Sbitany, Hani M.D.; Wang, Frederick M.D.; Saeed, Lina B.A.; Alvarado, Michael M.D.; Ewing, Cheryl A. M.D.; Esserman, Laura J. M.D., M.B.A.; Foster, Robert D. M.D.

Nipple-sparing mastectomy has been established as an oncologically safe procedure in patients with early-stage breast cancer, and as a safe risk-reducing technique in patients with a genetic predisposition to breast cancer.^1–3 More recently, studies have confirmed the safety of nipple-sparing mastectomy in patients with advanced-stage disease, showing similar locoregional recurrence rates when compared with skin-sparing mastectomy.^4–6 Our institutional preference is to routinely perform total skin-sparing mastectomy, a more oncologically thorough version of nipple-sparing mastectomy.⁷

When planning reconstruction on these patients, important considerations for choosing good aesthetic candidates include breast size and degree of ptosis.³ As experience grows with total skin-sparing mastectomy, so too do the indications for performing postmastectomy breast reconstruction on these patients.⁸ Implant-based reconstruction following total skin-sparing mastectomy and nipple-sparing mastectomy has now been shown to be safe in patients with prior circumareolar incisions.⁹,¹⁰ The effect of other prior operations, such as breast augmentation, on the safety of total skin-sparing mastectomy and immediate reconstruction has not been established to date. The improved body image and psychological adjustment that accompany preservation of the nipple-areola complex, over a completed reconstruction, are especially beneficial in this population.¹¹

We review our series of total skin-sparing mastectomy and immediate expander/implant-based breast reconstruction in patients with a history of augmentation mammaplasty. An outcomes analysis is performed and compared with that for women without a history of breast surgery undergoing total skin-sparing mastectomy with immediate expander placement. Specific attention is given to morbidity related to vascular compromise of the nipple-areola complex and mastectomy skin flaps.

PATIENTS AND METHODS

All patients treated with total skin-sparing mastectomy at the University of California, San Francisco, have been tracked prospectively since 2005, using our breast reconstruction outcomes database. From our database, we identified 34 patients who underwent total skin-sparing mastectomy and had a history of breast augmentation for either cosmetic or contralateral symmetry-producing reconstructive purposes, between 2006 and 2013 (Fig. 1). This cohort was matched to 68 patients who underwent total skin-sparing mastectomy without prior breast augmentation, over a similar period. Matching was performed based on mastectomy date, to achieve a representative control group, for comparison. All patients underwent immediate breast reconstruction with tissue expander placement performed by one of the two senior plastic surgeons (H.S., R.D.F.).

Fig. 1

Patient demographics and comorbidities, tissue expander coverage method (submuscular versus acellular dermal matrix–assisted), neoadjuvant and adjuvant oncologic treatment, expander size, intraoperative fill data, and length of expansion process were prospectively recorded. Outcome variables assessed for the two patient cohorts included hematoma, seroma, infection, wound breakdown, major/minor skin necrosis, partial/complete nipple necrosis, implant exposure, and explantation of the tissue expander.

Differences with respect to age, body mass index, smoking, diabetes, irradiation, and chemotherapy between the “prior implant” group and the “control” group were examined using the t test and Fisher’s exact test. Age and body mass index were continuous and reported as mean values. We reported the prevalence of categorical variables. We compared differences in tissue expander size and intraoperative fill volume with the t test, and these were reported as mean values. We also compared the risk of postoperative complications, which were tracked as dichotomous variables, and we reported these as whole values and percentages of total cases. Odds ratios were obtained for outcomes data, and hypotheses were analyzed with Fisher’s exact test. All statistical analyses were performed using Stata 13 (StataCorp LP, College Station, Texas). All p values were two-tailed, and a value of p < 0.05 was considered to be significant. This study, and the use of the prospectively collected breast reconstruction outcomes database, has been approved by the University of California, San Francisco Institutional Review Board.

