Timely cancer genetic counseling and testing for young women with breast cancer: impact on surgical decision-making for contralateral risk-reducing mastectomy

Genetic testing (GT) can identify individuals with pathogenic/likely pathogenic variants (PV/LPVs) in breast cancer (BC) predisposition genes, who may consider contralateral risk-reducing mastectomy (CRRM). We report on CRRM rates in young women newly diagnosed with BC who received GT through a multidisciplinary clinic. Clinical data were reviewed for patients seen between November 2014 and June 2019. Patients with non-metastatic, unilateral BC diagnosed at age ≤ 45 and completed GT prior to surgery were included. Associations between surgical intervention and age, BC stage, family history, and GT results were evaluated. Of the 194 patients, 30 (15.5%) had a PV/LPV in a BC predisposition gene (ATM, BRCA1, BRCA2, CHEK2, NBN, NF1), with 66.7% in BRCA1 or BRCA2. Of 164 (84.5%) uninformative results, 132 (68%) were negative and 32 (16.5%) were variants of uncertain significance (VUS). Overall, 67 (34.5%) had CRRM, including 25/30 (83.3%) PV/LPV carriers and 42/164 (25.6%) non-carriers. A positive test result (p < 0.01) and significant family history were associated with CRRM (p = 0.02). For the 164 with uninformative results, multivariate analysis showed that CRRM was not associated with age (p = 0.23), a VUS, (p = 0.08), family history (p = 0.10), or BC stage (p = 0.11). In this cohort of young women with BC, the identification of a PV/LPV in a BC predisposition gene and a significant family history were associated with the decision to pursue CRRM. Thus, incorporation of genetic services in the initial evaluation of young patients with a new BC could contribute to the surgical decision-making process.

Genetic testing (GT) is recommended for women with newly diagnosed BCs whose personal and family cancer history meet specific criteria, as GT results may impact care by providing information regarding risks for additional primary BCs [14][15][16][17]. Contralateral BC risks are influenced by multiple factors, including age at first BC diagnosis, family history of BC, previous treatments, and underlying genetic predisposition [4,[18][19][20][21][22]. The risk for a contralateral BC is approximately 10% within 20 years after the initial diagnosis for unselected women with no known hereditary BC predispositions, [22] but may be as high as 53-65% for BRCA1/2 PV/LPV carriers [4,23]. Contralateral BC risks have not been well-defined for most other BC predisposition genes, but may be increased for women with ATM, PALB2, TP53, and CHEK2 truncating PV/LPVs [13,[24][25][26][27].
Women with PV/LPVs in genes with established contralateral BC risks could consider contralateral risk-reducing mastectomy (CRRM) at the time of surgical BC treatment, which provides a > 90% reduction in future BC risk [17,[27][28][29]. However, CRRM is not routinely recommended for women without a BC predisposition, or for women with PV/ LPVs in moderate-risk genes with insufficient data regarding second primary BC risks or long-term benefits of CRRM [27,28,[30][31][32]. Despite the limited data on survival benefits of CRRM outside of certain hereditary BC predispositions, CRRM rates have been increasing in the United States over the past two decades across cohorts of women with unilateral BC treated in various geographical locations between the mid-1990s to early 2010s. The increase in CRRM rates ranges from of 4.2 to 9.6% over 8 years, to an increase of 5.4 to 37.5% over 15 years [27,[33][34][35][36][37][38][39].
Genetic testing availability has also increased over the past few decades, including increased utilization of multigene panel testing, which has increased the PV/LPV detection rates in genes other than BRCA1/2 [40][41][42]. However, limited information exists regarding the impact of PV/LPVs in other BC predisposition genes on surgical decision-making outcomes for BC patients [27,[43][44][45][46]. Regarding GT results, it is also important to consider the impact of variants of uncertain significance (VUSs) on surgical decision-making. As VUSs do not carry any clinical implications, patients with a VUS should follow management recommendations based on their personal and family history, rather than on the GT result [17,47,48]. However, VUS identification can lead to patients being recommended the same management considerations as PV/LPV carriers, including CRRM [49][50][51][52][53].
Given broader GT availability and subsequent improvements in the ability to identify PV/LPVs in BC predisposition genes other than BRCA1/2 in recent years, we sought to further explore the impact of GT on surgical decision-making. This study focused on a cohort of young female patients with a new diagnosis of unilateral BC, who were uniformly offered genetic counseling and GT as part of a multidisciplinary clinic designed to efficiently provide patients with comprehensive care at the time of diagnosis.

