Key Points
Question
Is prophylactic salpingo-oophorectomy (PSO) for patients with surgically resected, pathogenic germline BRCA1 or BRCA2 breast cancer associated with improved overall survival?
Findings
This cohort study of 480 consecutive patients with surgically treated BRCA1/2 breast cancer between 1972 and 2019 showed that PSO was significantly associated with improved overall survival compared with no PSO.
Meaning
These findings suggest that PSO should be offered to patients with BRCA1/2 breast cancer who undergo surgery, particularly those with the BRCA1 variant, to reduce the risk of death.
This cohort study assesses the prognostic outcomes of prophylactic salpingo-oophorectomy after surgical resection of breast cancer for patients carrying the BRCA1 or BRCA2 variant.
Abstract
Importance
Few studies have investigated whether prophylactic salpingo-oophorectomy (PSO) for patients with previously resected breast cancer who carry pathogenic germline BRCA1 or BRCA2 variants is associated with a reduced risk of cancer-specific death.
Objective
To assess the association of PSO and prophylactic mastectomy (PM) with prognosis after quadrantectomy or mastectomy as primary treatment for patients with BRCA1 or BRCA2 breast cancer.
Design, Setting, and Participants
This retrospective cohort study was performed in a single-institution, tertiary referral center. Consecutive patients with invasive breast cancer treated surgically between 1972 and 2019 were recruited and followed up prospectively after they were found to carry the BRCA1 or BRCA2 gene variant. The data analysis was performed between April 2022 and July 2023.
Exposure
Following breast surgery, some patients underwent PSO, PM, or both, whereas others did not.
Main Outcomes and Measures
The primary study end point was overall survival as measured by the Kaplan-Meier method. Secondary end points were crude cumulative incidence of breast cancer–specific mortality, ipsilateral breast tumor recurrence (IBTR), contralateral breast cancer, ovarian cancer, and ovarian cancer–specific mortality.
Results
Of 480 patients included in the cohort (median age at initial surgery, 40.0 years; IQR, 34.0-46.0 years), PSO was associated with a significantly reduced risk of death (hazard ratio [HR], 0.40; 95% CI, 0.25-0.64; P < .001). This reduction was most evident for patients carrying the BRCA1 variant (HR, 0.35; 95% CI, 0.20-0.63; P = .001), those with triple-negative disease (HR, 0.21; 95% CI, 0.09-0.46; P = .002), and those with invasive ductal carcinoma (HR, 0.51; 95% CI, 0.31-0.84; P = .008). Prophylactic salpingo-oophorectomy was not associated with risk of contralateral breast cancer or IBTR. Initial or delayed PM was associated with a reduced risk of IBTR but not with overall survival or breast cancer–specific mortality.
Conclusions
The study findings suggest that PSO should be offered to all patients with BRCA1/2 breast cancer who undergo surgery with curative intent to reduce risk of death. In particular, PSO should be offered to patients with the BRCA1 variant at the time of breast surgery.
Introduction
Breast cancer is the leading cause of cancer death among women.1 Approximately 8% of breast cancer cases are associated with pathogenic germline variants of the BRCA1 or BRCA2 genes. Women with the BRCA1 variant have lifetime risks of breast or ovarian cancer of 45% to 80% and 30% to 60%, respectively. Women with the BRCA2 variant have lifetime risks of breast and ovarian cancer of 35% to 60% and 10% to 25%, respectively.2,3 BRCA1 breast cancer is more often high grade and triple negative than sporadic disease, while BRCA2 breast cancer is often similar to sporadic, luminal-type disease that is positive for hormone receptors and does not express human epidermal growth factor receptor 2 (ERBB2 [formerly HER2 or HER2/neu]).4,5,6 Patients with BRCA1/2 breast cancer are also at greater risk of contralateral breast cancer and ipsilateral breast tumor recurrence (IBTR) than those with sporadic breast cancer,7,8,9,10 so treatment has shifted back to mastectomy.11,12,13
Few studies have investigated the outcomes of breast-conserving surgery, prophylactic mastectomy (PM), or prophylactic salpingo-oophorectomy (PSO) on outcomes in patients with BRCA1/2 breast cancer.14,15 Therefore, we conducted this cohort study to assess the prognostic outcomes of quadrantectomy vs mastectomy, PM vs no PM, and PSO vs no PSO.
Methods
Patient Population and Eligibility
This cohort study comprised 480 consecutive patients with breast cancer with BRCA1 or BRCA2 variants who were followed up prospectively from genetic testing after undergoing surgery for breast cancer between 1972 and 2019 at the Italian National Cancer Institute, Milan. The median age at initial surgery was 40 years (range, 19-77 years), and 290 patients (60.4%) had the BRCA1 variant, and 190 (39.6%) the BRCA2 variant. Patients with synchronous bilateral breast cancer, distant metastases at diagnosis, previous cancer at another site (except nonmelanoma skin cancer), or previous PSO were excluded. Tests for BRCA1/2 variants were performed for patients with a family history of breast cancer or ovarian cancer or with a young age at breast cancer diagnosis. Methods are described in more detail in the eMethods in Supplement 1.
The study was approved by the ethics committee of the Italian National Cancer Institute, Milan. All patients provided written informed consent. The study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline for cohort studies.
Treatments and Follow-Up
Patients underwent either quadrantectomy or mastectomy as the primary surgery. All patients were advised to undergo both PSO and PM. Disease status or cause of death were ascertained from clinical records or by contacting the general practitioners of patients no longer receiving follow-up.
Statistical Analysis
The data analysis was performed between April 2022 and July 2023. The primary study end point was overall survival (OS). Secondary end points were incidence of breast cancer–specific mortality, ovarian cancer–specific mortality, IBTR (as first event and at any time), contralateral breast cancer (at any time), and ovarian cancer (at any time). Kaplan-Meier curves were used to estimate OS and were compared using log-rank test. Crude cumulative incidence (CCI) curves for breast cancer–specific mortality, ovarian cancer–specific mortality, IBTR, contralateral breast cancer, and ovarian cancer were estimated as competing risks and compared using the Gray test. The OS and CCI estimates are reported as percentages with 95% CIs. Median times to IBTR, contralateral breast cancer, and ovarian cancer (calculated as time corresponding to one-half the value of the CCI curves at the latest event time) are also reported. The Kaplan-Meier and CCI curves were compared according to variant status and initial surgery.
