Introduction

Studies evaluating breast reconstruction have shown that the presence of a nipple is associated with improved body image and quality of life after mastectomy. This finding is postulated to be a result of a more finished appearance to the breast and better symmetry with the contralateral breast in patients undergoing unilateral mastectomy. Reconstruction of the nipple areolar complex (NAC) improves the psychological impact, but outcomes after nipple reconstruction are somewhat variable. Loss of projection of the papilla, change in color or shape when compared to the contralateral breast, and lack of sensation have all been associated with patient dissatisfaction [1].

Nipple preservation offers an appealing alternative to nipple reconstruction for appropriate patients undergoing mastectomy. Historically, nipple-sparing mastectomy (NSM) was performed for prophylactic indications in women deemed high risk for development of breast cancer based on genetic carrier status, family history, or other pathology. The group chosen for this procedure was based on a landmark trial by Hartmann and colleagues published in 1999 that demonstrated 93% risk reduction with the use of prophylactic mastectomy in the high-risk population [2]. Since that time, the indications for NSM have expanded to include certain patients undergoing therapeutic mastectomy for breast cancer. The purpose of this review is to evaluate the oncologic safety, risk of complications, and quality of life outcomes of NSM in this population.

Preoperative Decision-Making

In patients with breast cancer, the goal of NSM is to remove the same amount of tissue as traditional mastectomy without compromising oncologic outcomes. Retrospective trials show pathologic nipple involvement to occur 5–12% of mastectomies, but some are reported as high as 58% [Fig. 1] [3••]. The patient’s clinical presentation at diagnosis serves as primary methodology to determine eligibility for nipple preservation; contraindications include pathologic nipple discharge and skin changes consistent with Paget’s disease [4]. However, the majority of patients with invasive or in situ carcinoma within the base of the nipple lack visible symptoms. It can be difficult to predict preoperatively with breast examination and conventional imaging alone who is an appropriate candidate [5].

Fig. 1
figure 1

H&E stain of nipple skin showing scattered Paget cells within the epidermis. The Paget cells have pale cytoplasm and enlarged nuclei. Courtesy of Paul Musto, M.D.

Full size image

Preoperative breast MRI may play a role in selecting appropriate candidates. As demonstrated by a retrospective cohort analysis of 137 women who underwent breast MRI prior to definitive mastectomy, the presence of non-mass-like enhancement extending up to the nipple was associated with significant risk of nipple involvement (incidence 52.7%, OR 21.7, p = 0.003). If the enhancement did not extend directly to the nipple, the incidence of nipple involvement was 6.1% and did not appear to be significantly different with a 2 cm threshold (p = 0.46) [6]. These findings are in contrast to data published by Mariscotti and colleagues who noted that the distance from the nipple on MRI was significant on multivariate analysis. The authors suggested that a 1-cm distance from the nipple allows for 81% diagnostic accuracy with improved sensitivity and specificity for predicting nipple involvement [7].

There is no widely accepted location for incision for NSM; the decision regarding placement depends on multiple factors, including tumor location, breast size, ptosis, and planned method of reconstruction. Inframammary incisions allow for the scar to be located under the reconstructed breast which result in excellent esthetics. However, this approach can be technically difficult to resect tissue from the upper pole of the breast in patients with large, ptotic breasts, and require a counter incision for patients undergoing axillary surgery. Peri-areolar and radial incisions that split the breast allow for better visualization for patients with all breast sizes and allow for access to the axilla more easily [8]. These may also be utilized more commonly by surgeons with less experience performing NSM.

The National Comprehensive Cancer Network panel states that pathologic assessment of the nipple margin is mandatory for patients undergoing therapeutic NSM [4]. There is variation between practice settings as to whether frozen sectioning (FS) of the retro-areolar tissue is routinely performed versus routine permanent sectioning. Knowledge of pathology intraoperatively allows for immediate guidance regarding the need to excise the nipple, thereby saving the patient additional surgery. Resection of the nipple results in a smaller residual skin envelope that may affect implant size in patients undergoing immediate direct-to-implant reconstruction. Retrospective studies reveal FS to have moderate sensitivity (58–80%) and high specificity (88–100%). False-positive results are rare, and false-negative rates range from 2 to 6% [9,10,11]. These rates are likely due to sampling error, tissue loss during processing, or artifacts from diathermy and freezing. Suarez-Zamora and colleagues were the first to evaluate interobserver agreement in retro-areolar FS of 34 NSM; concordance between breast pathologists was higher than general pathologists (kappa value 0.87 vs 0.31, p < 0.0001) [9]. The decision whether to incorporate routine intraoperative FS is unique to each institutional practice setting based on resources; however, a cost analysis performed by Alperovich in 2016 showed routine FS in this setting adds $95 per breast [10]. In patients with positive nipple margins detected on permanent sectioning, resection of the nipple is recommended, as 40% of patients have disease within the nipple on final pathology of re-excision [11].

