Breast cancer treatment has evolved considerably in the past decades. Nipple-sparing mastectomies are associated with important benefits in terms of patient satisfaction and psychological adjustment to the trauma of mastectomy.1,2,3
The ideal candidates for nipple-sparing mastectomy include women with small to moderate-sized breasts, no ptosis, or grade I ptosis.4 Total breast reconstruction using only fat transfer consists of liposuction, fat processing, and injection to the recipient site. It is a validated technique that we have previously applied successfully for both irradiated and nonirradiated patients.5 One of the most heavily studied limitations to this technique is the lack of engraftment with an unpredictable reabsorption rate of infiltrated fat tissue in the recipient areas.6 Many authors describe alternative fat processing techniques and procedures as well as other tips and tricks to increase survival rate of adipocytes.7,8 Evidence is poor, however, and to the best of our knowledge, there are no unified standards or definitive results that would suggest the use of any of the said methods.9 The aim of this study was to report long-term clinical and aesthetic results in the application of our systematic fat transfer approach and to extend its scope to a new subgroup of patients with history of smoking.
PATIENTS AND METHODS
Between January of 2008 and December of 2019, we prospectively enrolled 79 patients (117 breasts) who received total breast reconstruction with fat transfer after nipple-sparing mastectomy. Inclusion criteria were women with small (≤400 cc) to moderate (401 to 600 cc) breasts, with indication for nipple-sparing mastectomy. Exclusion criteria were contraindications to breast reconstruction with free tissue transfer (i.e., previous abdominal surgical procedures, low body mass index, and insufficient volume at donor area), patients excluding implant-based reconstruction, comorbidities (e.g., diabetes and other vascular diseases that could negatively affect fat engraftment), and local or distant recurrences of oncologic disease. Any patient with mastectomy skin flap complications following nipple-sparing mastectomy was also excluded from the study. All patients underwent the same standardized protocol for the nipple-sparing mastectomy consisting of an infralateral incision to the inframammary fold, specifically from the 3 o’clock position to the 6 o’clock position for the left breast and from the 6 o’clock position to the 9 o’clock position for the right. Breasts were stratified into three groups: group A had 55 patients who underwent radiotherapy, group B had 14 patients who had a history of smoking, and group C was the control group and comprised 55 patients. All patients in the smokers category, group B, were active smokers in the last year before surgery; patients who had stopped smoking for over 1 year preoperatively were excluded from this category.
All patients were evaluated by a multidisciplinary team (i.e., radiologist, plastic surgeon, and breast surgeon) before beginning the fat transfer procedure and were monitored from the first fat transfer session every 6 months with ultrasonography and clinical examination each year at the end of the treatment period. Contralateral mammography was also performed annually on patients who underwent unilateral mastectomy. Age, body mass index, mastectomy weight, number of fat injection sessions, volume injected in each session, overall injected volume, time of follow-up, and incidence of complications were collected from any patient. Continuous and categorical variables per breast were analyzed with the Kruskal-Wallis test, the p value was corrected with the Bonferroni method, and posthoc pairwise comparisons were performed to determine the detailed differences across the groups. A value of p less than 0.05 was considered statistically significant.
Fat Transfer Protocol
As described in our previously published study, we began fat transfer sessions in nonirradiated patients within 6 months from the nipple-sparing mastectomy.8 Conversely, breast reconstruction efforts in irradiated patients were specifically delayed for at least 6 months after the last session of radiotherapy to allow tissues to heal properly. We decided to approach fat transfer for breast reconstruction in group B similarly to the way we approached patients in group C. The interval we chose to separate every fat transfer session was 3 months. The same standardized fat transfer protocol for the achievement of breast reconstruction was then used for all the nipple-sparing mastectomy patients included in the study (Fig. 1). We used 2.4-mm cannulas for fat harvesting and 1.2-mm cannulas for infiltration of the harvested fat after being processed using the Coleman lipostructuring technique.10 The amount of harvested fat was decided based on the weight of the resected breast tissue during mastectomy. We specifically injected a third of the original weight during the first fat transfer session. From the second session onward, we also implemented an additional 30 percent to the previous volume to take into account the fat reabsorption rate, as reported in literature.9
Some noteworthy improvements were added throughout the years to the technique described in our previous study.8 To avoid the creation of new scars, we decided to use the old inframammary fold scar used for the nipple-sparing mastectomy to place two of the required incisions and an additional incision on the upper border of the nipple-areola complex (Fig. 2). Moreover, we developed a so-called simplified lipostructure technique for fat tissue injection.11 It consists in the use of a three-way stopcock connector as a device to pass the fat tissue from a 10-cc syringe to a 1-cc syringe. All the above-mentioned steps were carried out by two separate teams working simultaneously: the first performing the liposuction and the second performing the actual infiltration of the harvested fat tissue in the breast area.