Univariable and multivariable Cox proportional hazards regression models for OS and the Fine-Gray model for breast cancer–specific mortality, ovarian cancer–specific mortality, IBTR, contralateral breast cancer, and ovarian cancer were run to assess the association of variant status, PSO, PM, and initial surgery with these outcomes. For all end points in time-to-event models, time was censored at latest follow-up for patients alive and free of any event. The median follow-up was estimated by the reverse Kaplan-Meier method from OS data.16 The statistical analyses were performed using SAS, version 9.4 (SAS Institute Inc) and R, version 4.1.2 (R Foundation for Statistical Computing) software. A statistical test was considered significant at 2-sided P < .05.
Results
Patient characteristics are summarized in Table 1 and eTable 1 in Supplement 1. Of the 480 patients included in the cohort (median age at initial surgery, 40.0 years [IQR, 34.0-46.0 years]), 272 underwent quadrantectomy and 208 underwent mastectomy as the primary surgery. Although all patients were advised to undergo both PSO and PM, not all chose to do so, with 300 (62.5%; 186 with BRCA1 and 114 with BRCA2 variant) receiving PSO and 163 (66.0%; 97 with BRCA1 and 66 with BRCA2 variant) receiving PM (bilateral if primary treatment was quadrantectomy). The median follow-up was 198 months (IQR, 125-307 months), and no patients were lost to follow-up. The flowchart in eFigure 1 in Supplement 1 shows patients according to variant status, initial surgery, subsequent surgery, and deaths.
Table 1. Characteristics of Consecutive Patients With Breast Cancer With Pathogenic Germline BRCA1 or BRCA2 Variants.
| Characteristic | Overall (n = 480) | BRCA1 (n = 290) | BRCA2 (n = 190) | SMD |
|---|---|---|---|---|
| Age at initial surgery, median (IQR), y | 40.0 (34.0-46.0) | 40.0 (34.0-46.0) | 40.0 (33.0-47.0) | 0.017 |
| Age at ovarian cancer surgery, median (IQR), y | 56.0 (50.0-63.0) | 55.0 (47.3-61.8) | 61.0 (54.5-66.0) | 0.652 |
| Age at prophylactic oophorectomy, median (IQR), y | 46.6 (42.1-54.2) | 46.7 (42.1-53.9) | 46.5 (42.0-54.5) | 0.074 |
| Prophylactic oophorectomy, No. (%) | ||||
| Yes | 300 (62.5) | 186 (64.1) | 114 (60.0) | 0.085 |
| No | 180 (37.5) | 104 (35.9) | 76 (40.0) | |
| Cancer at prophylactic oophorectomy, No. (%) | ||||
| Yes | 6 (2.0) | 5 (2.7) | 1 (0.9) | 0.137 |
| No | 294 (98.0) | 181 (97.3) | 113 (99.1) | |
| Prophylactic mastectomy, No. (%) | ||||
| No | 317 (66.0) | 193 (66.6) | 124 (65.3) | 0.027 |
| Yes | 163 (34.0) | 97 (33.4) | 66 (34.7) | |
| Initial surgery, No. (%) | ||||
| Mastectomy | 208 (43.3) | 104 (35.9) | 104 (54.7) | 0.386 |
| Quadrantectomy | 272 (56.7) | 186 (64.1) | 86 (45.3) | |
| Stage, No. (%) | ||||
| T1 | 275 (57.3) | 164 (56.6) | 111 (58.4) | 0.232 |
| T2 | 191 (39.8) | 120 (41.4) | 71 (37.4) | |
| T3 | 7 (1.5) | 5 (1.7) | 2 (1.1) | |
| T4b | 7 (1.5) | 1 (0.3) | 6 (3.2) | |
| Breast tumor size, median (IQR), mm | 17.0 (10.0-24.0) | 18.0 (12.0-24.0) | 15.5 (10.0-21.8) | 0.014 |
| Histology, No. (%) | ||||
| Invasive ductal carcinoma | 390 (81.3) | 233 (80.3) | 157 (82.6) | 0.465 |
| Invasive lobular carcinoma | 29 (6.0) | 12 (4.1) | 17 (8.9) | |
| Medullary carcinoma | 33 (6.9) | 31 (10.7) | 2 (1.1) | |
| Other invasive histology | 28 (5.8) | 14 (4.8) | 14 (7.4) | |
| Grade, No. (%) | ||||
| 1 | 5 (1.0) | 2 (0.7) | 3 (1.6) | 0.552 |
| 2 | 154 (32.1) | 65 (22.4) | 89 (46.8) | |
| 3 | 293 (61.0) | 202 (69.7) | 91 (47.9) | |
| Not available | 28 (5.8) | 21 (7.2) | 7 (3.7) | |
| Receptor status, No. (%) | ||||
| ER negative, PgR negative | 225 (46.9) | 185 (63.8) | 40 (21.1) | 1.085 |
| ER negative, PgR positive | 15 (3.1) | 10 (3.4) | 5 (2.6) | |
| ER positive, PgR negative | 26 (5.4) | 10 (3.4) | 16 (8.4) | |
| ER positive, PgR positive | 186 (38.8) | 64 (22.1) | 122 (64.2) | |
| Not available | 28 (5.8) | 21 (7.2) | 7 (3.7) | |
| ERBB2 status, No. (%) | ||||
| Positive | 25 (5.2) | 7 (2.4) | 18 (9.5) | 0.311 |
| Negative | 409 (85.2) | 252 (86.9) | 157 (82.6) | |
| Not available | 46 (9.6) | 31 (10.7) | 15 (7.9) | |
| Ki-67 status, No. (%) | ||||
| ≤20% | 103 (21.5) | 45 (15.5) | 58 (30.5) | 0.363 |
| >20% | 260 (54.2) | 168 (57.9) | 92 (48.4) | |
| Not available | 117 (24.4) | 77 (26.6) | 40 (21.1) | |
| pN, No. (%) | ||||
| Negative | 291 (62.3) | 196 (69.5) | 95 (51.4) | 0.378 |
| Positive | 176 (37.7) | 86 (30.5) | 90 (48.6) | |
| Surrogate molecular subtype, No. (%) | ||||
| ERBB2 positive | 9 (1.9) | 4 (1.4) | 5 (2.6) | 1.083 |
| Luminal A | 65 (13.5) | 24 (8.3) | 41 (21.6) | |
| Luminal B ERBB2 negative | 133 (27.7) | 51 (17.6) | 82 (43.2) | |
| Luminal B ERBB2 positive | 16 (3.3) | 3 (1.0) | 13 (6.8) | |
| Triple negative | 211 (44.0) | 177 (61.0) | 34 (17.9) | |
| Not available | 46 (9.6) | 31 (10.7) | 15 (7.9) |
Abbreviations: ER, estrogen receptor; ERBB2 (formerly HER2 or HER2/neu), human epidermal growth factor receptor 2; PgR, progesterone receptor; SMD, standardized mean difference.
