Impact of population-based testing for pathogenic variants in breast cancer genes on screening recommendations in the WISDOM study.

Abstract

10512 Background: The relative contributions of pathogenic variants (PV) in breast cancer (BC) genes, polygenic risk scores (PRS), and clinical risk factors to inform risk-stratified screening have not been well-studied on the population level. The WISDOM Study, a study of risk-based screening, used panel-based testing for PVs in 9 BC genes and the Breast Cancer Surveillance Consortium risk model modified by PRS (BCSC-PRS) to generate screening recommendations for > 30,000 participants undergoing risk-based screening. Women with PVs in high-penetrance genes were assigned q6 month screening (alternating MRI and mammogram). Those with PVs in moderate-penetrance genes were assigned q6 screening or yearly mammogram depending on family history. To determine the impact of PV detection on screening assignments, we examined the concordance between screening strategies informed by PV status vs. BCSC-PRS only. Methods: We analyzed participants with high-penetrance ( BRCA1, BRCA2, TP53, PALB2, STK11, CDH1 ), moderate-penetrance ( CHEK2, ATM ), or CHEK2 low-penetrance (I157T, S428T) PVs. We performed cross-tabulations of hypothetical (based on BCSC-PRS only) vs. actual (which considers PV) screening recommendation. To generate hypothetical screening recommendations, we calculated individual BCSC-PRS 5-year risk and applied the trial’s risk-based screening algorithm as if there were no PV. We considered initial (Year 1) screening recommendations to reflect that PV testing occurred at time of enrollment. We stratified analyses by PV category and age. Results: Our analysis included 714 women with a PV, of which 32% were high-penetrance, 39% were moderate-penetrance, and 29% were CHEK2 low-penetrance. Median age was 53 (IQR 45-62) years and 86% self-reported as non-Hispanic White. We found little overlap between screening recommendations based on PV status vs. BCSC-PRS. In women with a high-penetrance PV, 1.3% (3/234) would have received a q6 screening recommendation based on BCSC-PRS. In women with a moderate-penetrance PV, 33% (92/278) were assigned to q6 screening, of whom 6.5% (6/92) would have received the same assignment under BCSC-PRS. Additionally, 65% (182/278) were assigned to yearly mammogram, of whom 2.2% (4/182) would have received the same assignment under BCSC-PRS. In women with PVs aged 40-49, 67% (187/280) would have been recommended no screening until age 50 based on BCSC-PRS. In women with PVs aged 50-74, 90% (390/432) would have been recommended biennial screening. Conclusions: Most women with PVs in BC genes would not have otherwise been recommended for enhanced screening based on clinical and PRS risk. Our results highlight the importance of testing for PVs as a cornerstone of risk-based screening on the population level. Future studies will examine opportunities for risk reduction as we integrate genetic and non-genetic risk factors. Clinical trial information: NCT02620852 .