In addition to conflicting data about oncologic safety, there is also debate about the impact of reconstruction prior to PMRT on cosmetic outcomes. The main complications caused by radiation on the reconstructed breast include fat necrosis, impaired wound healing, contracture, fibrosis, volume loss, and architectural distortion [23]. There are data to support superior cosmetic results with delayed reconstruction compared with immediate reconstruction. Javaid et al. [23] in a systematic review of ten published reports of patients undergoing immediate and delayed reconstruction and PMRT found a higher incidence of breast fibrosis and contracture with immediate reconstruction. Similarly, Kronowitz et al. [16], in a systematic review of 49 articles, reported high rates of contracture and implant loss among patients undergoing immediate reconstruction prior to PMRT.

Other groups have also reported lower rates of complications after delayed reconstruction. Adesiyun et al. [24], in a review of 113 patients who underwent immediate or delayed breast reconstruction with PMRT, reported a lower rate of complications in the delayed-reconstruction group (32 % vs. 44 %, p = 0.18), although this difference was not statistically significant. The patients’ general satisfaction with their cosmetic outcome was similar in the two groups (68 %) [24]. Another group found no significant difference in complication rates with immediate or delayed reconstruction with TRAM flaps in patients who received PMRT, but the authors ultimately recommended delayed reconstruction because of possible low power of the study [25].

Compared with the aforementioned studies, other groups have reported acceptable cosmetic results and complication rates with immediate reconstruction. A meta-analysis of 11 studies by Barry et al. [26] concluded that postoperative outcomes did not differ depending on whether reconstruction was performed before or after PMRT. Autologous flaps appeared to have superior outcomes. Postoperative complications such as fibrosis, contracture, infection, fat necrosis, and reoperation were lower with autologous flap reconstruction than with implant reconstruction [26]. Thus, if immediate reconstruction is pursued, many authors advocate reconstruction with an autologous flap over a tissue expander/implant to enhance cosmetic results [6].

Although many authors have reported superior outcomes with flap reconstruction compared with implant reconstruction prior to PMRT, this does not necessarily imply that successful outcomes cannot be achieved with implant reconstruction. For example, Cordeiro et al. [27, 28] reported satisfactory aesthetic results with immediate tissue expander placement, followed by exchange for a permanent implant prior to radiotherapy. Aesthetic results were categorized as “good to excellent” in 80 % of patients, with an implant loss rate of 11 % [27].

In patients with inflammatory breast carcinoma, delayed reconstruction is recommended because of extensive skin involvement and a high risk of local recurrence [29]. The required resection of skin precludes a skin-sparing mastectomy. Furthermore, timely administration of radiotherapy is imperative, making the delay for healing after reconstruction undesirable. Therefore, reconstruction should be delayed in patients undergoing mastectomy for inflammatory breast cancer. This recommendation is reflected in the 2012 National Cancer Comprehensive Network guidelines [30].

There are two small series that have reported success with immediate reconstruction. Chin et al. [31] performed a retrospective analysis of 23 patients with inflammatory breast cancer who underwent immediate or delayed reconstruction. They reported similar rates of locoregional recurrence (29 % vs. 33 %, p not significant), suggesting no compromised oncologic outcome with immediate reconstruction. Another small series found no overall survival difference in patients who underwent immediate reconstruction, although six of ten patients did develop local recurrence [32]. Importantly, these small studies do not offer sufficient statistical power to conclusively demonstrate the safety of immediate breast reconstruction for patients with inflammatory breast cancer.

In conclusion, for patients who will likely require PMRT, immediate reconstruction remains controversial, owing to concerns of compromised radiotherapy delivery and impaired cosmetic outcome of the reconstructed breast. However, many authors have reported acceptable cosmetic outcomes and comparable rates of locoregional recurrence with immediate reconstruction. Immediate reconstruction is not recommended in patients with inflammatory breast cancer.

When selecting a candidate for NSM, one must consider the risk of cancer involvement of the NAC, and the size and degree of ptosis of the breast [37]. Candidates for NSM include patients undergoing risk-reducing mastectomy. Patients may pursue risk-reducing mastectomy because of high-risk factors such as a strong family history, the presence or history of a contralateral breast tumor, lobular carcinoma in situ, or previous radiation for Hodgkin lymphoma [38]. Selected patients with ductal carcinoma in situ (DCIS) or invasive breast cancer may also be candidates for NSM [38]. In appropriately selected patients, only 12 % will have tumor involvement at the NAC, precluding preservation [39, 40].

The factors associated with nipple involvement include tumors larger than 2–4 cm, a tumor–nipple distance of less than 2 cm, breast tumors overlapping more than one quadrant, grade 3 or undifferentiated cancers, stage III disease, human epidermal growth factor receptor 2 (HER2)/neu positivity, and an extensive intraductal component of greater than 25 % [41–43].

For patients with invasive cancer, small tumors located in the periphery of the breast have the lowest risk of NAC involvement. The lowest risk of NAC involvement occurs in tumors smaller than 2 cm, located at least 2.5 cm from the NAC [44]. Tumors located within 2 cm of the NAC, or larger than 4 cm, were found in one report to have occult tumor present at the nipple in 50 % of cases [44]. A pathologic analysis of 140 mastectomy specimens reported a 16 % rate of NAC involvement with cancer. In all cases, the primary tumor was located within 2.5 cm of the NAC [45].