Outcomes
eTable 2 in Supplement 1 shows the time to prophylactic surgery, the frequency and time to event, and the number of deaths and their causes. Of the 290 patients with the BRCA1 variant, 49 (16.9%) developed IBTR, 83 (28.6%) had contralateral breast cancer, and 77 (26.6%) had 80 second primary tumors (including 58 ovarian cancers).
Twenty-eight patients with the BRCA1 variant (9.7%) died of breast cancer, 32 (11.0%) died of ovarian cancer, 10 (3.4%) died of a non–ovarian cancer second primary cancer, and 1 (0.3%) died of a noncancer cause; 215 patients (74.1%) were alive and disease free at latest follow-up. The 25-year CCI rates of IBTR, contralateral breast cancer, and ovarian cancer were 25.6% (95% CI, 19.6%-33.5%), 28.7% (95% CI, 23.7%-34.8%), and 21.0% (95% CI, 16.6%-26.5%), respectively. Median times from breast surgery to IBTR, contralateral breast cancer, and ovarian cancer were 176.1 months (IQR, 103.2-248.9 months), 156.2 months (IQR, 74.3-362.8 months), and 111.3 months (IQR, 47.0-184.3 months), respectively (eTable 2 in Supplement 1).
Of the 190 patients with the BRCA2 variant, 29 (15.3%) developed IBTR, 46 (24.2%) had contralateral breast cancer, and 32 (16.4%) had 35 second primary tumors (including 19 ovarian cancers). Twenty-three patients (12.1%) died of breast cancer, 8 (4.2%) died of ovarian cancer, 3 (1.6%) died of a second primary cancer (not ovarian cancer), and 3 (1.6%) died of noncancer causes; 145 (76.3%) were alive and disease free at the latest follow-up. The 25-year CCI rates of IBTR, contralateral breast cancer, and ovarian cancer were 23.0% (95% CI, 15.3%-34.7%), 28.9% (95% CI, 22.4%-37.4%) and 11.2% (95% CI, 7.1%-17.6%), respectively. Median times from breast surgery to IBTR, contralateral breast cancer, and ovarian cancer were 258.6 months (IQR, 180.3-343.1 months), 135.1 months (IQR, 80.5-193.2 months), and 150.2 months (33.1-279.0 months), respectively (eTable 2 in Supplement 1).
The CCI curves for events and the Kaplan-Meier curves for OS according to variant status are shown in Figure 1. Events were similar for BRCA1 and BRCA2, except that patients with BRCA1 had a significantly higher CCI rate of ovarian cancer (25-year CCI, 21.1%; 95% CI, 16.6%-26.5%) than patients with BRCA2 (25-year CCI, 11.2%; 95% CI, 7.1%-17.6%; P = .03).
Figure 1. Crude Cumulative Incidence (CCI) for Breast Cancer and Ovarian Cancer Events and Overall Survival According to BRCA1 and BRCA2 Variant Status.
The CCI curves for IBTR and breast cancer–specific mortality and the Kaplan Meier curves for OS according to initial surgery are shown in Figure 2. The CCI rate of IBTR was significantly lower (25-year CCI as first event, 1.5% [95% CI, 0.1%-4.3%]; 25-year CCI as any event, 2.5% [95% CI, 0.1%-6.8%]) for patients who underwent mastectomy than for those who underwent quadrantectomy (25-year CCI as first event, 32.9% [95% CI, 24.5%-44.1%]; 25-year CCI as any event, 45.9% [95% CI, 36.3%-58.0%]; P < .001).
Figure 2. Crude Cumulative Incidence (CCI) for Ipsilateral Breast Tumor Recurrence (IBTR) and Breast Cancer–Specific Mortality and Overall Survival According to Mastectomy and Quadrantectomy as Primary Surgery.
The results of multivariable modeling to estimate whether interactions of BRCA1 or BRCA2 with histology and surrogate molecular subtype were associated with events are shown in Table 2. Due to the low numbers of events in subgroups, the subdistribution hazard ratios (sHRs) and the HR estimates have wide 95% CIs, which may be associated with the finding that interactions were not significant. Risk of IBTR was lower (sHR, 0.39; 95% CI, 0.17-0.90) for patients with BRCA2 vs BRCA1 luminal B ERBB2-negative disease. Conversely, risk of breast cancer–specific mortality was higher (sHR, 2.85; 95% CI, 1.24-6.56) for patients with the BRCA2 vs BRCA1 variant and triple-negative disease.
Table 2. Multivariable Models Estimating the Prognostic Outcomes of BRCA2 vs BRCA1 Variants, in Combination With Histology or Surrogate Molecular Subtype, Regarding IBTR, Contralateral Breast Cancer, Ovarian Cancer, Breast Cancer–Specific Mortality, and Overall Survival.