There has been some postulation that the use of NSM would result in increased rates of positive margins in other areas of the breast due to lack of visualization when compared to traditional skin sparing mastectomy (SSM). However, as demonstrated by a review from the National Cancer Database, there was no difference in rates of positive margins between the two approaches (4.1% NSM vs 3.9% SSM, p = 0.6) [12]. Additionally, as the indications expand for patients with more extensive disease, it is reassuring that utilization of NSM does not appear to delay receipt of adjuvant therapies such as chemotherapy, radiation, or endocrine therapy [12, 13].

Complications

Studies are conflicting as to whether NSM is associated with increased risk of surgical complications. The overall complication rates range from 20 to 25% as shown by two systematic reviews [2, 14]. This finding is likely a result of technical factors, as the approach required for enhanced cosmesis and nipple preservation may result in tension on the skin flaps and nipple, leading to necrosis. A Cochrane review comparing NSM to other types of mastectomy concluded that there did not appear to be an increase in adverse events with NSM [15]. This review is in stark contrast to the findings reported by others, including Agha et al., who noted a significant increase in complications with NSM when compared to SSM (22.6% vs 14.0%). The higher complication rate is attributed to the risk of nipple necrosis, which is not an issue when the nipple is removed [3••, 16•, 17••]. Interestingly, the potential increase in complications with NSM does not appear to translate to higher rates of readmission or reoperation [17••]. Moreover, these results may be time and experience dependent. Headon et al. noted a significant reduction in both overall complications and nipple necrosis when comparing studies performed prior to and after 2013. This finding was attributed to improvements in surgical technique and increasing surgeons’ comfort with the procedure [14].

Necrosis of the NAC is the most severe complication of NSM, as it can lead to deformity, hypopigmentation, or loss of the NAC [18]. Rates of nipple necrosis as showed by recent retrospective institutional series range from 0.2 to 16.7% (Table 1). Patients with central tumors close to the nipple may be at increased risk. As shown by Park and colleagues, for every 1 cm increase in tumor-nipple-distance, there was a 0.712-fold reduction in nipple necrosis (p = 0.012) [18]. These findings are similar to the study by Balci et al. that demonstrated patients with tumors within 2 cm of the nipple have increased risk of complete NAC necrosis (5.1% vs 0.7%, p = 0.028) [19]. Although the risk is elevated with central tumors, the risk of NAC necrosis is still low. Careful dissection of the tissue behind the nipple is crucial in order to preserve the blood supply while achieving negative margins.

Table 1 Incidence of nipple necrosis after nipple-sparing mastectomy
Full size table

Multiple authors have sought to evaluate whether the choice of skin incision in NSM affects risk of complications. Garwood and colleagues demonstrated that incisions that cross the areola or utilize > 30% of its border are associated with 3.8-fold higher rates of NAC necrosis [20]. These findings were confirmed by Park et al., who concluded that the type of incision was significantly associated with risk of nipple necrosis in 275 patients who underwent NSM with immediate reconstruction (p < 0.001). Peri-areolar incisions were associated with highest risk when compared to radial and inframammary incisions [8, 18].

Authors have reported other risk factors for complications after NSM. Larger breast size and weight of the mastectomy specimen have also been associated with increased risk of NAC necrosis on multivariate analysis (p = 0.014). Possible explanations of this relationship are anatomical, with longer distance of the source of blood supply to the nipple, and technical, with potential increased tension applied on the skin in order to achieve adequate visualization during surgery [8].

Smoking has been long associated with higher surgical morbidity secondary to subclinical microvascular disease. Not surprisingly, as viability of NAC and skin flaps after NSM relies on the microvasculature, multiple studies have confirmed a higher risk of complications in patients who smoke [3••, 14, 21]. Frey and colleagues noted a significant difference with a 10 pack-year history of tobacco use, current smokers, and those within 5 years of quitting [21].

