Association Between Genetic Testing for Hereditary Breast Cancer and Contralateral Prophylactic Mastectomy Among Multiethnic Women Diagnosed With Early-Stage Breast Cancer

PURPOSE:

Increasing usage of multigene panel testing has identified more patients with pathogenic or likely pathogenic (P or LP) variants in low-moderate penetrance genes or variants of uncertain significance (VUS). Our study evaluates the association between genetic test results and contralateral prophylactic mastectomy (CPM) among patients with breast cancer.

METHODS:

We conducted a retrospective cohort study among women diagnosed with unilateral stage 0-III breast cancer between 2013 and 2020 who underwent genetic testing. We examined whether genetic test results were associated with CPM using multivariable logistic regression models.

RESULTS:

Among 707 racially or ethnically diverse women, most had benign or likely benign (B or LB) variants, whereas 12.5% had P or LP and 17.9% had VUS. Racial or ethnic minorities were twice as likely to receive VUS. Patients with P or LP variants had higher CPM rates than VUS or B or LB (64.8% v 25.8% v 25.9%), and highest among women with P or LP variants in high-penetrance genes (74.6%). On multivariable analysis, P or LP compared with B or LB variants were significantly associated with CPM (odds ratio = 4.24; 95% CI, 2.48 to 7.26).

CONCLUSION:

Women with P or LP variants on genetic testing were over four times more likely to undergo CPM than B or LB. Those with VUS had similar CPM rates as B or LB. Our findings suggest appropriate genetic counseling and communication of cancer risk to multiethnic breast cancer survivors.

INTRODUCTION

Although the majority of breast cancer cases occur in patients without a known family history of breast cancer, approximately 5%-10% of breast cancers are considered hereditary, or because of an inherited pathogenic or likely pathogenic (P or LP) variant in a breast cancer susceptibility gene.¹ P or LP variants in BRCA1 and BRCA2 account for more than 80% of hereditary breast cancers, and confer an approximately 50%-60% absolute risk of developing breast cancer by age 80 years.^2,3 However, P or LP variants in other breast cancer susceptibility genes, including high-penetrance genes such as CDH1, PALB2, PTEN, STK11, and TP53, and moderate-penetrance genes such as ATM, CHEK2, NBN, and NF1, also confer increased risk of breast cancer.^4-6 In general, high-penetrance variants are associated with more than 40%-50% absolute lifetime risk of breast cancer, and moderate-penetrance variants with more than 20%-25% absolute lifetime risk.

National organizations, including the National Comprehensive Cancer Network (NCCN), ASCO, and American Society of Breast Surgeons (ASBrS), have published guidelines regarding genetic assessment for hereditary breast cancer among women affected by breast cancer.^7-9 Although NCCN and ASCO determine eligibility for genetic testing based upon family history of cancer, ancestry, age at diagnosis, and pathologic subtype of breast cancer, in 2019, ASBrS recommended that all women with a personal history of breast cancer, regardless of family history of cancer and age at diagnosis, should be offered genetic testing to inform the risks of future cancers in patients and their family members.

Following the 2013 US Supreme Court ruling against gene patents, genetic testing has become more accessible to patients, with increasing use of multigene panel testing.¹⁰ However, as multigene panel testing is increasingly used, rates of detection of variants of uncertain significance (VUS), which are higher in racial or ethnic minorities, and P or LP variants in low-moderate penetrance breast cancer susceptibility genes are also increasing, with unclear implications for management.^11,12 This trend has led to concern that patients and providers may overestimate breast cancer risk among those with VUS, leading to inappropriate use of risk-reducing strategies.

Contralateral prophylactic mastectomy (CPM) is one strategy to reduce the risk of contralateral breast cancer among women treated for unilateral, early-stage breast cancer. The risk of contralateral breast cancer among women with P or LP in BRCA1 or BRCA2 is two to three times higher than those without P or LP in BRCA1 or BRCA2 and ranges up to 36% at 15 years after diagnosis,^13,14 with greatest risk among those who were younger than 50 years at initial diagnosis and/or with first-degree family history of breast cancer.¹⁵ The NCCN only recommends CPM among women with P or LP variants in high-penetrance breast cancer susceptibility genes, such as BRCA1, BRCA2, PTEN, and TP53.⁷ Similarly, ASBrS recommends CPM for women with P or LP variants in BRCA1 or BRCA2, and could be considered for those with other high-penetrance genes including CHEK2, PALB2, and CDH1, but otherwise discourages CPM in average-risk women.¹⁶ However, the rate of CPM has consistently risen over the past two decades, particularly among women age < 50 years at diagnosis, without demonstrated survival benefit.^17-20 To address this understudied area, we evaluated how results of multigene panel testing affected rates of CPM among a racially or ethnically diverse population of women treated for unilateral early-stage breast cancer.