| Model | IBTR, any event | Contralateral breast cancer, any event | Ovarian cancer | Breast cancer–specific mortality | Overall survival | |||||
|---|---|---|---|---|---|---|---|---|---|---|
| sHR (95% CI) | P valuea | sHR (95% CI) | P valuea | sHR (95% CI) | P valuea | sHR (95% CI) | P valuea | HR (95% CI) | P valuea | |
| Variant and histologyb | ||||||||||
| BRCA2 vs BRCA1 with ILC | 1.20 (0.22-6.60) | .65 | 0.71 (0.23-2.21) | .50 | 0.74 (0.13-4.03) | .13 | 2.07 (0.23-18.88) | .78 | 0.93 (0.26-3.30) | .70 |
| BRCA2 vs BRCA1 with IDC | 0.76 (0.46-1.26) | 0.77 (0.52-1.13) | 0.49 (0.26-0.90) | 1.09 (0.60-1.98) | 0.82 (0.52-1.30) | |||||
| BRCA2 vs BRCA1 with medullary carcinoma | 3.10 (0.32-29.93) | Unable to estimate riskc | Unable to estimate riskc | Unable to estimate riskc | Unable to estimate riskc | |||||
| BRCA2 vs BRCA1 with other invasive histology | 0.50 (0.05-5.52) | 1.47 (0.28-7.81) | 1.45 (0.13-16.39) | Unable to estimate riskc | Unable to estimate riskc | |||||
| Interaction of variant with histology | NA | .62 | NA | .74 | NA | .63 | NA | .58 | NA | .86 |
| Variant and surrogate molecular subtypeb,d | ||||||||||
| BRCA2 vs BRCA1 with triple negative | 1.51 (0.65-3.48) | .13 | 0.53 (0.24-1.19) | .55 | 0.52 (0.19-1.40) | .23 | 2.85 (1.24-6.56) | .10 | 1.20 (0.61-2.38) | .88 |
| BRCA2 vs BRCA1 with luminal A | 1.51 (0.46-4.93) | 0.76 (0.29-2.00) | 0.23 (0.05-1.16) | 1.56 (0.23-10.82) | 1.99 (0.23-17.58) | |||||
| BRCA2 vs BRCA1 with luminal B ERBB2 positive | 0.14 (0.01-2.31) | 1.20 (0.61-2.33) | Unable to estimate riskc | Unable to estimate riskc | 0.47 (0.04-5.08) | |||||
| BRCA2 vs BRCA1 with luminal B ERBB2 negative | 0.39 (0.17-0.90) | Unable to estimate riskc | 0.62 (0.25-1.55) | 0.56 (0.20-1.57) | 0.72 (0.35-1.50) | |||||
| BRCA2 vs BRCA1 with not available | 0.60 (0.12-3.13) | 0.90 (0.38-2.14) | 1.32 (0.16-11.22) | 1.80 (0.30-10.78) | 1.01 (0.29-3.49) | |||||
| Interaction of variant with surrogate molecular subtype | NA | .10 | NA | .50 | NA | .61 | NA | .13 | NA | .78 |
Abbreviations: IBTR, ipsilateral breast tumor recurrence; ERBB1 (formerly HER2 or HER2/neu), human epidermal growth factor receptor 2; HR, hazard ratio; IDC, invasive ductal carcinoma; ILC, invasive lobular carcinoma; NA, not applicable; sHR, subdistribution hazard ratio.
From Wald test.
Tests of the null hypothesis that the variant coefficient and the coefficients for interaction between variant and histology or surrogate molecular subtype are 0.
Too few cases or events to estimate risks.
The ERBB2 subtype was omitted because it did not converge in any of the models due to low numbers of cases or events.
eTable 3 in Supplement 1 shows the results of multivariable models exploring associations of initial surgery with IBTR, breast cancer–specific mortality, and OS according to variant type, histology, and surrogate molecular classification. Mastectomy was significantly associated with reduced risk of IBTR (first event) only for patients with the BRCA1 variant (sHR, 0.04; 95% CI, 0.00-0.72) and with worse breast cancer–specific mortality (BRCA1: sHR, 4.95 [95% CI, 0.98-25.08]; BRCA2: sHR, 6.17 [95% CI, 1.51-25.24]) and OS (BRCA1: HR, 6.78 [95% CI, 1.61-28.52]; BRCA2: HR, 4.70 [95% CI, 1.19-18.61]).
Association of PSO With Outcomes
Of the 480 patients, 300 (62.5%) received PSO a median of 51.2 months (IQR, 17.8-134.6 months) after breast cancer surgery. The median age at PSO was 46.6 years (range, 27.5-79.2 years) (Table 1; eTable 2 in Supplement 1). For 6 patients (2.0%) (5 with the BRCA1 variant and 1 with the BRCA2 variant), ovarian cancer was found at PSO (serous tubal intraepithelial carcinoma in 4 patients, undifferentiated carcinoma in 1 patient, and G3 ovarian cystadenocarcinoma in 1 patient); all 6 patients were alive and disease free at the latest follow-up. Of the patients receiving PSO, 16 (5.3%) (9 with the BRCA1 variant and 7 with the BRCA2 variant) died of breast cancer, 7 (2.3%) died of a second primary cancer, and 1 (0.3%) died of a noncancer cause. A total of 276 patients receiving PSO (92%) were alive at the latest follow-up, 268 of whom were disease free. The median time to death after initial surgery was 385.1 months (IQR, 248.3-448.4 months), and the corresponding OS rate was 82.7% (95% CI, 65.4%-91.0%).
By univariable analysis (Table 3), PSO was strongly associated with better OS (HR, 0.40; 95% CI, 0.25-0.64; P < .001) but not with contralateral breast cancer or breast cancer–specific mortality. By multivariable analysis, with adjustment for age at initial surgery, tumor size, year of surgery, grade, initial surgery, and adjuvant treatment, there was a significant interaction only between PSO and the surrogate molecular subtype for OS (P = .02). Undergoing PSO was associated with better OS for patients with the BRCA1 variant (HR, 0.35; 95% CI, 0.20-0.63) but for not patients with the BRCA2 variant (HR, 0.60; 95% CI, 0.28-1.27). The OS advantage of PSO was highly significant for triple-negative disease (HR, 0.21; 95% CI, 0.09-0.46) and invasive ductal carcinoma (IDC) (HR, 0.51; 95% CI, 0.31-0.84), but not for luminal B ERBB2-negative disease (HR, 0.48; 95% CI, 0.20-1.19). Prophylactic oophorectomy was also significantly associated with higher contralateral breast cancer risk for patients with triple-negative disease (sHR, 1.73; 95% CI, 1.01-2.99) but lower contralateral breast cancer risk for invasive lobular carcinoma (sHR, 0.10; 95% CI, 0.01-0.78) and for patients with luminal B ERBB2-negative disease (sHR, 0.34; 95% CI, 0.12-0.96).
