More Reliable Breast Cancer Risk Assessment for Every Woman

Abstract

Two studies in The New England Journal of Medicine assessed the associations between germline variants in putative cancer susceptibility genes and the risk of breast cancer using large, unbiased cohorts. They consistently identified a small set of genes most informative for risk prediction, helping select high-risk women in the general population and developing effective cancer prevention strategies.

Breast cancer is the most common cancer for women in the United States, with an estimated 276,480 newly diagnosed cases and 42,179 cancer-related deaths in 2020 (Siegel et al., 2020). The best approach to manage this disease is to implement effective screening and risk reduction strategies for high-risk individuals, thereby preventing its occurrence or detecting it early when curative treatments are available. A major risk factor for breast cancer is germline mutations in specific cancer susceptibility genes that predispose the carrier to the development of cancer. Therefore, it is essential to identify reliable breast cancer susceptibility genes and accurately estimate the risks associated with their variants. However, the current risk estimations for breast cancer susceptibility genes are either based on high-risk populations, such as women with a family history of breast and ovarian cancer, or unselected population studies, but with limited cohort sizes. Due to these limitations, many putative susceptibility genes are controversial, the variant prevalence and risk estimate for established susceptibility genes are imprecise, and the risk associations for specific breast cancer subtypes are lacking, representing a critical knowledge gap in implementing effective cancer prevention and screening strategies for women in the general population.

Two population-based studies in The New England Journal of Medicine addressed this gap by characterizing putative breast cancer susceptibility genes in large, unbiased cohorts; and they directly sequenced the germline mutations of women with breast cancer (case patients) and unaffected women (controls) using multi-gene panels. One study is from the Breast Cancer Association Consortium (BCAC) (Breast Cancer Association et al., 2021), including 34 susceptibility gene candidates for 60,466 cases and 53,461 controls from 25 countries; and the other study is from the Cancer Risk Estimates Related to Susceptibility (CARRIERS) consortium (Hu et al., 2021), covering 28 genes for 32,247 cases and 32,544 controls from the United States (Fig. 1). Through a comparison of variant frequencies between case and control groups, these studies systematically assessed the associations of the susceptibility genes with breast cancer risk and estimated the risk models for both overall breast cancer and its subtypes. Impressively, the results from these two independent studies are highly consistent: the BCAC analysis identified 9 genes whose variants had a significant association with the risk of breast cancer, while the CARRIERS analysis identified 10 such genes, with 8 genes common to both studies. Furthermore, the ranks of odds ratios for the 8 genes for overall breast cancer risk in the two studies are strongly correlated (R = 0.98, P = 1×10⁻⁵, Fig. 1): BRCA1 variants exhibited the highest risk, followed by BRCA2 and PALB2, while CHEK2, ATM, BARD1, RAD51C, and RAD51D variants were associated with moderate risk. In addition, these 8 genes showed a consistent preference for risk-related breast-cancer subtypes (i.e., ER-positive or negative). In terms of study-specific findings, TP53 risk association was identified in the BCAC study but not in the CARRIERS study, presumably due to its low germline mutation prevalence in the general population; CDH1 association with ER-positive breast cancer was reported in the CARRIERS study.

Figure 1. A schematic summary of the BCAC and CARRIER case-control studies.

Susceptibility genes detected in both studies are highlighted by two circles. The color of gene symbol indicates the breast cancer subtype association (or preference), with dark red and blue highlighting those findings reported in both studies.

Because of the large cohort size and unbiased selection of breast cancer patients, the BCAC and CARRIER studies were able to (i) identify a shortlist of susceptibility genes with high confidence, (ii) estimate the prevalence and associated risk for variants in these genes more precisely, and (iii) estimate breast cancer subtype-specific risk. These findings have significant clinical implications. First, the two studies clearly define the genes that are most informative for the prediction of breast cancer risk in the general population and, therefore, should be the focus of genetic testing and counseling. As next-generation sequencing becomes more affordable, genetic testing based on multi-gene panels is becoming popular. However, many genes currently included in commercial panels may not be informative susceptibility genes (Easton et al., 2015). These two largest population-based studies collectively covered 38 putative susceptibility genes, and their results showed a high convergence for a small set of genes and detected no credible association for the remaining 27 genes (13 genes assessed in both studies and 14 genes in one study), suggesting that unknown genes with a strong predisposition effect for breast cancer are probably very limited (if any). Second, compared with those obtained from family-based studies, the life-time risk estimates for breast cancer from such large population-based studies are probably more reliable. These recalibrated estimates will help select cancer screening and risk-management strategies, from magnetic resonance imaging performed at an earlier age to preventative mastectomy (Costa and Saldanha, 2017), more wisely. Third, the studies establish the risk association of susceptibility genes not only for breast cancer but also for specific subtypes. Since breast cancer is a highly heterogeneous disease and different cancer subtypes have distinct treatment options (Sims et al., 2007), the reported subtype-specific risk associations will help choose more targeted preventative medications for different carriers, such as tamoxifen for ER-positive breast cancer (Hanker et al., 2020).

The BCAC and CARRIER studies represent a major step towards a more accurate breast cancer risk assessment for women in the general population, but they are just a beginning. First, not every germline variant in the same susceptibility gene confers the same risk, and there are no unifying guidelines to interpret the effect of a specific variant. For example, in the BCAC analysis, variants in a gene were classified as protein-truncating or missense variants, and these two classes were evaluated separately, whereas the CARRIER study considered the “pathogenic” annotation in the ClinVar database (Landrum et al., 2018). Therefore, further efforts are required to annotate the effect of individual variants, especially missense variants in the less-studied susceptibility genes. Second, despite a common set of identified susceptibility genes, the absolute odds ratios of some genes varied considerably between the two studies. Further efforts are warranted to revisit the relative risk estimates through meta-analysis or when more similar studies are available (Antoniou et al., 2003). Third, as a cancer risk is affected by other genetic (e.g., race and family history) and lifestyle (e.g., smoking history) factors, it is important to incorporate such information with data for germline variants in susceptibility genes to build a more reliable risk model for breast cancer (Lee et al., 2019). Finally, a similar population-based strategy should be extended to other major cancer types, such as lung and colorectal cancers. To this end, well-organized consortia, such as BCAC and CARRIER, are essential because of the coordination of large-scale sample cohorts and consistent efforts for data update and sharing. With these and more such efforts, we will be better situated to develop and implement more effective and generalizable cancer prevention and screening strategies, not only for breast cancer but for many other cancer types.