Methods
Clinical data for patients seen in the multidisciplinary Breast Cancer Specialty Care Clinic at UPMC Magee-Womens Hospital (Pittsburgh, PA) between November 2014 and June 2019 were abstracted. This chart review was performed under a Quality Improvement project to evaluate the utility of this clinic for young breast cancer patients. The project was reviewed and approved by the University of Pittsburgh Medical Center Quality Improvement Review Committee and was deemed exempt from approval by the Institutional Review Board.
The Breast Cancer Specialty Care Clinic was established specifically to deliver centralized, multidisciplinary care for newly diagnosed breast cancer patients under the age of 50. Patients were scheduled within 2-3 weeks of diagnosis, and were seen by specialists from various disciplines as necessary, including breast surgery, medical oncology, reconstructive surgery, radiation oncology, cancer genetics, and infertility, often on the same day. If genetic testing was pursued, a sample was routinely collected on the same day, with results available in 2-3 weeks on average. Surgical treatment plans were finalized after GT results were available.
Patients were included in the analysis if they were: (1) female, (2) newly diagnosed with non-metastatic, unilateral BC (stage 0-III), (3) had no prior BC diagnoses or mastectomies, (4) ≤ 45 years old, (5) had completed genetic counseling and GT, and (6) had completed surgical treatment of BC by the end of the study period. Genetic testing could have been done at any time, including prior to their cancer diagnosis, as long as results were available before the time of surgery. Patients were excluded if they had incomplete information regarding treatment, family history, or GT results.
The following data was collected via chart review: age; race and ethnicity; BC stage (AJCC staging manual, 7th edition for BCs diagnosed between November 2014 and December 2017; 8th edition for diagnoses after January 2018); type of breast surgery performed (lumpectomy, mastectomy, mastectomy with CRRM); and family history of BC and/or ovarian cancer (OC). Patients were considered to have a "significant" family history if they had at least one first-or second-degree relative with: (1) OC at any age, (2) BC diagnosed ≤ 40 years old, (3) male BC at any age, or (4) ≥ 2 relatives with BC at any age.
The specific GT performed was not standardized across the study cohort; rather, testing for BRCA1/2 or a multigene panel was tailored to each patient, based on their personal and family history of cancer, as well as patient preference. Results were: negative for a PV/LPV in a BC predisposition gene, positive for a PV/LPV in a BC predisposition gene, or uncertain (VUS). The classifications of all PV/LPV and VUS results were verified against ClinVar (https:// www. ncbi. nlm. nih. gov/ clinv ar), with no discrepancies identified. Given that the focus of this study was to evaluate the impact of GT results on BC risk management, PV/LPVs in genes that have not been shown to be associated with increased BC risks were considered to be negative. Of note, PV/LPVs in the following genes were considered to have an association with BC risk during the study time frame: ATM, BARD1, BRCA1, BRCA2, CDH1, CHEK2, NBN (Slavic founder variant), NF1, PALB2, PTEN, STK11, TP53 [4][5][6][7][8][9][10][11][12]. Negative and VUS results were considered together as uninformative results.
Statistical analysis was performed using Chi-squared tests, Fisher's exact test, or t-tests, as appropriate, to examine the associations between surgical decisions (CRRM vs. no CRRM) and self-reported race/ethnicity, age, family history (significant vs. not significant), GT outcome (positive vs. uninformative), or BC stage (0-I vs. II and III). Multivariate analysis was also performed to evaluate for any association between surgical decision and these factors. A p-value of < 0.05 was considered to be statistically significant. Additionally, the association between age, family history, stage, and surgical decision was examined among patients with uninformative results. All analyses were performed using the SAS statistical software package (SAS/STAT software, Version 9.4 SAS Institute Inc., Cary, NC).

Patient population
A total of 199 eligible patients were seen between November 2014 and June 2019. Five patients declined genetic testing, resulting in 194 patients included in the final analysis. A total of 82 patients (42.3%) had breast-conserving surgery (BCS), 45 (23.2%) had unilateral mastectomy (UM), and 67 (34.5%) had mastectomy with CRRM.
Compared with patients who chose a BCS/UM, patients who had a mastectomy with CRRM were younger (median age 37 vs. 39 years; p = 0.03), more likely to have a significant family history of BC/OC, (41.8% vs. 16.5%; p < 0.01), and similar stage distributions (38.8% early stage vs. 31% late stage; p = 0.27). However, only a positive GT result (OR 10.6 (95% CI 3.66-30.88, p ≤ 0.001) and significant family history was statistically significantly associated with mastectomy and CRRM in the multivariate analysis (Table 1).

