Abstract
This study uses Women’s Health Initiative data to compare the prevalence of pathogenic variants (PVs) in breast cancer susceptibility genes in postmenopausal women with vs without breast cancer to guide decisions about who should undergo PV testing.
Germline genetic testing for pathogenic variants (PVs) in cancer susceptibility genes after breast cancer diagnosis may inform cancer treatment, prevention, and testing of relatives. Whether testing should be performed depends partly on PV prevalence, which may be low in the general population but higher in women with risk factors (eg, young diagnosis age, family history). For the best-characterized breast cancer susceptibility genes, BRCA1, BRCA2, or both (BRCA1/2), a minimum PV prevalence of 2.5% to 10% has been recommended for testing.1 However, guidelines vary in testing all breast cancer patients2 vs only those with features suggestive of hereditary risk.3 Most guidelines do not address testing among postmenopausal women without hereditary risk factors, the most common subgroup of breast cancer patients, as PV prevalence data are lacking. This study’s purpose was to determine PV prevalence among women diagnosed with breast cancer after menopause vs the background prevalence among cancer-free postmenopausal women.
Methods
The Women’s Health Initiative (WHI) is a prospective study of morbidity and mortality that enrolled 161 808 postmenopausal women aged 50 to 79 years at 40 US sites from 1993 through 1998. Research was approved by each institution’s review board, and all participants provided written informed consent.4 We performed a nested case-control study of women without personal history of breast cancer at WHI enrollment who were diagnosed with invasive breast cancer (case participants) or remained cancer free (control participants) as of September 20, 2017. Case and control participants were unmatched and randomly selected from all WHI participants having banked DNA samples.
Next-generation sequencing and large rearrangement analysis was performed by Myriad Genetics using a panel of 28 genes; among these, BRCA1/2, ATM, BARD1, CDH1, CHEK2, NBN, PALB2, STK11, and TP53 were considered breast cancer associated.1,5 Variants were classified as pathogenic or likely pathogenic, of uncertain significance, or benign or likely benign.6
Gene-specific prevalence was reported as a percentage for case and control participants with 95% CIs and P values calculated using the exact binomial method for CIs and χ2 tests for P values. The percentages of PV carriers meeting National Comprehensive Cancer Network 2019 testing guidelines (most relevant to oncology practice and relying primarily on diagnosis age and family history) were analyzed.1 Association with age was examined. Tests for PV prevalence trend by age among BRCA1/2 and other breast cancer–associated genes were performed using χ2 tests (R version 3.5.3) (2-sided P < .05 was considered statistically significant).
Results
Among 4517 women, the median age at breast cancer diagnosis was 73 years for case participants (n = 2195) and 81 years at last follow-up for control participants (n = 2322). In the case group, 66.3% were white vs 84.9% in the control group. PVs were detected in 241 women (148 case participants [6.74%; 95% CI, 5.73%-7.87%] and 93 control participants [4.01%; 95% CI, 3.24%-4.88%]; P < .001). A PV was detected in any breast cancer–associated gene in 3.55% (95% CI, 2.82%-4.42%) of case participants and 1.29% (95% CI, 0.87%-1.84%) of control participants (P < .001). Of women with BRCA1/2 PVs, 30.8% of case participants and 20% of control participants met testing guidelines; of women with PVs in other breast cancer–associated genes, 34% of case participants and 16% of control participants met testing guidelines (Table 1).
Table 1. Pathogenic Variant Prevalence by Gene for Invasive Breast Cancer Case Participants and Cancer-Free Control Participants.
