and single-cell transcriptome-wide association studies, pathway enrichment, Mendelian randomization of modifiable factors and circulating proteins, and polygenic score and phenome-wide association analyses in UK Biobank. Results We identified 18 genome-wide significant loci, including four previously unreported loci, and prioritized one candidate gene at each locus. Transcriptome-wide association studies identified 46 genes associated with pancreatic cancer risk; ABO , PLEKHN1 , FBRSL1 , and KRT8 overlapped with genes identified by the gene-prioritization framework. Pathway enrichment implicated signal transduction, cell adhesion and migration, kinase-related pathways, and lipid metabolism. Mendelian randomization supported associations of genetically predicted fasting insulin, coffee consumption, obesity-related traits, and smoking-related traits with higher pancreatic cancer risk and moderate-to-vigorous physical activity with lower risk. Ten circulating proteins were associated with pancreatic cancer risk; eight—ABO, GRP, LGR4, CHST9, KRT18, IDUA, PCSK1, and FUT3—had strong colocalization support. Each one-standard-deviation increase in the polygenic score was associated with a 31% higher risk of incident pancreatic cancer (hazard ratio, 1.31; 95% confidence interval, 1.25–1.37), and phenome-wide association analyses identified 24 significant phenotypes. Conclusions These findings refine the genetic architecture of pancreatic cancer and provide insights into candidate genes, biological pathways, and potential etiologic risk factors.