METHODS

Study Design and Study Population

We conducted a retrospective cohort study of women diagnosed with stage 0-III breast cancer between January 2013 and May 2020 who also received genetic testing, and had electronic medical records available for review at Columbia University Irving Medical Center (CUIMC) in New York, NY. This study was conducted in accordance with STROBE guidelines (Data Supplement, online only). Patients were identified through the New York Presbyterian Hospital (NYPH) Tumor Registry at CUIMC and genetic testing portals (Color and Invitae). Patients with bilateral synchronous breast cancer, male breast cancer, distant metastases, lobular carcinoma in situ, missing breast surgery information, or missing genetic test results were excluded from the analysis. This study was approved by the CUIMC Institutional Review Board.

Demographic and Clinical Variables

We collected demographic and clinical information from the CUIMC electronic health record (EHR) and NYPH Tumor Registry. Demographic characteristics included age at diagnosis (< 40, 40-49, 50-59, and > 60 years), race or ethnicity (non-Hispanic White, non-Hispanic Black, Hispanic, Asian, and other or unknown), marital status (married and unmarried), and primary health insurance status (Medicare, Medicaid, private insurance, and uninsured or other or unknown). Other clinical characteristics included stage at diagnosis (0, I, II, III, or unknown), first- or second-degree family history of breast cancer (yes or no), and receipt of chemotherapy, endocrine therapy, and radiation therapy (yes or no).

Genetic test results were coded as pathogenic or likely pathogenic (P or LP), VUS, or benign or likely benign (B or LB) variants. Patients with both P or LP and VUS were documented as P or LP. We also documented the number of genes tested, and which genes were affected with P or LP or VUS. Genes with P or LP or VUS were further categorized as high penetrance, moderate penetrance, or other, according to prior published risk of breast cancer associated with each gene. The genes considered high penetrance were BRCA1, BRCA2, CDH1, PALB2, PTEN, STK11, and TP53. The genes considered moderate penetrance were ATM, CHEK2, NBN, and NF1.⁵ All other genes were considered low penetrance.

The primary outcome of interest was receipt of contralateral prophylactic mastectomy (CPM). The type of breast surgery (bilateral mastectomy, unilateral mastectomy, and lumpectomy) was ascertained from operative reports or outpatient breast clinic notes. Patients who underwent multiple breast surgeries were classified by their most advanced breast surgery (bilateral mastectomy > unilateral mastectomy > lumpectomy). To ascertain whether surgery was performed before or after genetic tests results were available, we compared the date of the genetic testing reports and the date of the most advanced breast surgery. When exact dates were not available, chart review of outpatient notes was conducted to ascertain temporality.

Statistical Analysis

Descriptive statistics of baseline characteristics were stratified according to most advanced surgery type received. Chi-squared test and Fisher's exact test for categories with five or fewer patients were used to assess differences in distribution of characteristics among women receiving different surgery types. Chi-squared test was also used to assess differences in genetic testing results by age, race or ethnicity, and insurance status.

We conducted univariable and multivariable logistic regression analyses to estimate the odds ratio associated with each variable on the receipt of CPM. We hypothesized that family history of breast cancer and stage would be potential confounders, and we controlled for these variables within our multivariable logistic regression analysis. We also ran a sensitivity analysis for the univariable and multivariable models for patients who received genetic testing before their most advanced breast surgery. To limit bias from external providers or input from friends or family, we ran additional sensitivity analyses with patients who had undergone surgery within 1 year of receiving genetic testing results, and with patients who had undergone surgery within 1 year after receiving genetic testing results. By limiting follow-up time on our sensitivity analyses, it provides less time for external factors unmeasured in this retrospective study to exert an influence on the outcome of interest. Variables were included in the multivariable logistic regression model if P values from the univariable logistic models were < 0.10 or if they had known clinical importance. All statistical analyses were conducted using SAS OnDemand (SAS Institute, Cary, NC), and a P value < .05 was considered statistically significant for the multivariable logistic regression.

RESULTS

From January 2013 to May 2020, a total of 826 women were diagnosed with breast cancer and also underwent genetic testing. Of these, 119 patients were excluded for stage IV disease (n = 59), missing surgical reports (n = 32), bilateral synchronous breast cancer (n = 11), duplicates in the tumor registry (n = 7), pending or canceled genetic test results (n = 9), or having a prior bilateral mastectomy (n = 1; Appendix Fig A1, online only).