Others have sought to determine whether the use of neoadjuvant chemotherapy prior to NSM impacted rates of surgical complications when compared to patients undergoing primary surgery. In a cohort of 832 patients over a 28-year span, Bartholomew and colleagues noted similar 30-day surgical complication rates with neoadjuvant chemotherapy (5.7%) vs. primary surgery (10.6%). The authors concluded that NSM was safe in this group of patients [22].

It is well established that radiation therapy (RT) to the breast either before or after surgery is a risk factor for wound healing complications due to fibrotic changes, loss of stem and parenchymal cells, and release of cytokines [23]. In the patients who develop in-breast recurrences after prior breast conserving surgery and RT, decisions must be made as to whether to preserve the nipple. Moreover, as the indications for NSM expand to patients with larger tumors and positive axillary nodes, a significant number of patients may undergo PMRT. The data regarding NSM in the irradiated breast is limited, consisting of single institutional series with small numbers and short follow-up. Two retrospective analyses evaluated the relationship of RT and complications after NSM. Tang et al. compared 166 women with a history of RT (69 prior to NSM and 97 PMRT) to 816 controls, with median follow-up of 23 months. Reish et al. compared 88 women with a history of RT (45 prior and 43 PMRT) to 517 controls, with median follow-up 22 months. As one would expect, the patients who underwent RT did have significantly increased risk of complications, including skin necrosis and nipple loss. However, in both studies, nipple preservation was successful in greater than 90%, and rates of reconstruction failure were less than 8%. Therefore, both studies concluded that history of RT either before or after surgery is not an absolute contraindication for NSM [24, 25].

Oncologic Outcomes

There are no randomized trials evaluating safety and efficacy of NSM for the treatment of breast cancer. The data is limited to institutional series, almost always retrospective, and often includes patients with prophylactic indications. Critics have questioned whether nipple preservation leaves residual terminal duct lobular units (TDLUs), thereby increasing risk of local recurrence (LR). Pathologic examination of nipple and retro-areolar tissue shows presence of TDLUs in only 8% of specimens [11]. Furthermore, the current data using retrospective analyses estimate LR to range from 0.6 to 11.9%, with recurrence in the nipple 0–4.9% (Table 2).

Table 2 Local recurrence after nipple-sparing mastectomy in retrospective series
Full size table

A Cochrane review published in 2016 included 11 studies and compared 6502 participants, including 2259 women who underwent NSM, 818 who underwent SSM, and 3671 women who underwent traditional mastectomy. The risk of LR in patients with NSM was 6.8%, with specific recurrence in the nipple in 1.8%. The conclusion of this review was that the data was insufficient to determine whether there was any difference in LR or survival between the groups. It deemed the quality of evidence to be very low, as there were variances in study design, significant risk of selection bias, and confounding factors that may have affected results. It also noted short study follow-up as a confounding factor, as only 6 of 11 studies within the analysis documented mean follow-up greater than 60 months [15]. To address this issue, Delacruz and colleagues performed a separate analysis of 1212 patients in 6 trials comparing NSM with traditional mastectomy with at least 5 years of follow-up (range 60–136 months). There was no difference in LR, disease-free survival, or overall survival between the 2 groups. Risk of recurrence in the NAC after NSM ranged from 0 to 3.7% [26].

Two additional systematic reviews have been published since the Cochrane analysis. The first, performed by Headon and colleagues, included 73 studies and 10,935 patients. The pooled rate of LR was 2.4% at a mean follow-up 38.3 months (range 7.4–156 months) [14]. Another systemic review and meta-analysis published in 2019 by Agha et al. included 3015 breasts, 1419 treated with NSM and 1596 with SSM. Local recurrence was 3.9% with NSM and 3.3% with SSM, with follow-up ranging from 18 to 101 months (mean not provided). Statistical analysis demonstrated no difference in LR (p = 0.45) or mortality (p = 0.34) between the 2 groups [3••]. Although both reviews suggest a low incidence of LR after NSM, they did not exclude patients undergoing prophylactic surgery, and had relatively short follow-up, which may affect interpretation of results.

Wu and colleagues sought to establish risk factors for LR within the NAC. In their retrospective analysis of 944 women who underwent NSM for invasive breast cancer, 39 (4.1%) developed recurrence at the NAC as first event, with median time to recurrence 35 months. The risk factors for NAC recurrence as shown by multivariate analysis were multifocal/multicentric disease (HR 3.3), presence of extensive intraductal component (HR 3.3), ER negative/HER-2 positive subtype (HR 3.05), and high grade (HR 2.6) [27].

