Evaluation of polygenic risk score for pancreatic ductal adenocarcinoma risk prediction

Abstract Background Pancreatic ductal adenocarcinoma (PDAC) is one of the deadliest malignancies, with 5-year survival rate of less than 10%. There is an urgent need to identify high-risk subgroups and tailor prevention strategies. Polygenic risk score (PRS) has shown promise in predicting susceptibility to several cancers mainly if integrated into risk models and non-O blood group alleles have been associated with increased PDAC risk. We conducted a case-control study to assess the predictive value of PRS for PDAC risk and clinical risk factors. Methods For this study, we enrolled 570 patients with histologically confirmed PDAC and 570 controls without cancer. We gave a questionnaire for medical history data and blood samples were collected. DNA has been extracted and genotyped using the Axiom Precision Medicine Diversity Array. PRS has been constructed using 30 SNPs, including those required to deduce ABO blood groups. Logistic regression models assessed the association between PRS percentiles and PDAC risk. A multifactorial risk score integrating PRS with smoking and diabetes status has been constructed. Analyses have been adjusted for demographic and clinical variables. Results To date, we have genotyped and analyzed 285 cases and 285 controls. We found that both scores were associated with an increased risk of PDAC. An OR of 2.23 (95% CI 1.15-4.34) for the highest compared to lowest quintile of the weighted PRS and an OR of 12.16 (95% CI 4.14-35.74) between highest and lowest quintile of the weighted multifactorial score. Conclusions These preliminary results highlight how PRS may help predict PDAC risk, especially when integrated into model with other risk factors like diabetes and smoking. This tool could refine risk stratification, enabling personalized prevention and early detection strategies, with potential public health benefits through risk-based screening and better resource allocation. This work is part of DARE - Digital Lifelong Prevention Project, B53C22006320001. Key messages • The development of a personalised risk prediction model could decrease the burden of disease by reducing the impact of the condition on the healthcare system. • disease by reducing the impact of the condition on the healthcare system. The use of an innovative tool such as the PRS could improve the identification of high-risk individuals.