Among the 707 evaluable women, the average age at diagnosis was 48.9 years (standard deviation = 11.2). Forty-five percent of women identified as non-Hispanic White, 12% non-Hispanic Black, 28% Hispanic, 11% Asian, and 4% other or unknown (Table 1). Fifty-two percent of women were married, and 52% percent had private insurance, 16% had Medicare, 26% had Medicaid, and 7% had no health insurance or unknown insurance status. The distribution of stage at diagnosis was 11% stage 0, 42% stage I, 37% stage II, 10% stage III, and < 1% with unknown stage. About 76.5% of patients received endocrine therapy, 60% had chemotherapy, and 63.5% had radiation therapy. More than half (55%) had a first- or second-degree family history of breast cancer. Approximately 69.5% had B or LB genetic test results, whereas 18.1% had VUS and 12.4% had P or LP variants. The median number of genes tested was 17 (range, 1-109). Definitive breast surgery occurred after receipt of genetic test results in 54% of patients. The median time from genetic testing results to definitive surgery was 24 days, with a mean of 52 days. The first quartile of patients had surgery 90 days before genetic testing results occurred and the third quartile of patients had surgery 121 days after genetic testing. Only 37 patients had surgery more than 1 year after receiving genetic testing results.

TABLE 1.

Baseline Characteristics of Women Who Were Diagnosed With Early-Stage Breast Cancer and Underwent Genetic Testing for Hereditary Breast and Ovarian Cancer (HBOC) at Columbia University Irving Medical Center, New York, NY, From 2013 to 2020, Stratified by Type of Breast Surgery

In total, 217 patients (30.7%) underwent bilateral mastectomy, ie, CPM. Patients undergoing CPM were younger at diagnosis (P < .001), more likely to have Medicaid and less likely to have Medicare (P < .001), and have stage II or higher disease (P < .001). Those who underwent CPM were more likely to have P or LP variants on genetic testing (P < .001) and more likely to have undergone definitive breast surgery after receipt of genetic testing results (P < .001). However, there were no statistically significant differences in CPM rates by race or ethnicity, marital status, family history of breast cancer, or year of diagnosis.

The Hosmer and Lemeshow test for goodness-of-fit for our final multivariable model resulted in a chi-square of 9.69 and P value of .287, demonstrating a good fit. In multivariable analysis, the strongest predictor of receipt of CPM was P or LP variants on genetic testing (Table 2). Compared to those with B or LB variants, patients with P or LP were over four times more likely to undergo CPM (odds ratio [OR] = 4.24; 95% CI, 2.48 to 7.26; P < .001). Patients with VUS underwent CPM at similar rates to those with B or LB variants (OR = 0.78; 95% CI, 0.47 to 1.29; P = .331). Among those with VUS or B or LB, the main factors associated with CPM were age at diagnosis, family history of breast cancer, and receipt of endocrine therapy. We conducted multiple sensitivity analyses to compare outcomes for women undergoing breast surgery after receiving genetic test results, women undergoing breast surgery within 1 year after receiving genetic test results, and women undergoing breast surgery any time within 1 year of receiving genetic test results. The results of all three sensitivity analyses were comparable with respect to the association between genetic test results and CPM rates, with women receiving P or LP genetic testing results being over four times more likely to undergo CPM compared to women with B or LB, and women with VUS having similar rates of CPM as those with B or LB on all sensitivity analyses (Appendix Table A2, online only).

TABLE 2.

Univariable and Multivariable Analyses of Factors Associated With Receipt of Bilateral Mastectomy Among Women Who Were Diagnosed With Early-Stage Breast Cancer and Underwent Genetic Testing for Hereditary Breast and Ovarian Cancer (HBOC) at Columbia University Irving Medical Center, New York, NY, 2013-2020

On multivariable analysis (Table 2), those with a family history of breast cancer were also more likely to undergo CPM (OR = 1.56; 95% CI, 1.07 to 2.29; P = .022). The younger the patients are at diagnosis, the more likely they are to undergo CPM. Patients undergoing CPM were half as likely to receive endocrine therapy and almost twice as likely to receive chemotherapy. Radiation therapy was not included in the multivariable analysis because patients who undergo mastectomy typically do not require radiation. There were no statistically significant associations between race or ethnicity, primary insurance status, and stage at diagnosis with receipt of CPM.

