Introduction

Contralateral prophylactic mastectomy (CPM) is an option for women undergoing mastectomy for newly diagnosed unilateral breast cancer (BC) to decrease the risk of contralateral BC development. Over the past decade, the number of women with unilateral BC electing CPM has markedly increased [1–3]. Bilateral prophylactic mastectomy (BPM) is a well-recognized risk reduction option for women at high risk of BC, such as those with known mutations in BRCA1 or BRCA2 or a strong family history of BC [4–6]. Rates of BPM are also increasing, likely because of increased awareness of familial risk and improved access to genetic counseling and genetic testing [7].

It has long been recognized that mastectomy does not remove all mammary tissue [8]. Studies have consistently shown that CPM and BPM reduce the risk of BC by 90–95 % [4, 6, 9, 10]. Therefore, prophylactic mastectomy (PM) markedly decreases but does not eliminate the possibility of a new BC on the side of the PM. Development of BC in the residual breast tissue, chest wall, or axilla on the side of a PM has been reported [11–15]. However, given the relative infrequency of its occurrence, little is known about the clinical characteristics, presentation, and management of BC following PM. The aim of the current study was to review our institutional experience with BC occurring after PM.

Methods

Study subjects

The Mayo Clinic Institutional Review Board approved this study. We have previously described our cohort of women who underwent either CPM or BPM at Mayo Clinic between January 1, 1960, and December 31, 1993 [6, 16]. BPM, either bilateral subcutaneous mastectomy or bilateral total mastectomy, was performed on 1065 women. A total of 1643 women were identified who had a unilateral therapeutic mastectomy (either radical or modified radical) and a CPM (contralateral subcutaneous mastectomy or total mastectomy performed with prophylactic intent) during the study. From 1995 through 1997, a study-specific questionnaire was sent to all women or, if they were deceased, to their next of kin. Additional follow-up surveys were sent at approximately 10 and 20 years after PM. In addition, given their diagnosis of BC, all patients who underwent CPM were also followed yearly through the Mayo Clinic Cancer Registry for subsequent BC events and outcomes. Patients were called if their questionnaire responses needed clarification. One of 2 nurses (J.L.J. and M.H.F.) reviewed medical records for procedure indications, BC risk factors, surgical complications, and cancer occurrence. All medical records of any woman thought to have a BC after PM were reviewed by the corresponding author.

Women who had a family history of BC were assigned to either a moderate-risk group or a high-risk group as described previously [6]. High-risk families have an autosomal dominant predisposition to BC and meet the following criteria: at least 1 relative with BC, young at cancer diagnosis, and a family history of ovarian cancer, bilateral BC, or BC in men [17, 18]. Women with a family history of BC who did not meet these criteria were considered to be at moderate risk.

Prophylactic surgical techniques

In subcutaneous mastectomy, most breast tissue (≥90 %) is removed except for tissue immediately beneath the nipple-areolar complex [19, 20]. In the era of this study, prophylactic total mastectomy involved resection of the breast, including the nipple-areolar complex, without axillary node dissection; the pectoralis muscles were preserved [21].

Review of pathologic specimens

All available specimens from PM and in subsequent cases of ipsilateral BC were reviewed. PM tissue was categorized as follows: normal, with nonproliferative changes; proliferative changes without atypia; or atypical hyperplasia, carcinoma in situ, or invasive disease. Included patients had PM specimens without any evidence of invasive or in situ cancer. If an incidental malignancy was identified at PM, the patient was excluded.

Outcome definitions and statistical analyses

We considered the following as BC after PM: ductal carcinoma in situ or invasive BC in the chest wall, reconstructed breast mound, or axilla ipsilateral to the PM. If patients underwent CPM after BC diagnosis, we included only locoregional events ipsilateral to the CPM if the patient had been without evidence of disease from the initial cancer. Recurrences ipsilateral to the initial tumor were not considered. Data were summarized descriptively as median (range) for continuous variables and frequency (percentage) for categorical variables. Disease-free survival was the time from diagnosis of BC after PM until any of the following: ipsilateral BC recurrence, distant metastases, or death from any cause. Disease-free survival was estimated with the Kaplan-Meier method with 95 % CIs calculated with the natural logarithm of the survival function. Time from PM to BC diagnosis after PM was compared between BPM and CPM patients with the Wilcoxon rank-sum test. P values <0.05 were considered statistically significant. Analysis was performed using SAS (Version 9.3; SAS Institute, Inc) and R (Version 3.0.2, www.r-project.org) software.

