The Oncological Safety of Nipple-Sparing Mastectomy: A Systematic Review of the Literature with a Pooled Analysis of 12,358 Procedures
Article information
Abstract
Nipple-sparing mastectomy (NSM) is increasingly popular as a procedure for the treatment of breast cancer and as a prophylactic procedure for those at high risk of developing the disease. However, it remains a controversial option due to questions regarding its oncological safety and concerns regarding locoregional recurrence. This systematic review with a pooled analysis examines the current literature regarding NSM, including locoregional recurrence and complication rates. Systematic electronic searches were conducted using the PubMed database and the Ovid database for studies reporting the indications for NSM and the subsequent outcomes. Studies between January 1970 and January 2015 (inclusive) were analysed if they met the inclusion criteria. Pooled descriptive statistics were performed. Seventy-three studies that met the inclusion criteria were included in the analysis, yielding 12,358 procedures. After a mean follow up of 38 months (range, 7.4–156 months), the overall pooled locoregional recurrence rate was 2.38%, the overall complication rate was 22.3%, and the overall incidence of nipple necrosis, either partial or total, was 5.9%. Significant heterogeneity was found among the published studies and patient selection was affected by tumour characteristics. We concluded that NSM appears to be an oncologically safe option for appropriately selected patients, with low rates of locoregional recurrence. For NSM to be performed, tumours should be peripherally located, smaller than 5 cm in diameter, located more than 2 cm away from the nipple margin, and human epidermal growth factor 2-negative. A separate histopathological examination of the subareolar tissue and exclusion of malignancy at this site is essential for safe oncological practice. Long-term follow-up studies and prospective cohort studies are required in order to determine the best reconstructive methods.
INTRODUCTION
In recent decades, advances in breast surgery and reconstruction have led to significant improvements in the quality of life for many patients who have undergone surgery for breast cancer. Although breast-conserving surgery has remained the treatment of choice for many, up to a third of patients still require, or request, a mastectomy in order to achieve local disease control. Currently, the indications for mastectomy include cases of invasive breast cancer (IBC) or ductal carcinoma in situ (DCIS) that are not amenable to breast-conserving surgery for a number of reasons, such as the size or distribution of disease, the possibility of a poor cosmetic outcome, local recurrence, contraindications to radiotherapy, or patient preference.
The primary aim of surgical intervention in breast cancer is to achieve optimal local disease control, although secondary aims have emerged, such as good cosmetic outcomes and high patient satisfaction. Mastectomy techniques have evolved from radical and modified mastectomy towards surgical methods that facilitate reconstruction and tend to lead to good cosmetic outcomes: the standard skin-sparing mastectomy (SSM) and, more recently, the nipple-sparing mastectomy (NSM).
Despite its advantages in terms of aesthetic outcomes, the use of NSM is controversial due to concerns regarding its oncological safety based on early reports of high rates of nipple malignancy, potentially leading to increased recurrence rates. However, more recent reports have suggested that the earlier studies may have exaggerated the risk [1]. Although patient satisfaction with the aesthetic outcome and oncological safety have historically been regarded as competing interests, more recently these two aims have evolved to become common goals and are no longer considered mutually exclusive outcomes, given appropriate patient selection. NSM is one such surgical option, designed to allow the local control of disease whilst optimising the cosmetic outcome. In NSM, the mammary tissue is removed, whilst the skin envelope and nipple-areola complex (NAC) are preserved. Preserving the NAC has been reported to be associated with multiple benefits, including better aesthetic outcomes, improved patient satisfaction, and psychosexual benefits, and its removal may, in selected cases, be an instance of overtreatment [2]. Additionally, leaving the NAC in situ can facilitate immediate breast reconstruction [3].
Currently, no universally agreed-upon criteria are used for patient selection for NSM. The suggested parameters include early-stage IBC (stage I to stage II) or DCIS where the tumour has a diameter smaller than 5 cm, is peripherally placed more than 2 cm from the areola, is oestrogen- and progesterone-positive and human epidermal growth factor-2 (HER2)-negative, with no evidence of multifocal or multicentric disease and no lymphovascular invasion [4]. The purpose of this study was to systematically review the literature in order to determine the rate of locoregional recurrence (LR) [5] following NSM as well as to determine the rates of overall complications and nipple necrosis in order to update standards of clinical practice when considering NSM.
METHODS
Search methodology
Electronic searches were performed using the PubMed and Ovid databases for studies evaluating NSM. The search terms that were used included 'nipple sparing mastectomy' and 'total skin sparing mastectomy' along with 'locoregional recurrence' and 'outcomes.' All studies published from January 1970 to January 2015 were reviewed and the references of appropriate articles were also evaluated for further relevant studies.
