Abstract

NGS-based identification of novel hereditary breast/ovarian cancer genes in patients with clinical features of genetic predisposition

Full text
BackgroundGermline variants in known breast/ovarian cancer (BC/OC) predisposing genes explain less than half of the familial clustering of these diseases. This study aimed to identify novel genes for hereditary BC/OC.MethodsWe developed a custom NGS panel composed of 25 candidate BC/OC-associated genes. Twelve of these 25 genes (AEN, ATF5, BRIP1, CEBPA, FANCM, GREB1, GRWD1, P4HTM, POLA2, RAD50, RAD54B, STK36) were selected from our whole exome sequencing (WES) dataset involving BRCA1/2 mutation-negative OC patients with pronounced response to platinum-based therapy. The remaining 13 genes (ATRIP, BRWD1, BTG4, CACNA1H, CLN5, FGL1, GLIS3, GRAPL, PMS1, SLX4, TEN1, TKTL1, TSGA10IP) were selected from UK Biobank data based on WES comparison of 3290 early-onset or familial BC cases vs. healthy controls (n=44950, UK Biobank). Targeted NGS was performed for 385 patients with early-onset, familial, or bilateral BCs (n=385) and 395 consecutive OCs negative Slavic founder BRCA1/2 mutations.ResultsThirty-five potentially deleterious mutations in 14 genes, including 16 truncating variants and 19 missense mutations with CADD score >25, were identified. Thirteen of these alleles were recurrent, although having exceptionally low population frequency. The recurrent nonsense ATF5 p.Arg207* germline mutation was accompanied by LOH of the remaining wild-type allele and high-level homologous recombination deficiency (HRD) in the tumor tissue. Other genes with apparent association with BC/OC risk were CEBPA, FANCM, BRIP1, ATRIP, RAD50, and RAD54B.ConclusionsThe results of NGS analysis suggest a contribution of a multitude of rare pathogenic variants in genetic predisposition to BRCA1/2 mutation-negative BC and OC.Legal entity responsible for the studyThe authors.FundingThis work was supported by grant #22-15-00278 from the Russian Science Foundation and UK Biobank Resource project 41601. The work of YSA and AVN was funded by PolyKnomics BV.DisclosureAll authors have declared no conflicts of interest.