Results

Clinical and pathologic characteristics at PM

We identified 25 patients with pathologically confirmed BC ipsilateral to PM. Outside correspondence involving three patients raised the possibility of BC after PM, but adequate medical records were not available to draw a definite conclusion; therefore, these patients were excluded from analysis. Median age at PM for the entire cohort was 43 years (range 32–69 years); for CPM patients and for BPM patients it was also 43 years. Median follow-up after PM for the entire cohort was 22 years (range 3–34 years); for CPM patients and for BPM patients, it was also 22 years.

Of the 25 patients, 13 underwent BPM; unilateral BC developed after PM in 12 patients, and bilateral BC developed in 1. Twelve patients with personal histories of BC treated with mastectomy underwent CPM and a subsequent BC developed ipsilateral to the CPM. Of the 25 patients, 12 (48 %) had a family history of BC (Table 1).

Table 1 Clinical and pathologic characteristics at prophylactic mastectomy
Full size table

Presentation of BC following PM

Median age at diagnosis was 56 years (range 38–81 years); it was 54 years for CPM patients and 58 years for BPM patients (Table 2). Median time to development of BC after PM was 7 years (range 1–25 years); it was not significantly different between BPM patients and CPM patients (6 and 8 years, respectively; P = 0.53). The most common presentation of BC following PM was a self-detected palpable abnormality (23 patients; 92 %). Although screening mammography was not routinely recommended following PM, 1 patient who underwent subcutaneous CPM had abnormal calcifications on a screening mammogram; BC was subsequently confirmed on excisional biopsy. Seventeen patients with BC after PM (68 %; 7 after CPM and 10 after BPM) presented with local disease only: 1 (4 %; after CPM) had both local and regional (axillary) disease; 4 (16 %; 3 after CPM and 1 after BPM) presented with disease limited to the axilla without evidence of local primary disease; and 3 (12 %; 1 after CPM and 2 after BPM) presented with synchronous local disease and distant metastases to bone.

Table 2 Characteristics of BC after PM
Full size table

Pathology of BC following PM

Median tumor size of resected BC after PM was 0.9 cm (range 0.3–3.5 cm) (Table 3); 20 % of patients had confirmed pathologically positive nodes. The histology of BC after PM was invasive ductal in 14 patients (56 %), invasive lobular in 2 (8 %), and unspecified in 9 (36 %). Estrogen receptor status was positive in 10 of 11 patients tested (91 %), although immunohistochemistry was performed in only 11 of 25 overall (44 %).

Table 3 Pathologic characteristics of breast cancer after prophylactic mastectomy
Full size table

Management of isolated local BC following PM

Surgery

The most common presentation of BC following PM was isolated local disease. Of the 17 patients who presented with isolated local BC after PM, 16 had prior subcutaneous mastectomy and 1 had prior total mastectomy. Completion or redo mastectomy was performed in 11 patients (65 %) and local excision of the tumor in 5 (29 %); surgical management was unknown for 1 (6 %) (Table 4). All 17 patients had undergone implant-based reconstruction. During surgical management of BC after PM, the implant was removed from 6 patients (35 %) and left in place in 7 (41 %); implant status was unknown for 4 (24 %).

Table 4 Management of isolated local breast cancer (clinically negative axilla) after prophylactic mastectomy
Full size table

Axillary staging was axillary lymph node dissection (ALND) in 10 of 17 patients (59 %), sentinel lymph node surgery in 1 (6 %), and no axillary staging in 1 (6 %); axillary management was unknown for 5 (29 %). Only 1 of 17 patients with a clinically negative axilla had a positive lymph node at axillary surgery.

Adjuvant therapy

Of 17 patients with isolated local BC after PM, 12 (71 %) received adjuvant therapy. Six (35 %) were treated with adjuvant radiotherapy. Radiotherapy was delivered to both the chest wall and the regional lymph nodes with a median dose of 6000 cGy (range 4200–6300 cGy). Of these 6 patients who received adjuvant radiotherapy, 4 also received systemic therapy (2 received chemotherapy; 1, tamoxifen; and 1, both). Of those patients who did not receive adjuvant radiotherapy, 2 received adjuvant tamoxifen and 1 received both adjuvant chemotherapy and tamoxifen. Chemotherapy was most commonly based on cyclophosphamide, methotrexate, and fluorouracil (CMF) or anthracycline. In 3 of the 17 patients, the use of adjuvant therapy was unknown.