Selection criteria
The inclusion and exclusion criteria were defined before data collection was initiated. Studies that evaluated outcomes following NSM were included. All forms of reconstruction were included. All studies must have clearly defined whether patients underwent NSM or SSM; if this was unclear, the article was not included. In articles where both SSM and NSM were reviewed, the article was required to clearly state the outcomes of the NSM cohort separately. Retrospective and prospective studies were included. Fig. 1 outlines the selection process. Studies were excluded if they did not report appropriate outcomes as stated, if they did not specify the number of patients and the number of procedures involved, were not printed in English, or were reports, commentaries, reviews or letters.
Data collection and analysis
The following data were collected: authors, study name, publication year, location of the study, journal of publication, type of study, number of patients, number of procedures, inclusion criteria for NSM, type of reconstruction, number of overall complications, nipple necrosis, LR, and aesthetic results. The pooled analysis of the rate of LR, the nipple necrosis rate, and the rate of overall complications was performed based on the number of patients included in each study. The primary outcome measures were the rate of overall LR recurrence, the overall complication rate, and the overall rate of nipple necrosis.
RESULTS
Seventy-three studies [5678910111213141516171819202122232425262728293031323334353637383940414243444546474849505152535455565758596061626364656667686970717273747576] reported LR rates, complication rates, and/or nipple necrosis rate following NSM. Table 1 presents the number of patients, the number of procedures, the LR rate, the complication rate, and the nipple necrosis rate. Almost all of the studies were retrospective (91%). The 73 studies yielded 12,358 procedures in 10,935 patients, and the indications included invasive breast cancer, risk-reduction surgery, and carcinoma in situ. The mean follow-up period was 38.3 months, with a range of 7.9–156 months. Pooled analysis demonstrated an overall LR rate of 2.38%, slightly higher than that found by Endara et al. [77] in 2013. The overall complication rate was 22% and the nipple necrosis rate was 5.9%, both lower than the rates found by Endara and colleagues. The majority of studies were published after 2011, reflecting the increasing popularity of NSM over time. A small subgroup analysis was carried out examining the average complication rates before and after 2013, and the results of this are shown in Fig. 2. A clear reduction was observed in the complication rate and the incidence of nipple necrosis after 2013.
DISCUSSION
According to our pooled analysis, NSM may be an oncologically safe surgical treatment in carefully selected patients with breast cancer, as we found a low pooled LR rate of 2.38%, which is comparable to that found for conventional mastectomy, for which the LR rates for have been reported to be as high as 16% [4]. It can also lead to improved aesthetic outcomes with minimal scarring when combined with reconstruction. The main oncological concern with NSM is the possibility of leaving residual tumour cells within the skin envelope, which may manifest later as LR.
Histological studies following conventional mastectomy have reported residual glandular tissue in 5% of all biopsies, indicating that more radical surgery may not be guarantee of complete clearance [78]. In SSM performed in patients with invasive breast cancer, the prevalence of residual breast tissue has been reported to be as high as 59.5%, with residual disease in 9.5% [79], a finding echoed by Ho et al. [80], who reported that skin flaps exhibited residual malignancy in 23% of cases, most commonly in the skin overlying the tumour. However, a large systematic review from 2012 reported that the overall incidence rate of LR was only 0.9% after a mean follow-up of 38.4 months and that the skin flap recurrence rate was 4.2% following SSM [4], which was much lower than had been reported in single-centre studies.
The concern has been raised that leaving the NAC in situ could provide another site for possible recurrence. We have found a low LR rate, which corresponds to the findings of long-term single-centre studies. Stanec et al. [65] found an LR rate of 3.7% in the breast and 1.2% in the NAC after a mean follow-up of 63 months. Other studies have found no cases of NAC recurrence [4869]. Most recently, Sakurai et al. [55] followed up 788 patients who underwent NSM for an average of 78 months, and reported an NAC relapse rate of 3.7% and an LR rate of 8.2%, stating that no significant difference was found in overall survival and disease-free survival between NSM patients and conventional mastectomy patients at 21 years.
Tumour size is a parameter that has been used in the past to select suitable patients for NSM. In the studies analysed that included data on tumour size, the largest recorded was an average of 3.4±2.2 cm, which showed a 10.3% LR rate over an 18-month follow up [56]. More recently, Leclere et al. [67] reported a 5.3% LR rate in 41 cases, 18 of which involved a tumour size larger than 3 cm Moreover, it appears that as NSM becomes more popular, the number of patients with larger tumours undergoing NSM is steadily increasing, as described by Agarwal et al. [81]. They found that although 50% of patients had tumours less than 2 cm in diameter, the number of patients with tumours larger than 2 cm undergoing NSM increased over time, perhaps because as surgeons have become more familiar and confident with the technique, it has been possible to widen the clinical indications for which NSM can be considered as a management option The currently available data therefore appear to support the use of NSM even with larger tumours, although more long-term follow-up studies with larger populations are required in order to ensure that the risk of LR is minimised.