Analysis was performed for each patient clinically by preoperative and postoperative digital photographs with frontal, lateral, and bilateral oblique views to grade aesthetic results (6 months after completion of treatment), compared with the opposite native or reconstructed breast by a separate blinded plastic surgeon team (two independent observers). Bilateral cases with one breast that received radiotherapy and one that did not were considered as part of the radiotherapy group for aesthetic analysis of global aesthetic score. The grading scale used was a modification of the one originally described by Garbay et al.12 and adapted by Carlson et al.13 Volume, shape, placement of the breast mound, inframammary fold, skin texture, and scar location were scored ranging from 0 to 2 and averaged for tabulation and analysis (Table 1); finally, the subscale averages were totaled to give an overall aesthetic score. A global aesthetic score, as described by Harris et al., was also assessed by classifying each case into one of four categories: excellent, good, fair, and poor (Table 2).14 The Kruskal-Wallis test was used to analyze the categorical variables considering a value of p less than 0.05 statistically significant. The p value was corrected with the Bonferroni method. Interrater reliability was calculated for each of the picture grading subscales and the total sum of the subscales using the Cohen Kappa.15 The strength of agreement was classified based on the j score as follows: less than 0.00, poor; 0.00 to 0.20, slight; 0.21 to 0.40, fair; 0.41 to 0.60, moderate; 0.61 to 0.80, substantial; and 0.81 to 1.00, almost perfect agreement.16 All statistical analyses were performed with SPSS version 24.0 software (IBM Corp., Armonk, N.Y.).
Subscale Analysis of Breast Reconstruction
|Subscale||Category 0||Category 1||Category 2|
|Volume of breast mound||Marked discrepancy relative to contralateral side||Mild discrepancy relative to contralateral side||Symmetrical volume|
|Shape of breast mound||Marked shape deformity||Mild shape deformity||Natural or symmetrical shape|
|Placement of breast mound||Marked displacement||Mild displacement||Symmetrical and aesthetic placement|
|IMF||Poorly defined/not identified||Defined but with asymmetry or lack of medial definition||Defined and symmetrical|
|Skin texture||Marked discrepancy relative to contralateral side||Mild discrepancy relative to contralateral side||Natural texture|
|Scar location||More noticeable||Less noticeable||Well-hidden|
Global Aesthetic Score Described by Harris et al.14
|Category 4||Category 3||Category 2||Category 1|
|Excellent: treated breast nearly identical to opposite breast||Good: treated breast only slightly different from opposite breast||Fair: treated breast clearly different from opposite breast but not seriously distorted||Poor: treated breast severely distorted|
The patient cohort included 79 patients, of whom 41 underwent monolateral breast reconstruction and 38 underwent bilateral breast reconstruction, for a total of 117 operated breasts (Table 3). Groups A, B, and C were homogenous (p > 0.05) for demographic characteristics (Table 4). Average operative time of surgery was 1.36 hours (range, 0.30 to 1.50 hours). The average time from the last fat transfer procedure to the last follow-up was 86.2 months for radiotherapy patients (range, 15.3 to 144.5 months), 45.9 months for patients with smoking history (range, 16.4 to 89.1 months), and 58.5 months for control group (range, 14.6 to 118.5 months). Total fat transfer volume (p = 0.002), fat transfer volume of first and second sessions (p = 0.003), and number of sessions (p < 0.001) showed statistically significant