Among the 88 patients with P or LP variants, 59 were in high-penetrance genes and 15 were in moderate-penetrance genes (Fig 1). Only one patient had P or LP variants in two genes (BRCA1 and MUTYH). Receipt of CPM varied by gene penetrance: 75% with P or LP variant in high-penetrance genes and 47% with P or LP variant in moderate-penetrance genes underwent CPM, compared to 43% with P or LP variant in other genes.

FIG 1.

Patients with pathogenic variants who underwent prophylactic mastectomy, stratified by gene penetrance. Total number of pathogenic variants shown is 89 because one patient had P or LP in two genes. P or LP, pathogenic or likely pathogenic.

The distribution of genetic test results (ie, P or LP, VUS, and B or LB) was also compared across age, racial or ethnic groups, and insurance (Appendix Table A1, online only). Those with B or LB were more likely to be 50 years or older at diagnosis, whereas those with P or LP or VUS were more likely to be younger, although this was not statistically significant. Racial or ethnic minorities, including non-Hispanic Blacks, Hispanics, and Asians, were more likely than non-Hispanic Whites to have a VUS result (P < .001). The frequency of VUS among non-Hispanic Whites was 10%, compared with 19% in non-Hispanic Blacks, 25% in Hispanics, and 27% in Asians. Patients with private insurance or Medicare were more likely to have B or LB, whereas patients with Medicaid or were uninsured were more likely to have VUS or P or LP (P = .002).

DISCUSSION

We observed that having a P or LP result on genetic testing was the strongest predictor of CPM among multiethnic women with early-stage breast cancer. Patients with P or LP variants were over four times more likely to undergo CPM than those with VUS or B or LB variants. Notably, those with VUS underwent CPM at similar rates to those with B or LB variants. Patients were also more likely to undergo CPM if they had a family history of breast cancer or were younger at diagnosis. When we examined the receipt of CPM among women with P or LP variants, we observed that thre quarters of those with P or LP variants in high-penetrance genes underwent CPM, as did nearly 50% with P or LP variants in moderate-penetrance genes. Of those with P or LP variants in high-penetrance genes, more than 80% were in BRCA1 or BRCA2.

Our study adds to the current literature on the impact of genetic testing on breast cancer surgical management by examining the effect of multigene panel testing on receipt of prophylactic surgery in a racially and ethnically diverse patient population. A main finding of our analysis was that women with early-stage breast cancer and P or LP variants on multigene panel testing were significantly more likely to undergo CPM than those with VUS or B or LB variants, consistent with current guidelines from the ASCO, NCCN, and American College of Medical Genetics and Genomics (ACMG).^7,21,22 In particular, most of the women with P or LP variants who underwent CPM had variants in hig-penetrance genes. Given concerns that with increasing uptake of multigene panel testing patients with VUS or P or LP variants in low-penetrance genes will undergo unnecessarily morbid surgeries,²³ it is reassuring that we did not observe that the receipt of VUS on multigene panel testing was associated with increased rates of CPM. This is particularly reassuring, given our large percentage of racial or ethnic minorities, among whom VUS are more common as a consequence of under-representation in studies validating multigene panel testing for hereditary breast cancer.^12,24 Also consistent with prior literature and with ASCO and NCCN guidelines, we observed that patients diagnosed at a younger age were more likely to undergo CPM, which is reasonable, given their higher chance of a second breast cancer event over time.^19,22

Our analysis also adds to the existing literature by characterizing the P or LP variants and VUS among women with breast cancer who did and did not undergo CPM by specific genes involved and by their penetrance. Previous studies have focused mainly on how BRCA1 and BRCA2 testing influenced rates of CPM,^25-28 and only more recently several have examined the impact of multigene panel testing results on CPM.^12,29,30 Similar to our findings, Kurian et al demonstrated that among women with early-stage breast cancer who underwent multigene panel testing, rates of CPM were highest among patients with P or LP, and patients with VUS underwent CPM at similar rates to those with B or LB. Kurian et al¹² reported that 79% of women with P or LP variants in BRCA1 or BRCA2 underwent CPM, compared with 37.6% of those with other pathogenic variants, 30.2% with VUS, and 35.3% with B or LB. However, their analysis combined all non-BRCA1 and BRCA2 genes as one category, such as PALB2 and BARD1, despite wide differences in associated breast cancer risk. Although variants have the potential to be reclassified with additional long-term data, examination of how these results guide surgical and other clinical decision making is important to inform recommendations, both at the level of national organizations and of individual or groups of providers. Current guidelines regarding prophylactic breast surgery largely address surgical management for patients with P or LP variants in BRCA1 or BRCA2 and select high-penetrance genes.^7,22 However, because guidelines on moderate- and low-penetrance genes are based upon low-quality evidence,²² more studies are needed to better assess the most appropriate management for patients with P or LP in other non-BRCA1 or BRCA2 genes.