Table 3. Univariable and Multivariable Models Estimating Prognostic Outcomes of Prophylactic Salpingo-Oophorectomy Regarding Contralateral Breast Cancer, Breast Cancer–Specific Mortality, and Overall Survival by Variant Type, Histology, and Surrogate Molecular Subtype.
| Model | Contralateral breast cancer, any event | Breast cancer–specific mortality | Overall survival | |||
|---|---|---|---|---|---|---|
| sHR (95% CI) | P valuea | sHR (95% CI) | P valuea | HR (95% CI) | P valuea | |
| Univariable modelb | ||||||
| Yes vs no | 1.04 (0.70-1.53) | .86 | 0.87 (0.47-1.62) | .67 | 0.40 (0.25-0.64) | <.001 |
| Multivariable model without interactionb | ||||||
| Yes vs no | 1.03 (0.69-1.52) | .90 | 0.96 (0.52-1.80) | .91 | 0.43 (0.27-0.67) | .001 |
| Prophylactic oophorectomy and variant typeb | ||||||
| Yes vs no, with BRCA1 | 1.12 (0.69-1.82) | .81 | 0.93 (0.41-2.13) | .98 | 0.35 (0.20-0.63) | .001 |
| Yes vs no, with BRCA2 | 0.86 (0.44-1.70) | 1.04 (0.43-2.47) | 0.60 (0.28-1.27) | |||
| Interaction of oophorectomy with variant type | NA | .53 | NA | .86 | NA | .27 |
| Prophylactic oophorectomy and histologyb | ||||||
| Yes vs no for ILC | 0.10 (0.01-0.78) | .10 | 0.70 (0.11-4.30) | .60 | 0.26 (0.05-1.23) | .008 |
| Yes vs no for IDC | 1.14 (0.74-1.77) | 0.90 (0.46-1.79) | 0.51 (0.31-0.84) | |||
| Yes vs no for medullary carcinoma | 2.63 (0.80-8.63) | 6.07 (0.37-99.12) | 0.21 (0.03-1.63) | |||
| Yes vs no, with other invasive histology | 1.19 (0.22-6.47) | Unable to estimate riskc | Unable to estimate riskc | |||
| Interaction of oophorectomy with histology | NA | .06 | NA | .40 | NA | .54 |
| Prophylactic oophorectomy and surrogate molecular subtypeb,d | ||||||
| Yes vs no for triple negative | 1.73 (1.01-2.99) | .09 | 0.70 (0.24-2.01) | .59 | 0.21 (0.09-0.46) | .002 |
| Yes vs no for luminal A | 0.94 (0.34-2.60) | 2.30 (0.40-13.36) | 1.34 (0.22-8.02) | |||
| Yes vs no for luminal B ERBB2 positive | Unable to estimate riskc | 3.88 (0.44-34.16) | 4.04 (0.37-44.62) | |||
| Yes vs no for luminal B ERBB2 negative | 0.34 (0.12-0.96) | 0.69 (0.21-2.23) | 0.48 (0.20-1.19) | |||
| Yes vs no for not available | 0.94 (0.35-2.56) | 1.90 (0.30-11.84) | 1.31 (0.40-4.31) | |||
| Interaction of oophorectomy with surrogate molecular subtype | NA | .050 | NA | .45 | NA | .02 |
Abbreviations: ERBB2 (formerly HER2 or HER2/neu), human epidermal growth factor receptor 2; HR, hazard ratio; IDC invasive ductal carcinoma; ILC, invasive lobular carcinoma; NA, not applicable; sHR, subdistribution hazard ratio.
From Wald test.
Tests of the null hypotheses that the prophylactic salpingo-oophorectomy coefficient, and coefficients for interactions between prophylactic salpingo-oophorectomy and variant status, histology, or surrogate molecular subtype, are 0.
Too few cases or events to estimate risks.
The ERBB2 subtype was omitted because it did not converge in any of the models due to low number of cases or events.
For patients younger than 50 years, PSO was also associated with a lower incidence of contralateral breast cancer for patients with luminal B ERBB2-negative disease (HR, 0.33; 95% CI, 0.10-0.97) and with better OS for those with triple-negative disease (HR, 0.27; 95% CI, 0.12-0.62) but not for patients aged 50 years or older, perhaps because of the small size of this subset. Prophylactic oophorectomy was not associated with lower IBTR incidence by either univariable or multivariable analysis, according to initial surgery (HR, 0.83 [95% CI, 0.51-1.34] and 0.98 [95% CI, 0.60-1.61], respectively). Furthermore, among patients receiving PSO, neither breast cancer–specific mortality nor OS differed between the variant groups (eFigure 2 in Supplement 1).
Of the 180 patients not receiving PSO, 77 (42.8%) (58 with the BRCA1 variant and 19 with the BRCA2 variant) developed ovarian cancer, and 40 (22.2%) died of this disease. In addition, 35 (19.4%) died of breast cancer, 6 (3.3%) died of a second primary cancer, and 3 (1.7%) died of noncancer causes. In the no-PSO group, the median age at ovarian cancer diagnosis was 55 years (range, 35-84 years) for those with the BRCA1 variant and 61 years (range, 50-78 years) for those with the BRCA2 variant. In the no-PSO group, the median OS was 273.4 months (IQR, 124.0-442.2 months), the median time to death after initial surgery was 205.9 months (IQR, 115.5-394.3 months), and the OS probability was 55.3% (95% CI, 47.4%-64.3%).
At the latest follow-up, 96 patients (53.3%) in the no-PSO group were alive, 92 of whom were disease free. Nine patients with the BRCA1 variant (8.7%) in the no-PSO group developed ovarian cancer at a young age (≤42 years), 5 of whom died of ovarian cancer, 1 died of breast cancer, 3 were alive at latest follow-up (2 of whom were disease free and 1 had peritoneal metastases from ovarian cancer). None of the patients with the BRCA2 variant in the no-PSO group developed ovarian cancer at a young age.
In the no-PSO group, patients with the BRCA1 variant had higher CCI rates for ovarian cancer and ovarian cancer–specific mortality and worse OS than patients with the BRCA2 variant, while breast cancer–specific mortality did not differ between the variants (eFigure 2 in Supplement 1). Of the 16 patients with medullary carcinoma (14 with the BRCA1 variant and 2 with the BRCA2 variant) in the no-PSO group, 15 (93.8%) developed ovarian cancer, and 11 (68.8%) died of this disease.
Association of PM With Outcomes
A total of 163 patients (34.0%) received PM (including 66 with bilateral mastectomy after quadrantectomy). For 13 patients (8.0%), disease was found (IDC ≤4 mm in 4 patients, ductal carcinoma in situ <3 mm in 7 patients, and lobular carcinoma in situ in 2 patients). The median time to PM was 11.2 months (IQR, 0.0-45.3 months) (eTable 2 in Supplement 1). Prophylactic mastectomy (time-dependent variable in univariable and multivariable models) was significantly associated with worse breast cancer–specific mortality (poor prognosis) for patients with triple-negative disease (sHR, 3.15; 95% CI, 1.22-8.15); of the 133 patients with luminal B ERBB2-negative disease (better prognosis), 0 of the 45 who received PM died, and 14 of 88 (16%) patients with luminal B ERBB2-negative disease who did not receive PM died. Prophylactic mastectomy was not significantly associated with OS in any subgroup (eTable 4 in Supplement 1).