| Gene With Pathogenic Varianta | Breast Cancer Case Group (n = 2195) | Control Group (n = 2322) | ||
|---|---|---|---|---|
| No. (%) [95% CI] | Met Testing Guidelines, No./Total (%)b | No. (%) [95% CI] | Met Testing Guidelines, No./Total (%)b | |
| BRCA1 and BRCA2 | 26 (1.18) [0.78-1.73]c | 8/26 (30.8) | 5 (0.22) [0.07-0.50]c | 1/5 (20.0) |
| BRCA1 | 7 (0.32) | 3/7 (42.9) | 1 (0.04) | 1/1 (100) |
| BRCA2 | 19 (0.87) | 5/19 (26.3) | 4 (0.17) | 0/4 |
| Other breast cancer–associated genes | 53 (2.41) [1.81-3.15]c | 18/53 (34.0) | 25 (1.08) [0.70-1.58]c | 4/25 (16.0) |
| ATM | 16 (0.73) | 6/16 (37.5) | 7 (0.30) | 2/7 (28.6) |
| BARD1 | 4 (0.18) | 1/4 (25.0) | 0 | 0 |
| BRIP1 | 4 (0.18) | 0/4 | 6 (0.26) | 1/6 (16.7) |
| CDH1 | 1 (0.05) | 0/1 | 0 | 0 |
| CHEK2 | 13 (0.59) | 3/13 (23.1) | 6 (0.26) | 0/6 |
| NBN | 1 (0.05) | 1/1 (100) | 3 (0.13) | 0/3 |
| PALB2 | 14 (0.64) | 7/14 (50.0) | 2 (0.09) | 1/2 (50.0) |
| STK11 | 0 | 0 | 1 (0.04) | 0/1 |
| TP53 | 0 | 0 | 1 (0.04) | 0/1 |
| Any breast cancer–associated gened | 78 (3.55) [2.82-4.42]c | 25/78 (32.1) | 30 (1.29) [0.87-1.84]c | 5/30 (16.7) |
| Lynch syndrome genes | 6 (0.27) | 7 (0.30) | ||
| MLH1 | 0 | 1 (0.04) | ||
| MSH6 | 3 (0.14) | 1 (0.04) | ||
| PMS2 | 3 (0.14) | 5 (0.22) | ||
| Other tested genes | 65 (2.96) | 56 (2.41) | ||
| APC | 21 (0.96) | 18 (0.78) | ||
| CDKN2A (P16) | 3 (0.14) | 0 | ||
| MYH | 38 (1.73) | 36 (1.55) | ||
| RAD51C | 3 (0.14) | 1 (0.04) | ||
| RAD51D | 0 | 1 (0.04) | ||
| Any tested gene | 148 (6.74) [5.69-7.79]c | 93 (4.01) [3.21-4.80]c | ||
No pathogenic variants were found among case or control participants in BMPR1A, CDK4, CDKN2A, MSH2, POLD1, POLE, PTEN, or SMAD4.
National Comprehensive Cancer Network guidelines for hereditary breast and ovarian cancer are relevant for breast cancer–associated genes and based primarily on age at breast cancer diagnosis and family cancer history.1
P < .001 for comparison of percentages of pathogenic variants in case vs control participants.
Defined as BRCA1, BRCA2, ATM, BARD1, BRIP1, CDH1, CHEK2, NBN, PALB2, STK11, and TP53. Among breast cancer case participants, 1 had pathogenic variants in BRCA2 and PALB2 and met testing guidelines and 1 had pathogenic variants in APC and BARD1 and did not meet testing guidelines. Among control participants, 1 had pathogenic variants in BRIP1 and PALB2 and did not meet testing guidelines.
For BRCA1/2, PV prevalence was 2.21% (95% CI, 0.82%-4.76%) among case participants diagnosed when younger than 65 years and 1.09% (95% CI, 0.67%-1.68%) among case participants diagnosed at 65 years or older. There was no trend by age for PV prevalence in BRCA1/2 (P = .34) or other breast cancer–associated genes (P = .54) (Table 2).
Table 2. Pathogenic Variant Prevalence for Women Diagnosed With Breast Cancer by Diagnosis Age.
| Genea | Age at Breast Cancer Diagnosis, No. (%), yb | P Value for Trend | ||||||
|---|---|---|---|---|---|---|---|---|
| 50-59 | 60-64 | 65-69 | 70-74 | 75-79 | 80-84 | ≥85 | ||
| BRCA1 and BRCA2 (BRCA1/2) | 2 (2.33) | 4 (2.16) | 4 (1.06) | 6 (1.11) | 5 (1.08) | 3 (1.00) | 2 (1.32) | .34 |
| BRCA1 | 2 (2.33) | 0 | 3 (0.79) | 1 (0.19) | 0 | 1 (0.33) | 0 | |
| BRCA2 | 0 | 4 (2.16) | 1 (0.26) | 5 (0.93) | 5 (1.08) | 2 (0.67) | 2 (1.32) | |
| Other breast cancer–associated genes | 0 | 3 (1.62) | 7 (1.85) | 17 (3.15) | 13 (2.82) | 6 (2.00) | 2 (1.32) | .54 |
| ATM | 0 | 2 (1.08) | 2 (0.53) | 5 (0.93) | 1 (0.22) | 3 (1.00) | 1 (0.66) | |
| BARD1 | 0 | 1 (0.54) | 0 | 3 (0.56) | 0 | 0 | 0 | |
| BRIP1 | 0 | 0 | 1 (0.26) | 1 (0.19) | 1 (0.22) | 0 | 0 | |
| CDH1 | 0 | 0 | 0 | 1 (0.19) | 0 | 0 | 0 | |
| CHEK2 | 0 | 0 | 0 | 3 (0.56) | 7 (1.52) | 0 | 1 (0.66) | |