Another selection parameter that has been suggested is a tumour-to-nipple distance no smaller than 2 cm. Currently, when performing an NSM, a frozen section of the retroareolar tissue is required for intraoperative histological analysis in order to verify that the borders of the NAC are clear of disease. A recent study by Chattopadhyay et al. [66] found a 0% LR rate after 28.5 months in patients with a median tumour-to-nipple distance of 3.8 cm This low LR rate is echoed by Fortunato et al. [56], who found an LR rate of less than 1% after 26 months, despite a tumour-to-nipple distance of just 1 cm in 65% of cases suggesting that previous guidelines requiring a minimum distance of 2 cm may be outdated for carefully selected patients. In order to preoperatively evaluate the tumour-to-nipple distance, it has been suggested that patients undergo magnetic resonance imaging (MRI) before surgery as a way of predicting who would be suitable for NSM. Moon et al. [82] found a significant correlation between NAC enhancement on MRI and malignant invasion of the NAC, reporting a sensitivity of 93.8% and a specificity of 85.7%, indicating that presurgical MRI may be useful in selecting appropriate patients However, not all authors agree on the use of MRI, meaning that a combined approach, utilising both clinical evaluation and imaging techniques, may be best when determining patient suitability for NSM.
The tumour location within the breast is also considered an important parameter when assessing suitability for NSM, as centrally located tumours have been reported to be significantly associated with nipple margin involvement: one study found that centrally located tumours were associated with residual malignancy of the NAC in 40% of cases [68]. Most studies reporting on tumour location support this proposal [483]. Mallon et al. [4], in their large systematic review, stated that a retroareolar or central location was a factor influencing occult nipple malignancy, with an overall incidence of 35.2%, compared to only 9.7% for peripherally located tumours One study suggested that tumour location was the most important parameter in predicting NAC involvement, as only centrally located tumours showed a significant association with nipple involvement [83]. A retrospective study of 219 mastectomies demonstrated that peripheral tumours were found to involve the NAC in only 2.5% of cases, whereas for centrally located lesions the incidence was 68% [84]. In contrast, a recent small study of 28 patients concluded that NSM was an oncologically safe procedure regardless of tumour location [85]. More research is required in larger cohorts before central tumours can be deemed safe for NSM.
The histological subtype of the breast cancer may also have an impact on the LR rate after NSM. A systematic review suggested that the overall LR rate for invasive ductal carcinoma was 14.9%, whilst for DCIS it was 15.3%. The highest LR rate was for invasive lobular carcinoma, with an LR rate of 17.2% [4]. Recently, a study of 934 patients also investigated the LR rate of different histological types with a mean follow-up of 50 months, and found a 3.6% LR rate in the breast and a 0.8% recurrence rate in the NAC in patients who had had invasive carcinoma compared to a 4.9% LR rate in the breast and a 2.9% recurrence rate in the NAC in those who had had intraepithelial neoplasia, supporting the proposal that histological subtype may be an important factor when considering the suitability of NSM [86]. However, Kneubil et al. [45] found that having DCIS as a primary tumour was a significant predictor of NAC recurrence, although the LR rate was still acceptably low at 8.7%. Overall, the evidence regarding the impact of histological type on LR rate is scarce, and currently no adequate study has evaluated the differences in LR between all the different histological subtypes.
A further parameter that has been highlighted is the tumour receptor status, with tumours that are HER2-negative and oestrogen receptor-(ER) and progesterone receptor (PR)-positive having the best outcomes following NSM. Mallon et al. [4] found a significant increase in the LR rate in HER2-positive patients, with an overall incidence of 19.7%, compared to 10.1% in HER2-negative patients. Additionally, ER positivity was found to be protective, with an overall LR rate of 10.8%, compared to 14% in ER-negative patients, a trend also echoed by PR positivity. More recent studies have found similar results. A positive HER2 status has been consistently found to influence LR rates, and most authors agree that HER2 positivity should rule out NSM as an option for surgical management [86]. Petit et al. [86] found in their study that out of 11 patients that had LR, 9 exhibited HER2 overexpression, indicating a strong correlation between HER2 and LR in NSM. The evidence linking ER and PR status and NSM outcomes is currently scant, although a number of studies have examined the risk of LR in patients undergoing conventional mastectomy. One such study found that, with the addition of post-mastectomy radiation, the cumulative incidence of LR was 8.6% vs. 4.4% for ER-negative tumours and ER-positive tumours, respectively, and 8.5% vs. 3.4% for PR-negative and PR-positive tumours respectively [87]. Consequently, it can be concluded that a HER2-negative status coupled with a positive ER and PR status confers the best protection against LR and therefore would be suitable for NSM, providing the patient fits the other criteria discussed above.