Management of locoregional BC following PM

Of 4 patients who presented with isolated axillary disease without evidence of local primary disease, 3 underwent excisional biopsy of affected lymph nodes and 1 underwent ALND. Adjuvant therapy for these 4 patients was (1) adjuvant radiotherapy directed to the axilla followed by oophorectomy; (2) adjuvant radiotherapy directed to the axilla and chest wall followed by CMF chemotherapy; (3) tamoxifen alone; and (4) CMF and tamoxifen.

One patient presented with both local and axillary BC after PM and underwent completion or redo mastectomy with implant removal and ALND. Adjuvant therapy was unknown for this patient.

Of the 3 patients who presented with synchronous local and distant disease to bone, 1 underwent local excision of disease from the PM site without axillary surgery (leaving the implant intact), and the other 2 underwent palliative chest wall radiotherapy only.

Outcomes for patients presenting with BC following PM

For the 22 patients who presented with locoregional BC following PM, median follow-up time after diagnosis of BC following PM was 6.5 years. BC subsequently recurred in 5 patients treated with curative intent. For 2 patients, isolated local recurrence was the first event; for 3 patients, distant recurrence was first.

In 1 patient with isolated local recurrences, BC developed 3.3 years after CPM, with isolated axillary disease on the CPM side managed with excisional biopsy of 2 axillary nodules followed by axillary radiotherapy and oophorectomy. Subsequently, multiply recurrent chest wall disease developed on the CPM side and the patient died of distant metastatic disease 8 years following presentation of BC after PM at age 58. The second patient had a subcutaneous BPM and presented with isolated local BC involving the dermis and subcutaneous fat of the chest wall. She underwent total mastectomy and ALND, which showed a 1.4-cm invasive ductal adenocarcinoma and 13 negative nodes, followed by adjuvant radiotherapy to the chest wall and regional lymph nodes and doxorubicin-cyclophosphamide chemotherapy. Recurrent disease was identified on the chest wall 3 years after initial treatment, and distant metastases were identified the following year. She ultimately died of the disease 7 years following presentation of BC after PM at age 45.

In 3 patients, distant metastases developed as a first event following presentation and definitive management of isolated axillary BC after CPM (1 patient) and isolated local BC after BPM (2 patients). One patient presented with axillary fullness less than 1 year after CPM and underwent excisional biopsy of 2 pathologic lymph nodes, the largest measuring 3 cm. This was followed by CMF chemotherapy and radiotherapy to the chest wall and regional lymph nodes. Lung metastases developed only 6 months after completion of chemotherapy, and the patient died of disease the following year at age 53. A second patient presented with a 2.5-cm subareolar ductal adenocarcinoma 20 years after subcutaneous BPM and was treated with total mastectomy, ALND (all 18 nodes excised were negative), and adjuvant doxorubicin-cyclophosphamide-docetaxel chemotherapy followed by chest wall and regional nodal radiotherapy. She did not tolerate adjuvant tamoxifen. Bone metastases were identified 7 years after diagnosis, and she died of disease the following year at age 66. A third patient presented 3.7 years following subcutaneous BPM with disease localized to the inframammary region of the reconstructed chest wall. She underwent a total mastectomy and ALND for multifocal mammary adenocarcinoma (≤0.3 cm) identified with 9 negative nodes. She received no adjuvant therapy. Pulmonary metastases were diagnosed 8 years later, her health rapidly declined, and she died of disease at age 64.

The 5-year disease-free survival estimate was 69 % (95 % CI 52–94 %) overall (Fig. 1); the estimate for the 11 patients with isolated locoregional BC after BPM was 90 % (95 % CI 73–100 %) compared with 52 % (95 % CI 29–94 %) for the 11 patients with isolated locoregional BC after CPM (P = 0.23) (Fig. 2).

Fig. 1
figure 1

Disease-free Survival. Kaplan-Meier curve for disease-free survival is shown for the 22 patients who presented with isolated locoregional breast cancer after prophylactic mastectomy (PM) and were treated with curative intent. Error bars indicate 95 % CIs at 5 and 10 years

Full size image
Fig. 2
figure 2

Disease-free Survival. Kaplan-Meier curve for disease-free survival is shown for the 11 patients with isolated locoregional breast cancer after bilateral prophylactic mastectomy (BPM) and the 11 patients with isolated locoregional breast cancer after contralateral prophylactic mastectomy (CPM). Error bars indicate 95 % CIs at 5 and 10 years. PM prophylactic mastectomy

Full size image

Discussion

Our review of 1065 women who underwent BPM and 1643 women with unilateral BC treated with therapeutic mastectomy who underwent a CPM at our institution identified 25 patients in long-term follow-up who were confirmed to have locoregional invasive BC ipsilateral to the PM. Given the relative rarity of BC after PM, little has been published on this clinical entity. Our study, the largest known report of BC after PM, provides details on the most common presentations, pathology, treatment, and outcomes.