Retrospective studies have been performed looking specifically at outcomes of NSM in patients with other high-risk features such as those who undergo neoadjuvant chemotherapy, have history of prior RT, or need PMRT. As demonstrated in an analysis of 39 women who underwent neoadjuvant chemotherapy prior to surgery, the use of chemotherapy resulted in significant reduction in the median tumor size (3.1 to 0.9 cm). In addition, in the women deemed cautionary for NSM prior to chemotherapy based on tumor location within 2 cm of the nipple, 90% was able to successfully undergo nipple preservation. At mean follow-up 67.2 months, local recurrence was seen within the tumor bed in 2 of 39 patients (5.1%), and no nipple recurrences were seen [28]. In the two studies evaluating the relationship of RT and outcomes after NSM, local recurrence rates were acceptably low at median follow-up 22–23 months (2.9% pooled result of both studies). No LR was seen in the patients with prior RT and 4 seen in the patients with PMRT [24, 25]. These outcomes are within acceptable ranges for patients with high-risk features.

Patient Satisfaction

The main benefits of nipple preservation and reconstruction are psychological; therefore, patient satisfaction (PS) should be considered an important outcome measure. Assessment of this measure is difficult from a statistical standpoint as it relies on use of patient questionnaires which introduce significant bias. There are also different methods to measure satisfaction, making comparison of different studies more complex.

The BREAST-Q© is a commonly utilized instrument for assessment of patient-reported outcomes after breast surgery. It includes domains focusing on PS and different aspects of health-related quality of life. Wei and colleagues used this instrument to compare PS after surgery in 52 women who underwent NSM to a reference group of 202 women who underwent SSM and NAC reconstruction. On multivariate analysis after adjustment for confounding factors, patients after NSM reported higher psychosocial and sexual well-being compared to their SSM counterparts. Interestingly, there was no significant difference in the domain evaluating physical qualities of the breast reconstruction, such as the softness of the breast and symmetry with the contralateral breast [1]. A matched-pair analysis reported by Yoon-Flannery et al. also noted improved postoperative satisfaction scores in regard to sexual well-being, but no difference in the other domains [29].

The above studies included results from postoperative scores only; it could be questioned whether results may be altered if the preoperative results were included as a comparison. Two separate reviews using the National Cancer Database showed that patients who undergo NSM more likely to be younger, healthier, have smaller body mass indices and more likely to have smaller, earlier stage cancers that are less likely to require adjuvant chemotherapy and radiation [12, 13]. These are all clinical factors that can affect satisfaction.

To assess this baseline difference, an analysis of 513 patients with both preoperative and postoperative BREAST-Q© scores was performed by Romanoff and colleagues (72 patients after NSM and 443 patients after SSM). As predicted above, there were differences in baseline characteristics of the 2 groups, and while patients undergoing NSM had higher median baseline scores, they did not differ significantly from SSM. Additionally, after adjusting for the preoperative scores and clinical variables, NSM was only associated with higher satisfaction of psychological well-being (p = 0.035); the remainder of the scores did not differ significantly [16•].

Satteson et al. performed a more indirect comparison of patient satisfaction after NSM and SSM. Their systemic review included 23 studies evaluating NSM and methods of NAC reconstruction. Since there were multiple instruments utilized within the studies, the authors developed a method to calculate a “satisfaction score” for each study. Overall, patient satisfaction scores were significantly higher in studies evaluating NSM when compared to those evaluating NAC reconstruction (weighted average 80.5% vs 73.9%, p = 0.079) [30].

Patient-reported assessment of nipple-specific outcomes was performed by Peled et al. in 28 women who underwent NSM with immediate expander reconstruction. At 1 year, nearly all patients reported satisfaction with the shape, appearance, and feel of their nipples. However, only 40% of patients reported satisfaction with nipple sensation [31]. Loss of nipple sensation was also seen in a series of 35 patients who underwent NSM. Pinprick tests revealed partial recovery of nipple sensation in 44% of patients at 6 months after surgery and 60% of patients at 1 year [32].

Conclusion

In patients undergoing therapeutic mastectomy for breast cancer, there are multiple benefits to nipple preservation including improved esthetics, prevention of need for additional surgery, and improved psychological well-being. Although there are no randomized trials evaluating oncologic safety of NSM in this group, as presented here, the data is promising. The risks of local recurrence and complications are well within acceptable ranges and do not appear to be statistically different when compared to patients undergoing SSM; however, longer follow-up is needed. There is no longer an ideal candidate for NSM; multiple factors should be considered when assessing for eligibility for nipple preservation, including preoperative presentation and risk factors for nipple involvement.