The main strengths of this study are the large racially or ethnically diverse cohort of women diagnosed with early-stage breast cancer, and a recent data set with patients diagnosed after 2013, with the expansion of multigene panel testing.¹⁰ In addition, we had detailed clinical data documented, including breast surgery type and results of multigene panel testing. Furthermore, unlike prior large studies that relied on patient self-reported outcomes,²⁹ we were able to ascertain temporality of genetic testing results and surgery with exact dates and ascertain surgery type through operative reports. Limitations include that this was a retrospective study conducted at a single urban academic institution, with potential patient and provider biases that might limit the generalizability of our results to other patient populations. For example, we are unable to capture individual patient preferences and the influence of discussion with family or friends within this study. We are also unable to account for the influence of other providers that the patients may see. These questions are out of the scope of this paper; however, future studies should include qualitative assessments of patients and providers to better understand these additional influences on how particular patients choose to undergo CPM or not. We tried to limit this bias by running sensitivity analyses that restricted the patient population to those who had undergone surgery within 1 year after receiving genetic testing results. Because we found similar conclusions through this analysis, we believe that our conclusions are still valid despite this limitation. Although some patients received genetic testing results after definitive surgery, we included these patients in our sample because these patients had the option to undergo delayed CPM but chose not to. Furthermore, our sensitivity analysis including only women who had breast surgery after receiving genetic test results demonstrated similar findings.

In conclusion, we examined the association between results of multigene panel testing for hereditary breast cancer and receipt of CPM among women diagnosed with unilateral early-stage breast cancer and found that women with P or LP variants were four times more likely than those with VUS or B or LB variants to undergo bilateral mastectomy. In addition, women with VUS, a group enriched for racial or ethnic minorities, did not undergo risk-reducing breast surgery at higher rates than those with B or LB variants. Our findings suggest that genetic counseling and communication of cancer risk to multiethnic breast cancer survivors was effective. Given the current lack of consensus regarding the clinical management options for patients with P or LP variants in moderate- and low-penetrance cancer susceptibility genes and VUS results, future studies should investigate whether the impact of genetic test results on receipt of CPM affects long-term clinical outcomes in breast cancer survivors.

APPENDIX

FIG A1.

CONSORT diagram. GT, genetic testing.

TABLE A1.

Genetic Testing Results Stratified by Age, Race or Ethnicity, and Insurance

TABLE A2.

Sensitivity Analyses for Patients Receiving Genetic Testing Before Surgery, Patients Receiving Genetic Testing Within 1 Year of Surgery, and Patients Receiving Genetic Testing Within 1 Year Before Surgery

AUTHOR CONTRIBUTIONS

Conception and design: Tarsha Jones, Rita Kukafka, Katherine D. Crew

Finanical support: Meghna S. Trivedi, Katherine D. Crew

Provision of study materials or patients: Meghna S. Trivedi, Wendy K. Chung, Elana Levinson, Carrie Koval, Donna Russo, Ilana Chilton, Katherine D. Crew

Collection and assembly of data: Julia E. McGuinness, Boya Guo, Meghna S. Trivedi, Wendy K. Chung, Elana Levinson, Carrie Koval, Donna Russo, Ilana Chilton, Katherine D. Crew

Data analysis and interpretation: Vicky Ro, Julia E. McGuinness, Boya Guo, Meghna S. Trivedi, Tarsha Jones, Wendy K. Chung, Roshni Rao, Katherine D. Crew

Manuscript writing: All authors

Final approval of manuscript: All authors

Accountable for all aspects of the work: All authors

AUTHORS' DISCLOSURES OF POTENTIAL CONFLICTS OF INTEREST

Association Between Genetic Testing for Hereditary Breast Cancer and Contralateral Prophylactic Mastectomy Among Multiethnic Women Diagnosed With Early-Stage Breast Cancer

The following represents disclosure information provided by authors of this manuscript. All relationships are considered compensated unless otherwise noted. Relationships are self-held unless noted. I = Immediate Family Member, Inst = My Institution. Relationships may not relate to the subject matter of this manuscript. For more information about ASCO's conflict of interest policy, please refer to www.asco.org/rwc or ascopubs.org/op/authors/author-center.

Open Payments is a public database containing information reported by companies about payments made to US-licensed physicians (Open Payments).