Leila Dorling, PhD
In a study reported in TheNew England Journal of Medicine, Leila Dorling, PhD, of the Centre for Cancer Genetic Epidemiology, Departments of Public Health and Primary Care, University of Cambridge, United Kingdom, and colleagues in the international Breast Cancer Association Consortium (BCAC), identified the risk of breast cancer associated with germline protein–truncating and rare missense variants of putative susceptibility genes in a large population of breast cancer cases and controls from BCAC studies.1
The entire study population consisted of 53,461 controls and 60,466 women with invasive (90.3%) or in situ tumors (6.9%) or tumors of unknown invasiveness (2.7%) from 44 BCAC studies, including 14 enriched by patients with a family history of breast cancer. The population-based analysis reviewed herein included 48,826 patients with invasive or in situ breast cancer and 50,703 controls from 30 BCAC population-based studies (ie, studies with no oversampling of patients with a family history of breast cancer) performed in individuals from 25 predominately European countries. Sequencing of germline DNA was performed on a panel of 34 putative susceptibility genes. Risks of breast cancer overall and for tumor subtypes were assessed for protein-truncating variants and rare missense variants (those with population frequency < 0.001) of these genes.
Findings for Protein-Truncating Variants
Protein-truncating variants in five genes were associated with risk of breast cancer overall with a P value of less than .0001: ATM (odds ratio [OR] = 2.10), BRCA1 (OR = 10.57), BRCA2 (OR = 5.85), CHEK2 (OR = 2.54; OR = 2.66 for c.1100delC variant and 2.13 for other variants), and PALB2 (OR = 5.02). Evidence of an association with increased risk of breast cancer overall was found for variants in seven additional genes: BARD1 (OR = 2.09, P = .00098), RAD51C (OR = 1.93, P = .0070), RAD51D (OR = 1.80, P = .018), PTEN (OR = 2.25, P = .10; P = .0040 in population from all 44 studies), NF1 (OR = 1.76, P = .068), TP53 (OR = 3.06, P = .17), and MSH6 (OR = 1.96, P = .013).
Of the remaining 22 genes (ABRAXAS1, AKT1, BABAM2, BRIP1, CDH1, EPCAM, FANCC, FANCM, GEN1, MEN1, MLH1, MRE11, MSH2, MUTYH, NBN, PIK3CA, PMS2, RAD50, RECQL, RINT1, STK11, XRCC2), all but 3 (AKT1, MSH2, and STK11, none with statistically significant odds ratios) had an upper limit of the 95% confidence interval of the odds ratio of < 2.0 (no estimated moderate or high risk).
Among the 12 genes with evidence of an association with breast cancer overall, odds ratios were higher for estrogen receptor (ER)-positive vs ER-negative disease for variants in ATM (2.33 vs 1.01, P = .00055 for difference) and CHEK2 (2.67 vs 1.64, P = 3.6 × 10−5 for difference). CHEK2 also had evidence of an association with ER-negative, non–triple-negative breast cancer (OR = 2.53, 95% confidence interval [CI] = 1.75–3.67), but not with triple-negative disease. Odds ratios were higher for ER-negative vs ER-positive disease for variants in BARD1, BRCA1, BRCA2, PALB2, RAD51C, and RAD51D (P < .05 for all). Of these genes, three had a stronger association with triple-negative disease vs ER-negative, non–triple-negative disease: BARD1 (P = .044 for difference), BRCA1 (P = 9.6 × 10−17), and BRCA2 (P = 7.8 × 10−5). For variants in BRCA1, BRCA2, and PALB2, odds ratios were higher for invasive vs in situ tumors; for ATM and CHEK2, odds ratios were similar for invasive vs in situ tumors.
Odds ratios for breast cancer declined significantly with increasing age for variants in BRCA1, BRCA2, CHEK2, PALB2, PTEN, and TP53 (P < .01 for all). In an analysis combining age-specific estimated odds ratios with reported population incidence in the United Kingdom, the estimated absolute risk of overall breast cancer by age 80 years was > 30% (high risk) for individuals with variants in BRCA1, BRCA2, and PALB2 and within the moderate risk range of 17% to 30% for variants in ATM, BARD1, CHEK2, RAD51C, and RAD51D.
Findings for Missense Variants
Evidence of an association with breast cancer overall for aggregated rare missense variants was found for six genes: CHEK2 (OR = 1.42, P = 2.5 × 10−11; P = 2.9 × 10−18 in population from all 44 studies), ATM (OR= 1.06, P = .051; P = .0010 in all studies), TP53 (OR = 1.10, P = .32; P = .00080 in all studies), BRCA1 (OR = 1.11, P = .010; P = .027 in all studies), CDH1 (OR = 1.10, P = .096; P = .042 in all studies), and RECQL (OR = 1.12, P = .047; P = .036 in all studies). CHEK2 (P = 9.1 × 10−5 for difference) and CDH1 (P = .012 for difference) exhibited a stronger association with ER-positive vs ER-negative disease, and BRCA1 (P = .01 for difference) had a stronger association with ER-negative vs ER-positive breast cancer. Odds ratios decreased with increasing age for BRCA1 (P = .0026), CHEK2 (P = .00022), and TP53 (P = .00023).
In an analysis of missense variants in BRCA1, BRCA2, and TP53, there was strong evidence of association with breast cancer overall for those variants classified as pathogenic according to standard criteria vs other aggregated missense variants in the same genes: odds ratios were 16.11 (95% CI = 5.83–44.50) vs 1.06 (95% CI = 0.98–1.15) for BRCA1, 5.68 (95% CI = 2.62–12.29) vs 0.97 (95% CI = 0.92–1.02) for BRCA2, and 2.91 (95% CI = 1.71–4.98) vs 0.94 (95% CI = 0.77–1.14) for TP53. The risk for breast cancer associated with the guideline pathogenic missense variants was similar to that associated with the protein-truncating variants for the three genes.
The investigators concluded: “The results of this study define the genes that are most clinically useful for inclusion on panels for the prediction of breast cancer risk, as well as provide estimates of the risks associated with protein-truncating variants, to guide genetic counseling.”
DISCLOSURE: The study was funded by European Union Horizon 2020 programs, Wellcome Trust, and Cancer Research UK. Dr. Dorling reported no conflicts of interest.