Factors associated with surgical decisions in patients with uninformative results
Of the subset of 164 patients with uninformative results, 122 (74.4%) opted for BCS/UM, and 42 (25.6%) had a mastectomy with CRRM (Table 2, Fig. 2

Discussion
In this study, the identification of a PV/LPV in a BC predisposition gene was significantly associated with surgical decision for young women newly diagnosed with a BC. Patients with a PV/LPV in any BC predisposition gene were more likely to opt for mastectomy with CRRM at the time of surgery. Overall, 15.5% of patients were found to have a PV/LPV in a moderate-or high-risk BC predisposition gene, which is similar to the 12-18% rate previously reported in other premenopausal BC cohorts [54][55][56]. Pathogenic/likely pathogenic variants in BRCA1/2 accounted for approximately two-thirds of PV/LPVs identified in this study, and most patients with a BRCA1/2 PV/ LPV opted to have a CRRM. Although there are limited data on long-term survival benefits, consideration of a CRRM is an established recommendation for women with a BRCA1/2 PV/LPV [4,[27][28][29]57]. The rates of CRRM in our study was also similar to those reported previously for individuals with a BRCA1/2 PV/LPV who had been diagnosed with BC [49,51,58].
In addition to BRCA1/2, PV/LPVs in moderate-risk genes, such as ATM and CHEK2, may confer increased risks for a second primary BC. However, data on estimated risks are limited, and the long-term benefit of CRRM in this setting is uncertain [8, 9, 25-28, 30, 43, 57]. Despite this lack of data, more than half of the patients with a PV/ LPV in a moderate-risk BC predisposition gene also opted to have CRRM, a rate similar to that previously reported among BC patients with a PV/LPV in a moderate-risk BC predisposition gene [44][45][46]. The number of patients identified to have a PV/LPV in a BC predisposition gene other than BRCA1/2 in this study was small, which precluded our ability to adequately assess the association between the identification of a PV/LPV in a moderate-risk BC predisposition gene and the decision to pursue CRRM at the time of surgical BC treatment. However, patients with a BRCA1/2 PV/LPV have been shown to be more likely than patients with PV/LPVs in other BC genes to choose CRRM [44][45][46]. Our study adds to the limited data available on surgical decisions on newly diagnosed patients with a PV/LPV in a moderate-risk BC predisposition gene.
Decisions regarding risk-reducing surgical intervention are complex, and likely influenced by factors other than knowledge of a genetic predisposition [59]. The risks of developing another primary BC after the first diagnosis have been shown to be higher in patients with younger age at first diagnosis, or a family history of BC/OC [18][19][20][21][22]. Consequently, earlier age at diagnosis and family history of BC/OC have been shown to be positively correlated with the decision to have CRRM in women with BC, regardless of genetic testing outcome [23, 36-38, 46, 60, 61]. In this study, we did not observe any association between age at diagnosis and surgical decision when adjusted for GT result. However, all patients included in this study were 45 years old or younger, which hindered the ability to evaluate the impact of age at diagnosis on surgical decision-making.
Similar to previous studies [23,[36][37][38][60][61][62], our data showed that a family history of BC diagnosed at age 40 or younger and/or OC was associated with a decision to have CRRM. Thus, although comprehensive GT could identify patients with increased risk for a second primary breast cancer, young patients diagnosed with BC might choose to have CRRM based on their perception of a high risk for contralateral BC in the presence of a significant family history [63]. Thus, the provision of pre-surgical genetic counseling and GT in this study may have helped to clarify estimated contralateral BC risks, and promoted risk-appropriate surgical decision-making among those without a genetic predisposition.
Higher CRRM rates with early stage disease have previously been reported [23,37,61], however, disease stage was not shown to be associated with CRRM in our study. Approximately a third of the patients had stages 0 or I at the time of diagnosis. Although there seemed to be a trend towards more CRRM among those with early staged disease, this finding did not reach statistical significance, possibly due to small number of patients.
Although cancer GT can provide valuable information in risk assessment and promote risk-appropriate management, some test results have the potential to lead to unnecessary interventions. Variants of uncertain significance (VUSs) in cancer predisposition genes do not have any clinical implications, and should not be used in risk assessment or risk management considerations. With appropriate pre-and post-test counseling, patients with VUSs in BC predisposition genes did not elect to undergo CRRM more often than average-risk women [45,46,50,51,64]. However, interpretation of a VUS result can often be challenging, and high-risk management recommendations based on VUS findings, have been reported, including CRRM [42,49,52,53]. As part of the multidisciplinary clinic, all patients in our study received pre-and post-test genetic counseling from a genetics specialist to discuss test implications and possible results. In our study, among those with an uninformative result, the CRRM rate was not significantly different between patients with a VUS in any gene, compared with patients with a negative result. Thus, incorporation of pre-and post-test genetic counseling as part of the multidisciplinary evaluation for young women newly diagnosed with BC could help ensure appropriate testing and accurate risk assessment.
Strengths of this study included systematic provision of genetic counseling to all patients. Additionally, all patients were seen by a genetic specialist. Limitations of this study included a small sample size comprised primarily of non-Hispanic White patients at one institution, which limits the ability to generalize the findings. Additional studies are needed to understand the impact of GT on surgical decision-making in more diverse populations. While the type of GT offered was not standardized, the majority (86.4%) were tested via a multigene panel. This exploratory study focused on the utility of genetic counseling and GT at the time of diagnosis for surgical decision-making among young patients with BC; as such, other factors that could potentially influence decision-making regarding CRRM, such as socioeconomic status, reproductive planning, psychosocial factors, or concerns regarding potential cosmetic outcomes, were not explored here. In addition, the few patients seen in the multidisciplinary Breast Cancer Specialty Care Clinic who declined genetic counseling and/or testing were excluded from the study population, thus, we were not able to evaluate whether patients with breast cancer diagnosed at age ≤ 45 years who opted not to have genetic counseling and testing would select mastectomy with CRRM a similar or different rate compared to women included in our study.