Discussion
The most important finding in this cohort study was that risk of any-cause death after breast cancer surgery for patients with BRCA1/2 variants who also received PSO was less than one-half (HR, 0.40; 95% CI, 0.25-0.64; P < .001) (Table 3) that of patients who did not receive PSO. Furthermore, although the interaction between variant status and PSO relative to OS was not significant (P = .27) (Table 3), the reduction in risk of death associated with PSO was greater for patients with the BRCA1 variant (HR, 0.35; 95% CI, 0.20-0.63) than for patients with the BRCA2 variant (HR, 0.60; 95% CI, 0.28-1.27).
The 180 patients in the no-PSO group were more likely to die of ovarian cancer than of breast cancer, with 77 (42.8%) developing ovarian cancer during follow-up and 40 (22.2%) having died of this disease, while 35 (19.4%) died of breast cancer. Furthermore, outcomes in the no-PSO group were worse for those carrying BRCA1 vs BRCA2, with higher ovarian cancer incidence, higher ovarian cancer–specific mortality, and lower OS (eFigure 2C-E in Supplement 1). However, breast cancer–specific mortality did not differ between the variants (eFigure 2F in Supplement 1), even though 61% of patients with the BRCA1 variant had triple-negative disease. Thus, patients with the BRCA1 variant who do not undergo PSO may die of ovarian cancer rather than of their poor-prognosis breast cancer.
The relatively high ovarian cancer–specific mortality for patients who did not undergo PSO correlates with early diagnosis of ovarian cancer being problematic, even with periodic transvaginal ultrasonography and cancer antigen 125 test. Thus, although various screening strategies can detect ovarian cancer relatively early,17,18,19 they fail to reduce ovarian cancer–specific mortality.17,20 Somewhat paradoxically, all 6 of the patients with incidentally detected OC at PSO were alive and disease free at the latest follow-up, suggesting that clinically occult ovarian cancer can be cured by surgery. Finch et al21 also reported that PSO was associated with an 80% reduction in mortality risk due to ovarian, fallopian tube, or peritoneal cancer and a 77% reduction in all-cause mortality for patients carrying the BRCA1 or BRCA2 variant.
We found that fewer (5.3%) patients undergoing PSO died of breast cancer than those who did not (19.4%). Nevertheless, PSO was not significantly associated with breast cancer–specific mortality by either univariable or multivariable analysis with interaction with variant status (Table 3). Reasons for this finding may be that deaths were few or PSO was often delayed (median time to PSO after initial surgery, 51 months), which may have reduced any positive outcome of PSO associated with breast cancer mortality. Notably, multivariable analysis revealed a significant interaction (P = .02) between PSO and surrogate molecular subtype associated with OS (Table 3) in which PSO was strongly associated with better OS for patients with triple-negative disease (HR, 0.21; 95% CI, 0.09-0.46) and ductal carcinoma (HR, 0.51; 95% CI, 0.31-0.84), even though the interaction between PSO and histology was not significant. The associations of PSO with better OS for patients with triple-negative disease (and luminal B ERBB2-negative disease) were evident for patients younger than 50 years at surgery but not for older patients, perhaps due to the small number of older patients. Taken together, these findings suggest that PSO may have a beneficial association with breast cancer–specific mortality, at least in some subgroups. This possible benefit is supported by Metcalfe et al,15 who found that for patients with breast cancer carrying a BRCA variant, PSO was associated with a reduced risk of breast cancer–specific mortality (as well as ovarian cancer mortality), with the strongest protection observed in women with estrogen receptor (ER)–negative breast cancer. Valentini et al22 found in a case-control study of pregnancy and survival of patients with BRCA breast cancer, in which approximately one-half of the cases and controls had received bilateral PSO, PSO was associated with significantly reduced breast cancer–specific mortality compared with no PSO (adjusted HR, 0.20; 95% CI, 0.06-0.62). Similar findings were reported by Huzarski et al23 for patients with BRCA1 breast cancer in whom those receiving oophorectomy had considerably lower mortality (asserted to be mainly fewer breast cancer deaths) than those without oophorectomy (HR, 0.30; 95% CI, 0.12-0.75). These authors noted that there appeared to be no biological mechanism by which oophorectomy could influence breast cancer mortality, particularly in patients with mainly ER-negative disease. Perhaps hormone-responsive cells are present in metastases (even from ER-negative and triple-negative disease) that respond to oophorectomy by becoming quiescent.
With regard to PM, 163 patients underwent the operation (including 66 originally receiving quadrantectomy who had bilateral mastectomy). Prophylactic mastectomy (time-dependent variable in the models) was not associated with mortality or OS, either overall or by variant type, histology, or surrogate molecular subtype (eTable 4 in Supplement 1). There were exceptions, however, for patients with luminal B ERBB2-negative disease who experienced no breast cancer–specific mortality, indicating a positive outcome that could not be investigated by modeling. Prophylactic mastectomy was also associated with significantly worse breast cancer–specific mortality for patients with triple-negative disease, as observed for 3 patients who developed metastases soon after PM and died early.
Prophylactic mastectomy is often offered to patients with BRCA1/2 breast cancer to prevent a second breast cancer. However, a Cochrane review24 concluded that there is insufficient evidence that PM improves breast cancer survival. In contrast to our findings and the conclusions of the Cochrane review, a retrospective analysis of 390 patients with BRCA breast cancer14 found that the 20-year survival rate was 88% for women receiving PM vs 66% for those who did not receive PM. The operation was associated with a 48% reduction in breast cancer mortality. We suggest that PM be offered to women with good-prognosis breast cancer, who remain disease free after a long follow-up, and who fear IBTR or contralateral breast cancer (and possible chemotherapy).
As expected,25 we found that IBTR was more frequent in patients undergoing quadrantectomy than mastectomy; however, OS and breast cancer–specific mortality did not differ between these 2 groups (Figure 2). Furthermore, we found no significant association between PSO and IBTR for patients receiving quadrantectomy.
