Abstract
Identification of germline pathogenic variants in cancer patients is critical for treatment planning, genetic counseling, and health policymaking. However, previous estimates of the prevalence of germline etiology of pancreatic ductal adenocarcinoma (PDAC) were biased because they were based only on sequencing data of protein-coding regions of known PDAC candidate genes. To determine the percentage of patients with PDAC carrying germline pathogenic variants, we enrolled the inpatients from the digestive health clinics, hematology and oncology clinics, and surgical clinics of a single tertiary medical center in Taiwan for whole genome sequencing (WGS) analysis of genomic DNA. The virtual gene panel of 750 genes comprised PDAC candidate genes and those listed in the COSMIC Cancer Gene Census. The genetic variant types under investigation included single nucleotide substitutions, small indels, structural variants, and mobile element insertions (MEIs). In 8 of 24 (33.3%) patients with PDAC, we identified pathogenic/likely pathogenic variants, including single nucleotide substitutions and small indels in ATM, BRCA1, BRCA2, POLQ, SPINK1 and CASP8, as well as structural variants in CDC25C and USP44. We identified additional patients carrying variants that could potentially affect splicing. This cohort study demonstrates that an extensive analysis of the abundant information yielded by the WGS approach can uncover many pathogenic variants that could be missed by traditional panel-based or whole exome sequencing-based approaches. The percentage of patients with PDAC carrying germline variants might be much higher than previously expected.