RESULTS

We identified 51 breasts undergoing total skin-sparing mastectomy in 34 patients, with a history of prior saline or silicone breast implants placed for breast augmentation, labeled as the “history of prior augmentation mammaplasty” group. In all cases, the implant was present at the time of mastectomy, and was removed at that operation. Simultaneously, we identified 117 breasts undergoing total skin-sparing mastectomy in 68 patients, who underwent total skin-sparing mastectomy without prior breast augmentation, labeled as the “no history of prior breast implant” group. Preoperative characteristics and demographics between the two groups were similar (Table 1). Mean age and body mass index were compared with the t test. Mean body mass index was lower in the augmentation mammaplasty group, and this was statistically significant (21.6 ± 3.2 versus 24.1 ± 4.1, p = 0.004). Other comorbidities including diabetes, current and prior smoking history, irradiation history, and neoadjuvant or adjuvant chemotherapy were compared using Fisher’s exact test and were not found to be significantly different between the two groups.

Table 1

Baseline treatment characteristics were also obtained for both groups (Table 2). The mean follow-up time was 24.4 months in the nonimplant group and 24.2 months in the augmentation mammaplasty group; this similar follow-up time was expected, given that cases were matched to controls. A higher percentage of patients in the augmentation mammaplasty group underwent breast irradiation before undergoing total skin-sparing mastectomy (14 percent versus 9 percent; p = 0.404). The number of prophylactic cases, postoperative irradiation, and acellular dermal matrix–assisted cases were statistically similar between the two groups.

Table 2

All patients in both groups underwent tissue expander placement at the time of mastectomy, and expander size and volume are listed in Table 3. The mean tissue expander size used for patients in the augmentation mammaplasty group (488 ± 93 ml) was larger than the mean tissue expander size used for patients in the nonimplant group (424 ± 106 ml), and this difference was statistically significant. The mean intraoperative fill volume for tissue expanders in the augmentation mammaplasty group (26.7 ± 12.5 percent) was also larger than the intraoperative fill volume for tissue expanders in the nonimplant group (19.9 ± 10.9 percent). The overall time to completion of reconstruction was shorter in the prior implant group (151 ± 122 days) than in the no–prior implant group (187 ± 137 days; p = 0.1112), but this was not statistically significant.

Table 3

There were no significant differences in the risk of complications when compared between the two groups (Table 4), even with a greater percentage of cases with a history of irradiation in the augmentation mammaplasty group. The overall rate of serious infection, defined as patients requiring intravenous antibiotics or an unexpected return to the operating room, was 9 percent in the nonimplant group and 12 percent in the augmentation mammaplasty group (p = 0.576). The rate of delayed wound healing was 7 percent in the nonimplant group and 4 percent in the augmentation mammaplasty group (p = 0.725). The rate of major skin necrosis requiring operative débridement was 5 percent in the nonimplant group and 4 percent in the augmentation mammaplasty group (p = 1.0). The rate of tissue expander/implant exposure was 5 percent in the nonimplant group and 2 percent in the augmentation mammaplasty group (p = 0.677). The rate of tissue expander/implant explantation and loss was 6 percent in the nonimplant group and 10 percent in the augmentation mammaplasty group (p = 0.515).

Table 4

One of the main threats to validity and a drawback of this study is the issue of low power when studying rare outcomes in limited cases. The highest level of power for the postoperative outcomes in this study is estimated to be approximately 0.1, and the sample size needed to detect the effect sizes observed in this study is estimated to be greater than our total experience of total skin-sparing mastectomy. This limitation clearly decreases our ability to detect a true difference in outcomes, and we acknowledge that we may be making a type II error in this study.

DISCUSSION

As experience with nipple-sparing mastectomy and total skin-sparing mastectomy grows, the oncologic indications for offering these procedures continue to expand. We routinely offer total skin-sparing mastectomy to the majority of women without direct tumor involvement of the nipple-areola complex. This technique preserves all the external skin elements of the nipple and areola but intraoperatively involves inversion of the nipple from the underside and complete excision of all nipple tissue at the dermal junction.⁷ In this way, all parenchymal elements are removed, and only the skin of the nipple-areola complex remains, thus allowing the patient to retain their original nipple-areola complex externally. Given the more aggressive removal of nipple tissue in total skin-sparing mastectomy (relative to nipple-sparing mastectomy), this technique is offered to all women presenting to our institution with tumors not directly involving the nipple-areola complex.