Many series of carefully selected patients have reported low rates of NAC involvement, ranging from 6 to 10 % [37, 38, 46–49]. In one series of patients with peripheral tumors and clinically node-negative disease, a low rate (less than 2 %) of NAC involvement was reported [48]. Therefore, the risk of NAC involvement is lower in patients with low-grade, unicentric, small, peripheral tumors, with clinically uninvolved axillary lymph nodes, who have not undergone neoadjuvant chemotherapy [39, 48, 50, 51]. Patients who will likely undergo radiotherapy are not ideal candidates, as they have advanced disease that portends a higher probability of NAC involvement. Furthermore, radiotherapy may result in distortion and asymmetric displacement of the NAC. A proposed algorithm for patient selection is illustrated in Fig. 7.2.

Identification of NAC tumor involvement precludes NAC preservation. Intraoperative pathologic assessment with frozen section of the retroareolar ducts can be useful to identify the presence of NAC tumor involvement at the initial surgery [39, 42, 52]. Dissection of the retroaerolar ducts should be done sharply, as cautery can cause thermal damage to the NAC [52]. Coring of the nipple ducts may be facilitated by everting the nipple [52].

Frozen-section analysis is 91 % sensitive and 99 % specific for assessing tumor involvement of the NAC [53]. Reported rates of positive frozen section range from 2.5 to 12 % in well-selected patients [36, 39, 54, 55]. With careful patient selection and the use of preoperative MRI, Wijayanayagam et al. [56] reported a low rate of NAC involvement of 3 %. NAC tumor involvement may not be identified until final surgical pathologic analysis, necessitating NAC resection at a second surgery. When the NAC is involved with tumor, the histologic finding is usually DCIS, although atypical ductal hyperplasia and invasive breast carcinoma may also be identified [39, 43, 54, 57].

Multiple series with less than 3 years of follow-up have reported recurrence rates of 5 % or less after NSM, comparable to rates of recurrence after skin-sparing mastectomy [36, 40, 55, 58]. Voltura et al. [55] reported a 5 % recurrence rate at 24 months in patients with aggressive triple-negative tumors. Sacchini et al. [58] reported recurrences in only two of 123 patients undergoing NSM, with a median follow-up of 25 months. Recurrences did not occur at the NAC [58]. Breast cancer occurred in two patients who underwent risk-reducing mastectomies, located in peripheral locations [58]. In another series of 96 patients who underwent NSM with a median follow-up of 34 months, only one patient developed a locoregional recurrence, and two patients developed distant metastases [40].

The reported recurrence rates of longer-term studies, with follow-up of at least 3 years, range from 5 to 28 % [39, 42, 59, 60]. In a review of 112 patients who underwent NSM and had tumors located at least 2 cm from the nipple, 5 % of patients has recurrence at a mean follow-up of 59 months [42]. Recurrences occurred in the chest wall, upper breast, and inframammary fold, with only one recurrence in the NAC [42]. The location of these recurrences highlights the importance of considering the potential for elevated risk at the periphery of the breast after NSM, as access to the peripheral breast may be more difficult if a small periareolar incision is used.

Studies with long-term follow-up of patients who undergo NSM are limited, and have not definitively demonstrated the long-term oncologic safety of NSM. In a series with a follow-up of 5.5 years, Caruso et al. [59] reported a recurrence rate of 12 % in 50 patients. Recurrences occurred at the NAC in one patient, and distant metastases developed in four patients. In a prospective trial with a median follow-up of 13 years, Benediktsson and Perbeck [53] reported a high overall locoregional recurrence rate of 28 %. This may suggest that NSM is not oncologically safe in the long term, but this high rate may have been due to patient selection. Patients at high risk of recurrence were included, with tumors larger than 3 cm or multicentric disease [53]. Patients in this study who received PMRT had a local recurrence rate of 8.5 %, similar to reported rates after skin-sparing mastectomy [53].

Petit et al. [60] recently published an update of their experience with 934 patients who underwent NSM with a median follow-up of 50 months. These investigators routinely treat the NAC intraoperatively with electron intraoperative treatment if the frozen section is negative, and preserve the NAC even if final pathologic investigation reveals tumor involvement [60]. For patients with invasive ductal cancer, 3.6 % had recurrence in the breast at 5 years, and 0.8 % had recurrence at the NAC [60]. Of the patients who had recurrence at the NAC, most had an extensive intraductal component and had HER2/neu positivity [60]. For patients with DCIS, the rate of locoregional recurrence at 5 years was high: 8 % [60]. The rate of recurrence was 4.9 % in the breast and 2.9 % at the NAC [60]. These high recurrence rates may cause one to pause before offering this procedure to patients with DCIS. Predictors of breast recurrence among patients with DCIS included age under 40 years, positive retroareolar margins, estrogen receptor negativity, progesterone receptor negativity, high-grade histologic findings, HER2/neu positivity, and Ki-67 index greater than 20 % [60].

In conclusion, several studies support the short-term oncologic safety of NSM, with locoregional recurrence rates similar to those of skin-sparing mastectomy, and rare recurrences occurring at the NAC. However, the long-term oncologic safety of this procedure has not been determined, and the recent data of Petit et al. [60] may be a reason for caution in patients with DCIS. More studies with longer-term follow-up are needed, as the literature to date is not yet definitive on the oncologic safety of NSM in the long term.