Of the 16 patients in the no-PSO group (14 with the BRCA1 variant and 2 with the BRCA2) with medullary carcinoma, 15 (93.8%) developed ovarian cancer, of whom 11 (68.8%) died of this disease and 1 died of breast cancer. The incidence of ovarian cancer was higher for patients with this histology than for those with IDC, invasive lobular carcinoma, or other invasive types. Therefore, it seems reasonable to propose that patients with a BRCA variant, particularly those with BRCA1, with medullary carcinoma be offered PSO as an initial treatment option. Only approximately 2% of sporadic breast cancers are medullary26,27 but comprised more than 10% of our BRCA1 group. A transcriptomic study27 found that most medullary carcinomas are triple negative, consistent with the finding of the present study and others15,23 that patients with triple-negative (or ER-negative) disease receiving PSO have better OS and breast cancer–specific survival.
Strengths and Limitations
The strengths of our study are its large population size and long follow-up (median, 16.5 years) over a 47-year study period in which no patient was lost to follow-up. Patients were treated and followed up at a single center, with homogeneous treatments and data collection, although treatments and tumor characterization methods changed markedly over the study period.
This study also had some limitations. Although PSOs were performed to prevent ovarian cancer (not as a treatment for breast cancer), avoiding possible bias due to differences in breast cancer characteristics influencing whether PSO was performed, an unknown number of patients with a BRCA variant who refused variant testing could not be included in the study, which may have introduced bias. Another limitation is that although the cohort was followed up prospectively after identification of a variant, patients were not randomized to treatments. The lack of randomization to quadrantectomy vs mastectomy was compensated for by use of the inverse probability treatment weighting method. Finally, despite the use of covariate adjustments in the multivariable models, unknown systematic differences between groups cannot be ruled out as potentially biasing our findings.
Conclusions
In this cohort study, we have found that PSO for patients with BRCA1 or BRCA2 breast cancer was associated with significantly better OS compared with patients not receiving PSO. We also found that PSO was associated with better OS for patients with triple-negative disease and IDC and, less strongly, with other breast cancer types, suggesting that PSO might reduce breast cancer mortality. We conclude that for patients carrying a BRCA1 or BRCA2 variant undergoing breast cancer surgery, PSO should be performed as early as possible, particularly for those with the BRCA1 variant, to reduce overall mortality.
eMethods.
eReferences.
eTable 1. Patient Clinical Characteristics According to Initial Surgery (Quadrantectomy vs Mastectomy) and Surgery During Follow-Up (Prophylactic Mastectomy, Bilateral Salpingo-Oophorectomy)
eTable 2. Times From Initial to Prophylactic Surgeries, Frequencies and Times of Events in Follow-Up, and Causes of Death, Overall and According to Variant Type
eTable 3. Results of Multivariable Models Exploring the Influence of Initial Surgery (Mastectomy vs Quadrantectomy) on Ipsilateral Breast Tumor Recurrence, Breast Cancer–Specific Mortality, and Overall Survival, According to Variant Type, Histology, and Surrogate Molecular Subtype
eTable 4. Results of Univariable and Multivariable Models Exploring Effect of Prophylactic Mastectomy on Breast Cancer–Specific Mortality and Overall Survival, According to Variant Type, Histology, and Surrogate Molecular Subtype
eFigure 1. Flow Chart for the 480 Patients With Breast Cancer With Pathogenic BRCA Variant Included in the Study, Showing Variant Type, Initial Breast Surgery, Subsequent Surgery, and Deaths
eFigure 2. Results for Patients With BRCA1 and BRCA2 Who Received Salpingo-Oophorectomy or Not Received Salpingo-Oophorectomy
Data Sharing Statement
References
- 1.DeSantis CE, Ma J, Goding Sauer A, Newman LA, Jemal A. Breast cancer statistics, 2017, racial disparity in mortality by state. CA Cancer J Clin. 2017;67(6):439-448. doi: 10.3322/caac.21412 [DOI] [PubMed] [Google Scholar]
- 2.Kuchenbaecker KB, Hopper JL, Barnes DR, et al. ; BRCA1 and BRCA2 Cohort Consortium . Risks of breast, ovarian and contralateral breast cancer for BRCA1 and BRCA2 mutation carriers. JAMA. 2017;317(23):2402-2416. doi: 10.1001/jama.2017.7112 [DOI] [PubMed] [Google Scholar]
- 3.Begg CB, Haile RW, Borg A, et al. Variation of breast cancer risk among BRCA1/2 carriers. JAMA. 2008;299(2):194-201. doi: 10.1001/jama.2007.55-a [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Gonzalez-Angulo AM, Timms KM, Liu S, et al. Incidence and outcome of BRCA mutations in unselected patients with triple receptor-negative breast cancer. Clin Cancer Res. 2011;17(5):1082-1089. doi: 10.1158/1078-0432.CCR-10-2560 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Atchley DP, Albarracin CT, Lopez A, et al. Clinical and pathologic characteristics of patients with BRCA-positive and BRCA-negative breast cancer. J Clin Oncol. 2008;26(26):4282-4288. doi: 10.1200/JCO.2008.16.6231 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Metcalfe K, Lynch HT, Foulkes WD, et al. Oestrogen receptor status and survival in women with BRCA2-associated breast cancer. Br J Cancer. 2019;120(4):398-403. doi: 10.1038/s41416-019-0376-y [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Metcalfe K, Gershman S, Lynch HT, et al. Predictors of contralateral breast cancer in BRCA1 and BRCA2 mutation carriers. Br J Cancer. 2011;104(9):1384-1392. doi: 10.1038/bjc.2011.120 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Basu NN, Ingham S, Hodson J, et al. Risk of contralateral breast cancer in BRCA1 and BRCA2 mutation carriers: a 30-year semi-prospective analysis. Fam Cancer. 2015;14(4):531-538. doi: 10.1007/s10689-015-9825-9 [DOI] [PubMed] [Google Scholar]
- 9.Graeser MK, Engel C, Rhiem K, et al. Contralateral breast cancer risk in BRCA1 and BRCA2 mutation carriers. J Clin Oncol. 2009;27(35):5887-5892. doi: 10.1200/JCO.2008.19.9430 [DOI] [PubMed] [Google Scholar]
