Abstract

Clinical genetic testing readily detects germline genetic variants. Yet, the evidence available for variant classification as benign or pathogenic is often limited by the rarity of individual variants, leading to many "variant of uncertain significance" (VUS) classifications. VUS cannot guide clinical decisions, complicating counseling and management. Laboratory assays can potentially aid reclassification, but require benchmarking against variants with definitive interpretations to have sufficient predictive power for clinical use. Of all clinically identified germline variants in hereditary breast cancer gene PALB2 (Partner and Localizer of BRCA2), ~50% are VUS and ~90% of VUS are missense. Truncating PALB2 variants have homologous recombination (HR) defects and instead rely on error-prone non-homologous end-joining (NHEJ) for DNA damage repair (DDR). Recent reports show some missense PALB2 variants may also be damaging, but thus far functional studies have lacked benchmarking controls. Using the Traffic Light Reporter (TLR) to quantify cellular HR and NHEJ using fluorescent markers, we assessed variant-level DDR capacity in hereditary breast cancer genes. We first determined the TLRs dynamic range using BRCA2 missense variants of known significance as benchmarks for normal/abnormal HR function. We then tested 37 PALB2 variants, generating functional data for germline PALB2 variants at a moderate level of evidence for a pathogenic interpretation (PS3_moderate) for 8 variants, or a supporting level of evidence in favor of a benign interpretation (BS3_supporting) for 13 variants, based on the ability of the assay to correctly classify PALB2 validation controls. This new data can be applied in subsequent variant interpretations for direct clinical benefit.