Our results suggest a high level of risk reduction is achieved with prophylactic mastectomy (both subcutaneous and total mastectomy), as previously published from these and other cohorts [4, 6, 16, 22]. Subcutaneous PM was performed much more frequently in the era of this study than in our current practice, where skin-sparing and nipple-sparing mastectomies, with more complete removal of breast tissue, are now used. We cannot compare the relative efficacy of subcutaneous mastectomy and total mastectomy due to limitations in follow-up of some patients and the small number of events. Although 80 % of BC after PM occurred in patients who underwent subcutaneous mastectomy, subcutaneous mastectomy was by far the more common procedure in the BPM patients [6] and was also performed in a significant number of CPM patients [16]. The fact that BC after PM occurred in patients who underwent total mastectomy highlights that BC after PM will continue to occur in the present era of risk-reducing surgery, particularly in the context of increasing rates of CPM and BPM [7]. Therefore, consideration of optimal management of these patients is important.

The most common presentation of BC after PM was locally confined disease, most frequently in various parts of the chest wall or in the subareolar region. This is likely due to residual breast tissue left at mastectomy because all PM specimens were negative for malignancy. For local BC after PM, wide local excision or redo mastectomy (or both) is appropriate. We currently recommend performing preoperative axillary ultrasonography to rule out regional disease. The accuracy of sentinel lymph node biopsy in patients with prior mastectomy is not established. Only 1 of 17 patients who presented with local BC after PM had axillary involvement and no patients had regional recurrence of disease, suggesting that modern imaging and sentinel lymph node biopsy may be reasonable approaches in patients with negative nodes to avoid the morbidity of axillary lymph node dissection.

Definitive conclusions cannot be made on the role of adjuvant therapies in patients with isolated local BC after PM because of the retrospective nature of this study, the small numbers, and the heterogeneity of the population. Both tumor biology and patient-specific factors, such as age and concurrent comorbidities, must be considered. Six of 17 patients (35 %) with isolated local BC after PM received adjuvant radiotherapy. The ability to obtain wide margins of resection while maintaining optimal cosmesis may be challenging in patients with BC after PM. In this study, the implant was left intact in 41 % of patients, which may limit the extent of resection. In addition, depending on the nature of the presentation, it may not be possible to fully exclude that the BC after PM was a recurrence from a previously occult breast malignancy not identified in the PM specimen. After breast-conserving surgery, adjuvant radiotherapy reduces recurrence and improves survival following wide local excision [23]. In addition, radiotherapy is routinely administered for isolated locoregional recurrence of BC following therapeutic mastectomy [24]. Therefore, adjuvant radiotherapy should be a consideration in multidisciplinary management discussions of isolated local BC after PM. By the same rationale, observation may be an appropriate option for elderly patients with small estrogen receptor-positive local BC after PM resected with negative margins and treated with endocrine therapy [25].

Many advances in systemic therapy will likely improve management of BC after PM in the future. For example, HER2 testing was available for only 12 % of patients in this study, none of whom had tumors overexpressing HER2 by immunohistochemistry. Therefore, no patients received HER2-directed therapy. In the present and future eras, disease biology will play a more important role in determining optimal locoregional and systemic management of BC after PM.

This study has several strengths, including the long-term median follow-up of more than 22 years, in addition to a median follow-up of more than 6 years after the diagnosis of BC after PM. Limitations include the retrospective nature of the study spanning several decades and the heterogeneity of the patient population, the lack of information on BRCA1 or BRCA2 status for the majority of patients, the incomplete tumor biology information, and the improvements in imaging, surgical techniques, radiotherapy, and systemic therapies in the years since patients with BC after PM were treated.

In conclusion, BC can occur after PM. Although the risk reduction from PM appears to be substantial, with the increase in rates of CPM and BPM, more cases of BC in women who have undergone PM could potentially occur in the future. With the increasing rates of PM, understanding the management of BC after PM is important. The most common presentation, disease localized to the breast, can be managed with surgical resection followed by consideration of adjuvant therapy.