Reduced penetrance BRCA1 and BRCA2 pathogenic variants in clinical germline genetic testing

Abstract

Prior studies have suggested the existence of reduced penetrance pathogenic variants (RPPVs) in BRCA1 and BRCA2 (BRCA) which pose challenges for patient counseling and care. Here, we sought to establish RPPVs as a new category of variants. Candidate BRCA RPPVs provided by two large clinical diagnostic laboratories were compiled to identify those with the highest likelihood of being a RPPV, based on concordant interpretations. Sixteen concordant candidate BRCA RPPVs across both laboratories were systematically assessed. RPPVs included missense, splice site, and frameshift variants. Our study establishes RPPVs as a new class of variants imparting a moderately increased risk of breast cancer, which impacts risk-informed cancer prevention strategies, and provides a framework to standardize interpretation and reporting of BRCA RPPVs. Further work to define clinically meaningful risk thresholds and categories for reporting BRCA RPPVs is needed to personalize cancer risks in conjunction with other risk factors.

Fig. 1. Subway chart outlining the study design.

HWA, cancer family and personal history weighing algorithms; RPPV, reduced penetrance pathogenic variant.

Fig. 2. Clinical history curves for seven RPPVs.

Aggregate clinical data of variant carriers (blue line) is plotted relative to the distribution of carriers of known pathogenic (red curves) and known benign (green curves). A History Weighing Algorithm (HWA; Ambry Genetics) assessment of aggregate family history for carriers of variants indicated (blue vertical line) plotted relative to the distribution of simulated control groups for benign (green curve) and pathogenic (red curve) variants. Dashed and solid vertical lines represent the conservative 95th and 99th percentile confidence bounds for pathogenic and benign curves for the control groups. Each count refers to one simulated dataset with the same number of probands as the test dataset. Each benign and pathogenic curve is comprised of 100,000 simulated datasets with the same number of probands as the test dataset and count refers to the number of those simulated datasets. LLR, Log Likelihood Ratio. B History Weighing Algorithm (HWA; Myriad Genetics) assessment of aggregate family history for carriers of variants indicated (dark blue vertical line) plotted relative to the distribution of simulated control groups for benign (green curve) and pathogenic (red curve) variants.

Fig. 3. Functional evidence for pathogenicity for missense RPPVs.

Available published and unpublished cumulative functional evidence for three RPPVs in BRCA1 (A) and BRCA2 (B) were compiled and coded per ACMG-AMP codes (PS3 functional evidence towards pathogenic; BS3 functional evidence towards benign) and strength (down-weighted from baseline ‘strong’ to either ‘moderate’ or ‘supporting’ where denoted)^,. A final call on functional strength is provided based on the cumulative functional evidence.