ETD

• genetics
• intraductal papillary mucinous neoplasm
• pancreatic cancer
• rare variants
• susceptibility loci

Pancreatic ductal adenocarcinoma (PDAC) is the fourth most common cause of cancer death in Europe and is projected to become the second by 2030. This cancer is typically asymptomatic in its early stages, which usually leads to late diagnosis at advanced stages of cancer. The most effective treatment is surgery, yet only around 20% of the patients are fit for surgery due to late diagnosis. Early detection is therefore critical. For this reason, Intraductal papillary mucinous neoplasms (IPMNs) represent an important opportunity for intervention since they are very common cystic lesions with a potential evolution into PDAC. The malignant progression of IPMN is a complex process driven by genetic and environmental factors. Clinically, progression is monitored through the development of worrisome features (WF; radiological markers of increased malignant risk) and/or high-risk stigmata (HRS; clinical/imaging findings warranting surgical evaluation). Therefore, early detection, accurate identification of individuals at high risk of malignant progression, and personalized adjustment of follow-up frequency represent critical steps toward effective disease management and ultimately preventing the transition of IPMN to PDAC. The aim of this study was to identify genetic variants associated with clinical progression of IPMN and build a risk score through a whole exome sequencing approach. Experimental design consisted of a total of 496 IPMN patients separated in two cohorts as follows; 393 IPMN patients under surveillance with outcome defined as development of new WFs and/or HRS (cohort 1), and 103 patients who underwent surgical resection with final pathology defined as high-grade dysplasia vs. low-grade dysplasia (cohort 2). Following quality controls of sequencing data, read mapping, alignment and variant calling procedures were carried out with Dragen Germline tool. Identified variants were then subjected to statistical analysis. Cox proportional hazard and logistic regression analyses were carried out to identify significant associations. Additionally, non-genetic factors collected at the time of diagnosis were subjected to statistical analyses. Genetic variant annotation was performed using OpenCRAVAT [1], which integrates multiple annotation resources including dbSNP [2], CADD [3], PolyPhen-2 [4], SIFT [5], GTEx [6], and dbGene. Cross validation of the SNPs among the two cohorts was carried out. Finally, a deleterious allele score that included all deleterious variants was computed to identify the cumulative effect of the deleterious variants. Statistical analyses identified 29,334 and 3065 significantly (p<0.05) associated variants in cohort 1 and cohort 2, respectively. The cross-validation identified seven variants significantly associated in both cohorts. One notable association was rs1058808 in ERBB2,a gene involved in cell survival and proliferation. Additionally, individuals within the highest quartile of the deleterious allele score were significantly associated with progression: Hazard Ratio =12.54 (95% CI: 6.32-24.90), p = 4.89 × 10-13. Consistently, Kaplan-Meier estimates showed that individuals in the first and second quartiles had approximately 70% probability of remaining progression-free after nearly three years. In contrast, patients in the fourth quartile had less than 25% probability of remaining progression-free after 18 months.