Necrotic complications associated with NSM
The overall complication rate and the incidence of nipple necrosis were included as secondary outcomes of the pooled analysis. The overall complication rate was 22.3% and the nipple necrosis rate was 5.9%. Due to the extensive undermining of the NAC during NSM, it is thought that NSM may lead to an increased incidence of necrotic complications. A number of studies have reported data on nipple necrosis, with incidence rates ranging from 3.5% for total nipple necrosis to 12.1% for partial nipple necrosis [30606488]. Necrosis can occur as a quite early complication, with Radovanovic et al. [30] finding a major skin necrosis rate of 3% after just 6 weeks The concern with nipple necrosis is that it can lead to loss of the NAC at a later date [67].
Some studies have assessed whether any risk factors may be associated with an increased risk of nipple necrosis following NSM, because the identification of such risk factors could help clarify patient selection criteria. Some studies have linked a higher nipple necrosis rate with obesity [64], which may be due to the larger breast volume. Another factor that has been found to increase the risk is a positive retroareolar margin [53]; therefore, if possible, it is important to perform frozen section analysis intraoperatively in order to minimise this risk.
Consequently, it would appear that despite the risk of necrotic complications, the actual incidence of necrosis remains low, meaning that NSM may still be a viable option. Those at a higher risk, such as those with a higher body mass index or large breast volume, should be individually assessed for suitability with the options of an autologous tissue flap or two-stage reconstruction discussed in order to minimise the possibility of revisional surgery.
Radiotherapy and NSM
Radiotherapy is an important treatment option in the management of breast cancer, as it has been found to reduce the LR rate, with one study finding that patients who underwent radiotherapy had a LR rate of 8.5% compared to 28.4% in those that did not undergo radiotherapy over a 13-year follow-up period [15]. The use of radiotherapy in NSM has been a major topic of discussion, as it has been associated with an increase in long-term complications; if radiotherapy needs to be administered after a mastectomy, the reconstructed breast can also complicate the planning and delivery of the radiotherapy [89]. Radiotherapy is also associated with complications in the reconstructed breast, such as fat necrosis and volume loss in reconstructions using autologous tissue [90] and capsular contracture in those using implants [91]. In terms of nipple necrosis, however, it appears that including radiotherapy in the treatment of the patient does not increase the risk of NAC necrosis [92].
Preventing complications in NSM
One way in which the complication risk could be reduced is through the use of an acellular dermal matrix (ADM) in the reconstruction process. Small trials examining the use of ADMs with NSM have found that immediate breast reconstruction can be achieved with minimal complications [93]. One advantage is that use of an ADM allows the surgeon to use prepectoral placement, which has been reported to be associated with good patient satisfaction and excellent cosmetic results [94]. An example of NSM with immediate ADM-assisted implant based reconstruction can be seen in Fig. 3. No evidence is yet available on the effect of ADM use on necrotic complication rates.
The use of an ADM should therefore be considered when planning an NSM and immediate reconstruction procedure with the goal of reducing complications.
CONCLUSIONS AND RECOMMENDATIONS FOR PRACTICE
A growing body of evidence suggests that NSM is oncologically adequate for early-stage IBC and DCIS in carefully selected patients. Patients with a peripherally located tumour less than 5 cm in diameter, located more than 2 cm from the NAC, not showing HER2 overexpression, and exhibiting a positive ER and PR status may be considered for NSM with or without adjuvant radiotherapy. NSM can facilitate immediate breast reconstruction, providing an aesthetically superior treatment option for women who are not suitable for breast-conserving surgery. The optimal integration of NSM and radiotherapy has yet to be established, but radiotherapy is not associated with an increased risk of post-mastectomy nipple necrosis. Dual procedures combining NSM and immediate breast reconstruction afford many advantages, including fewer hospital admissions and a reduced need for contralateral breast adjustment in order to achieve symmetry. In order to ensure the best outcomes, it is important to highlight appropriate patient selection in order to minimise the risk of LR, as well as requiring coordination between oncological and reconstructive surgeons and an effective multidisciplinary team. In the future, a meta-analysis investigating all of the factors that affect LR recurrence after NSM should be carried out in order to facilitate optimal long-term outcomes.
Notes
No potential conflict of interest relevant to this article was reported.