| NBN | 0 | 0 | 0 | 1 (0.19) | 0 | 0 | 0 | |
| PALB2 | 0 | 0 | 4 (1.06) | 3 (0.56) | 4 (0.87) | 3 (1.00) | 0 | |
| Lynch syndrome genes | 0 | 1 (0.54) | 0 | 0 | 0 | 4 (1.33) | 0 | |
| MSH6 | 0 | 0 | 0 | 0 | 0 | 2 (0.67) | 0 | |
| PMS2 | 0 | 1 (0.54) | 0 | 0 | 0 | 2 (0.67) | 0 | |
| Other tested genes | 3 (3.49) | 8 (4.32) | 5 (1.32) | 17 (3.15) | 14 (3.04) | 10 (3.33) | 5 (3.29) | |
| APC | 1 (1.16) | 1 (0.54) | 1 (0.26) | 6 (1.11) | 4 (0.87) | 4 (1.33) | 3 (1.97) | |
| P16 | 0 | 1 (0.54) | 0 | 0 | 1 (0.22) | 1 (0.33) | 0 | |
| MYH | 2 (2.33) | 6 (3.24) | 2 (0.53) | 10 (1.86) | 9 (1.95) | 5 (1.67) | 2 (1.32) | |
| RAD51C | 0 | 0 | 2 (0.53) | 1 (0.19) | 0 | 0 | 0 | |
| Any tested gene | 5 (5.81) | 15 (8.11) | 16 (4.22) | 39 (7.24) | 32 (6.94) | 23 (7.67) | 9 (5.92) | .55 |
| 95% CI | 1.91-13.05 | 4.61-13.02 | 2.43-6.77 | 5.20-9.76 | 4.80-9.66 | 4.92-11.28 | 2.74-10.94 | |
No pathogenic variants were found among breast cancer cases in APC, BMPR1A, CDK4, CDKN2A, EPCAM, GREM1, MLH1, MSH2, POLD1, POLE, PTEN, RAD51D, SMAD4, STK11, and TP53.
For BRCA1/2, the prevalence of pathogenic variants among case participants diagnosed at younger than 65 years vs 65 years or older (<65 years: n = 6 [2.21%]; 95% CI, 0.82%-4.76% vs ≥65 years: n = 20 [1.09%]; 95% CI, 0.67%-1.68%; P = .21).
Discussion
In this study, 3.55% of unselected, postmenopausal patients with breast cancer carried PVs in breast cancer–associated genes, a 3-fold higher prevalence than among cancer-free control participants and without decrease by age. Women diagnosed when younger than 65 years had similar probability of BRCA1/2 PVs as Ashkenazi Jewish individuals (≈2.5%), for whom testing is supported. Limitations: women chose to participate in WHI so the study may not represent all US women, a small number of PVs, and wide CIs in some subgroups.
These data on the prevalence of PVs in breast cancer susceptibility genes among postmenopausal women should inform testing guidelines. Among postmenopausal patients with breast cancer, PV prevalence may be high enough to warrant testing even in the absence of early diagnosis age or family history.
Section Editor: Jody W. Zylke, MD, Deputy Editor.
References
- 1.Daly MB, Pilarski R, Berry M, et al. . NCCN guidelines insights: genetic/familial high-risk assessment: breast and ovarian, version 2.2017. J Natl Compr Canc Netw. 2017;15(1):9-20. doi: 10.6004/jnccn.2017.0003 [DOI] [PubMed] [Google Scholar]
- 2.American Society of Breast Surgeons Consensus Guideline on Genetic Testing for Hereditary Breast Cancer. Published 2019. Accessed November 1, 2019. https://www.breastsurgeons.org/docs/statements/Consensus-Guideline-on-Genetic-Testing-for-Hereditary-Breast-Cancer.pdf
- 3.Owens DK, Davidson KW, Krist AH, et al. ; US Preventive Services Task Force . Risk assessment, genetic counseling, and genetic testing for BRCA-related cancer: US Preventive Services Task Force recommendation statement. JAMA. 2019;322(7):652-665. doi: 10.1001/jama.2019.10987 [DOI] [PubMed] [Google Scholar]
- 4.Hays J, Hunt JR, Hubbell FA, et al. . The Women’s Health Initiative recruitment methods and results. Ann Epidemiol. 2003;13(9)(suppl):S18-S77. doi: 10.1016/S1047-2797(03)00042-5 [DOI] [PubMed] [Google Scholar]
- 5.Easton DF, Pharoah PD, Antoniou AC, et al. . Gene-panel sequencing and the prediction of breast-cancer risk. N Engl J Med. 2015;372(23):2243-2257. doi: 10.1056/NEJMsr1501341 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Eggington JM, Bowles KR, Moyes K, et al. . A comprehensive laboratory-based program for classification of variants of uncertain significance in hereditary cancer genes. Clin Genet. 2014;86(3):229-237. doi: 10.1111/cge.12315 [DOI] [PubMed] [Google Scholar]