Breast reconstruction in these patients offers a particular challenge when patients have a history of breast surgery. The literature has shown that total skin-sparing mastectomy and immediate expander/implant-based reconstruction is safe and effective in patients with a history of circumareolar incisions.⁹,¹⁰

Our current series shows that total skin-sparing mastectomy and immediate expander based reconstruction in patients with a history of breast augmentation offers a safety profile similar to that in patients without a surgical history of the breast. Sparing the nipple-areola complex skin does not increase the risk of morbidity, including delayed wound healing, partial or complete nipple necrosis, and ultimately expander/implant loss.

In this patient population, it is the practice of both plastic surgeons (H.S. and R.D.F) to place a submuscular tissue expander at the time of mastectomy, rather than a permanent implant. Although the preexpanded skin envelope will in some cases allow for permanent implant placement, we find that expander placement offers less skin tension and thus minimizes morbidity, such as nipple necrosis and mastectomy skin flap necrosis. Intraoperative perfusion assessment of the breast skin following completion of mastectomy, using indocyanine green and the SPY Elite System (LifeCell Corp., Branchburg, N.J.) was performed in selected cases at the surgeon’s discretion. These images confirmed that in the majority of cases, the tissue perfusion was reduced with the expander filled to a fraction of the prior implant size. Thus, this justified expander placement versus permanent implant placement.

Based on our cumulative experience, we have now developed an algorithm for assessing and treating these patients, both in terms of incision placement and plane of expander placement (Fig. 2). In patients presenting with a history of submuscular implant placement through a lateral inframammary incision (the most common scenario in our cohort), we perform both total skin-sparing mastectomy and expander placement through the same inframammary incision (Fig. 3). In the majority of these cases, the pectoralis is either disinserted (prior dual-plane breast augmentation) or thin and attenuated. In such circumstances, we perform acellular dermal matrix–assisted tissue expander coverage, with placement of the acellular dermal matrix along the lower pole of the expander as an inferolateral hammock, thus offering improved coverage of the expander in the region of the incision. In addition, this offers more precise definition of the inframammary fold, which is of particular significance in this patient population, given that the breast has been surgically manipulated multiple times, and the tissue is more likely attenuated.

Fig. 2

Fig. 3

In those patients with prior implant placement through an inframammary incision and subglandular implant placement, both the implant and capsule are completely removed, and a new subpectoral pocket is created for expander placement. In these patients, tissue expander coverage is achieved with either total submuscular coverage or acellular dermal matrix assistance, at the discretion of the surgeon.

Patients with a history of prior intraareolar or periareolar incisions for implant placement are categorized. In those who have undergone a small intraareolar incision to place a subglandular saline implant in the past, the mastectomy and expander placement is again most routinely performed through a new inframammary incision. The expander is then placed in a submuscular position, potentially with acellular dermal matrix–assisted coverage. In the rare instance in these patients where the patient has a significant “snoopy deformity” and significant descent (ptosis) of breast tissue over the implant, we will consider performing the mastectomy through a superior periareolar (crescent mastopexy) incision. The most important consideration in planning this incision is that it encompasses less than 30 percent of the entire diameter of the areola complex. Our experience has shown that this will help minimize the risk of nipple-areola necrosis.⁷

The final clinical scenario routinely encountered is patients having undergone prior augmentation/mastopexy, with a complete circumareolar incision, either alone or as part of a vertical pattern. In such patients, we plan the mastectomy and expander placement through a superior periareolar incision, again encompassing less than 30 percent of the overall circumference of the areola complex. Experience has shown in these patients that using the vertical limb of the prior pattern carries a higher rate of areolar and lower pole skin flap necrosis. These patients then undergo tissue expander placement in a submuscular position.

In our series, there was a clinical difference between the two cohorts with regard to history of premastectomy radiation therapy (9 percent for the no-implant group and 14 percent for the prior implant group; p = 0.404). Despite the higher rate of preoperative irradiation in the cohort with prior breast augmentation, there was a similar morbidity profile between the two groups. Specifically, there were similar rates of partial and complete nipple necrosis between both and no cases of superficial nipple necrosis in the augmented patients. This included patients who had undergone previous saline breast augmentation through a prior intraareolar incision.

The lack of superficial nipple necrosis in the prior augmentation cohort is potentially a result of the prior stretch placed on the nipple-areola complex in these patients from the underlying implant over time. This in essence may create a relative, tension-induced “delay phenomenon” of the nipple skin. Thus, the skin may be better able to tolerate the reduced vascularity from the mastectomy, and the greater tension from the underlying expansion volume, and the risk of isolated epidermolysis is less.