- 10.Menes TS, Terry MB, Goldgar D, et al. Second primary breast cancer in BRCA1 and BRCA2 mutation carriers: 10-year cumulative incidence in the Breast Cancer Family Registry. Breast Cancer Res Treat. 2015;151(3):653-660. doi: 10.1007/s10549-015-3419-y [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Chiba A, Hoskin TL, Hallberg EJ, et al. Impact that timing of genetic mutation diagnosis has on surgical decision making and outcome for BRCA1/BRCA2 mutation carriers with breast cancer. Ann Surg Oncol. 2016;23(10):3232-3238. doi: 10.1245/s10434-016-5328-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.Kirova YM, Savignoni A, Sigal-Zafrani B, et al. Is the breast-conserving treatment with radiotherapy appropriate in BRCA1/2 mutation carriers? long-term results and review of the literature. Breast Cancer Res Treat. 2010;120(1):119-126. doi: 10.1007/s10549-009-0685-6 [DOI] [PubMed] [Google Scholar]
- 13.Metcalfe K, Lynch HT, Ghadirian P, et al. Risk of ipsilateral breast cancer in BRCA1 and BRCA2 mutation carriers. Breast Cancer Res Treat. 2011;127(1):287-296. doi: 10.1007/s10549-010-1336-7 [DOI] [PubMed] [Google Scholar]
- 14.Metcalfe K, Gershman S, Ghadirian P, et al. Contralateral mastectomy and survival after breast cancer in carriers of BRCA1 and BRCA2 mutations: retrospective analysis. BMJ. 2014;348:g226. doi: 10.1136/bmj.g226 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15.Metcalfe K, Lynch HT, Foulkes WD, et al. Effect of oophorectomy on survival after breast cancer in BRCA1 and BRCA2 mutation carriers. JAMA Oncol. 2015;1(3):306-313. doi: 10.1001/jamaoncol.2015.0658 [DOI] [PubMed] [Google Scholar]
- 16.Schemper M, Smith TL. A note on quantifying follow-up in studies of failure time. Control Clin Trials. 1996;17(4):343-346. doi: 10.1016/0197-2456(96)00075-X [DOI] [PubMed] [Google Scholar]
- 17.Pinsky PF, Yu K, Kramer BS, et al. Extended mortality results for ovarian cancer screening in the PLCO trial with median 15 years follow-up. Gynecol Oncol. 2016;143(2):270-275. doi: 10.1016/j.ygyno.2016.08.334 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Skates SJ, Greene MH, Buys SS, et al. Early detection of ovarian cancer using the risk of ovarian cancer algorithm with frequent CA 125 testing in women at increased familial risk—combined results from two screening trials. Clin Cancer Res. 2017;23(14):3628-3637. doi: 10.1158/1078-0432.CCR-15-2750 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19.Olivier RI, Lubsen-Brandsma MA, Verhoef S, van Beurden M. CA125 and transvaginal ultrasound monitoring in high-risk women cannot prevent the diagnosis of advanced ovarian cancer. Gynecol Oncol. 2006;100(1):20-26. doi: 10.1016/j.ygyno.2005.08.038 [DOI] [PubMed] [Google Scholar]
- 20.Menon U, Gentry-Maharaj A, Burnell M, et al. Ovarian cancer population screening and mortality after long-term follow-up in the UK Collaborative Trial of Ovarian Cancer Screening (UKCTOCS): a randomised controlled trial. Lancet. 2021;397(10290):2182-2193. doi: 10.1016/S0140-6736(21)00731-5 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21.Finch AP, Lubinski J, Møller P, et al. Impact of oophorectomy on cancer incidence and mortality in women with a BRCA1 or BRCA2 mutation. J Clin Oncol. 2014;32(15):1547-1553. doi: 10.1200/JCO.2013.53.2820 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Valentini A, Lubinski J, Byrski T, et al. ; Hereditary Breast Cancer Clinical Study Group . The impact of pregnancy on breast cancer survival in women who carry a BRCA1 or BRCA2 mutation. Breast Cancer Res Treat. 2013;142(1):177-185. doi: 10.1007/s10549-013-2729-1 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Huzarski T, Byrski T, Gronwald J, et al. Ten-year survival in patients with BRCA1-negative and BRCA1-positive breast cancer. J Clin Oncol. 2013;31(26):3191-3196. doi: 10.1200/JCO.2012.45.3571 [DOI] [PubMed] [Google Scholar]
- 24.Lostumbo L, Carbine NE, Wallace J. Prophylactic mastectomy for the prevention of breast cancer. Cochrane Database Syst Rev. 2010;11(11):CD002748. doi: 10.1002/14651858.CD002748.pub3 [DOI] [PubMed] [Google Scholar]
- 25.Pierce LJ, Phillips KA, Griffith KA, et al. Local therapy in BRCA1 and BRCA2 mutation carriers with operable breast cancer: comparison of breast conservation and mastectomy. Breast Cancer Res Treat. 2010;121(2):389-398. doi: 10.1007/s10549-010-0894-z [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Pedersen L, Zedeler K, Holck S, Schiødt T, Mouridsen HT. Medullary carcinoma of the breast. prevalence and prognostic importance of classical risk factors in breast cancer. Eur J Cancer. 1995;31A(13-14):2289-2295. doi: 10.1016/0959-8049(95)00408-4 [DOI] [PubMed] [Google Scholar]
- 27.Romero P, Benhamo V, Deniziaut G, et al. Medullary breast carcinoma, a triple negative breast cancer associated with BCLG overexpression. Am J Pathol. 2018;188(10):2378-2391. doi: 10.1016/j.ajpath.2018.06.021 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
eMethods.
eReferences.
eTable 1. Patient Clinical Characteristics According to Initial Surgery (Quadrantectomy vs Mastectomy) and Surgery During Follow-Up (Prophylactic Mastectomy, Bilateral Salpingo-Oophorectomy)
eTable 2. Times From Initial to Prophylactic Surgeries, Frequencies and Times of Events in Follow-Up, and Causes of Death, Overall and According to Variant Type
eTable 3. Results of Multivariable Models Exploring the Influence of Initial Surgery (Mastectomy vs Quadrantectomy) on Ipsilateral Breast Tumor Recurrence, Breast Cancer–Specific Mortality, and Overall Survival, According to Variant Type, Histology, and Surrogate Molecular Subtype
eTable 4. Results of Univariable and Multivariable Models Exploring Effect of Prophylactic Mastectomy on Breast Cancer–Specific Mortality and Overall Survival, According to Variant Type, Histology, and Surrogate Molecular Subtype
eFigure 1. Flow Chart for the 480 Patients With Breast Cancer With Pathogenic BRCA Variant Included in the Study, Showing Variant Type, Initial Breast Surgery, Subsequent Surgery, and Deaths
eFigure 2. Results for Patients With BRCA1 and BRCA2 Who Received Salpingo-Oophorectomy or Not Received Salpingo-Oophorectomy
Data Sharing Statement