Conclusions
For young women newly diagnosed with BC seen in a multidisciplinary clinic, the identification of a BC genetic predisposition was associated with the decision to pursue CRRM at the time of surgical BC treatment. Young patients with uninformative results had lower rates of CRRM, even among those with a significant family history. Thus, early incorporation of genetic services in the treatment planning process for young patients newly diagnosed with BC has the potential to lead to more appropriate risk assessment, which in turn could promote more risk-appropriate management.
Our multidisciplinary Breast Cancer Specialty Care Clinic is designed specifically to serve young women with breast cancer. As the utility of timely genetic testing has been demonstrated, it is critical to provide timely GT to all BC patients for whom testing is indicated. Given the shortage of cancer genetics providers and limited access to cancer genetic services, other approaches to genetic services such as oncology provider-led genetic testing and telegenetic services have been employed with promising outcomes. Additional works are needed to solidify a process whereby timely GT is available to all appropriate patients. Furthermore, educational efforts to identify and refer individuals without a cancer diagnosis who might benefit from GT based on family history are needed, as the identification of a hereditary cancer predisposition before a cancer diagnosis occurs can lead to proactive implementation of risk management interventions, which could in turn result in improved quality of life and reduced cancer-related morbidity and mortality.
Author contributions SAD collected, analyzed, and interpreted the data, and prepared the manuscript; DLT interpreted the data and participated in manuscript preparation; PFM, JGS, and RRJ contributed to clinical data collection, and reviewed the manuscript; EJD conceived and designed the study, provided clinical data, critically reviewed the data, and reviewed the manuscript, PLM conceived and designed the study, analyzed and interpreted the data, and prepared the manuscript. All authors approved the final manuscript.
Funding This research project was supported in part by funding from the Hackers for Hope Pittsburgh, a 501c(3) charity.

Data availability
The datasets generated during and/or analyzed during the current study are not publicly available to protect patient privacy; however, a de-identified dataset is available from author PLM upon reasonable request.

Conflict of interest
The authors have no relevant financial or non-financial interests to disclose.
Ethical approval This study was performed in line with the principles of the Declaration of Helsinki. The project was reviewed and approved by the University of Pittsburgh Medical Center Quality Improvement Review Committee and was deemed exempt from approval by the Institutional Review Board.

Consent to participate
This was a quality control study. The University of Pittsburgh Medical Center Quality Improvement Review Committee and Institutional Review Board has confirmed that informed consent was not required.

Consent to publish
Not applicable; no identifying information is included in this report or associated data, figures, or tables.