Both cohorts were similarly matched with regard to minor and major skin flap necrosis. This suggests that patients with a significant history of breast surgery can still safely undergo total skin-sparing mastectomy when properly performed. Our low rates of nipple and skin necrosis in these cases are likely a result of both our breast surgeon’s significant experience performing this procedure, and our multidisciplinary approach in tailoring the operation and the reconstructive plan for each patient with the surgical oncologists. Although skin flap necrosis was categorized into minor and major for descriptive terms, mainly to differentiate between partial-thickness and full-thickness necrosis, respectively, we do concede that any degree of skin necrosis is significant, as it threatens implant integrity.

The infection rate between the two groups was clinically different with regard to those requiring oral antibiotics (11 percent for the no-implant group and 22 percent for the prior implant group; p = 0.094) for treatment. Although not statistically significant, this was a notable clinical difference. Infections requiring intravenous antibiotics were similar between the two groups, and infections requiring operative débridement were higher in the control group. The greater rate of cellulitis in the prior augmentation group was likely attributable in part to the history of multiple operative procedures on the breast, the performance of a capsulectomy and more extensive operation in many of these patients, and the thinned-out nature of the pectoralis muscle in these patients with a history of submuscular augmentation. With the possible existence of a biofilm associated with the prior implant, and the thinned, weaker internal tissues, transmission of these organisms to the overlying skin is theoretically more likely. This likely in part accounts for the higher degree of cellulitis in these patients.

In addition, the higher rate of preoperative radiation history and intraoperative expander fill volume in this prior augmentation cohort may in some cases lead to either a true cellulitis, or either skin changes or congestion that is mistaken as cellulitis. A high level of vigilance with these patients then leads to treatment with oral antibiotics. In the end, the cases in this category were all successfully treated and did not progress further.

Ultimately, the group with prior implants did have a higher rate of reconstructive implant loss (10 percent) than the group without prior augmentation (6 percent). Although this difference was not statistically significant (p = 0.515), it was a notable clinical variation. Again, this is likely accounted for by the history of operative procedures and attenuated pectoralis major muscle, thus leading to less reliable internal vascularized coverage of the expander/implant. The higher rate of irradiation in this cohort is almost certainly a contributing factor to this as well.

Of the five cases of implant loss in the prior implant group, three occurred in patients with a history of external beam irradiation and one occurred in a patient receiving postmastectomy radiation therapy. The other patients with a history of preoperative irradiation and augmentation mammaplasty underwent successful total skin-sparing mastectomy and two-stage expander/implant reconstruction (Fig. 4).

Fig. 4

This lends merit to the fact that both a history of breast surgery and a history of irradiation will require the surgeon to proceed cautiously, as potential risk is significant. We continue to look for alterations in technique to improve reconstructive success in such patients with a history of radiation therapy, including lengthening the time before implant exchange following radiation completion.¹² In addition, in those patients with a history of radiation therapy to the breast, the tissue expander is usually filled with no volume, and at most 50 cc of saline, at the time of intraoperative placement. We find that minimal tension on the irradiated pectoralis muscle in this way, during the initial healing process, helps to reduce the risk of wound dehiscence and implant exposure. This is in contrast to those patients without a history of irradiation, who undergo expansion as much as possible at the time of placement, without producing tension on the skin flaps, to allow for maximal preservation of mastectomy skin flap position and shape, and to prevent the accumulation of redundant and loose skin with excessive underlying dead space.

Among the patients in the prior breast augmentation cohort without any irradiation, all but one experienced successful reconstructive procedures. This one patient experienced postmastectomy cellulitis and, because of a delay in follow-up, was not immediately treated. By the time this patient was assessed and treated, the infection had spread, and the tissue expander was involved and thus removed. No specific source of the cellulitis was identified.

The high success rate in this subset of patients with prior augmentation and no irradiation history again further suggests that without a history of irradiation, performing two-stage expander/implant-based breast reconstruction following total skin-sparing mastectomy in this population is safe and effective. Given these findings, we promote total skin-sparing mastectomy with immediate two-stage expander/implant-based breast reconstruction as a safe and effective technique in patients with a presenting history of breast augmentation.

REFERENCES