differences among groups. During pairwise analysis, this statistical significance was confirmed between radiotherapy versus control and smoking group but not between smoking versus control group. To be more specific, mean total volume of injected fat tissue was 578.3 cc for group A (range, 230 to 1015 cc), 450.3 cc for group B (range, 280 to 970 cc), and 413.6 cc for group C (range, 220 to 690 cc). Mean total volume of first and second sessions was 186.1 cc for group A (range, 90 to 275 cc), 243.8 cc for group B (range, 160 to 390 cc), and 235.5 cc for group C (range, 130 to 380 cc). Mean number of fat transfer sessions was 5.1 for group A (range, 3 to 8), 3.7 for group B (range, 3 to 6), and 3.4 for group C (range, 3 to 6) (Table 5). We report one recipient-site infectious complication in a patient from the radiotherapy patient group, which required prolonged antibiotic treatment and two additional fat transfer sessions to recover the volume loss caused by the complication. Moreover, we report two cases (one in the radiotherapy patient group and one in the smoking history group) of development of a superficial oil cyst that were treated by aspiration. No donor-site complications were observed in any of the patients. We also encountered the same difficulties as we did in our previous study in patients exposed to radiotherapy, which hindered the amount of fat tissue we could inject, especially during the first two fat transfer sessions. Volume (1.65 versus 1.86 versus 1.95, p = 0.064), shape (1.52 versus 1.93 versus 1.87, p = 0.114), position of the breast mound (1.65 versus 1.93 versus 1.93, p = 0.068), inframammary fold (1.61 versus 1.93 versus 1.87, p = 0.153), and scar location (1.78 versus 1.71 versus 1.69, p = 0.503) subscales obtained high score evaluation without significant differences among the groups, whereas the skin texture subscale showed lower score evaluations in group A and group B than in group C (1.0 versus 1.93 versus 1.98, p = 0.003). Total subscale showed a statistically significant difference in favor of group B and group C (9.22 versus 11.57 versus 11.35, p = 0.004), whereas the difference for the global score was not significant among the groups (3.22 versus 3.93 versus 3.93, p = 0.145) (Table 6). During pairwise analysis, the skin texture subscale and total score were confirmed statistically significant different between radiotherapy versus control and smoking group but not between smoking versus control group. Figure 3 shows photographs from a patient in group A, Figure 4 shows those of a patient from group B, and Figure 5 shows photographs from a patient in group C as a series through their multiple rounds of fat transfer.
79 Patients and 117 Breasts Distributed among Groups*
|Group A (radiotherapy)||13||13||3 + 6||12|
|Group B (smokers)||7||7||7||14|
|Group C (control group)||21||21||22 + 6||50|
*The six patients in bold refer to patients who had one breast that underwent radiotherapy and one that did not.
|Group||Age (yr), Median (IQRs)||BMI (kg/m2), Median (IQRs)||Mastectomy Weight (g), Median (IQRs)|
|Radiotherapy (A)||48 (39.5–49)||22.9 (21.2–23.5)||255 (165–390)|
|Smoking (B)||42 (39–47)||23 (21.1–23.8)||200 (155–257.5)|
|Control (C)||42 (36–48)||23.1 (21.2–24.1)||240 (165–320)|
BMI, body mass index; IQR, interquartile range.
Patients’ Operative Data
|Group||Sessions, Median (IQR)||Total Fat Volume (cc), Median (IQR)||1 and 2 Fat Volume (cc), Median (IQR)||Follow-Up (mo), Median (IQR)|
|Radiotherapy (A)||5 (4–6)||580 (390–702)||165 (140–247.5)||69.7 (30.6–140.5)|
|Smoking (B)||3 (3–4)||385 (343–510)||235 (210–257.5)||39.4 (25.5–70.4)|
|Control (C)||3 (3–4)||390 (350–470)||230 (220–260)||40.5 (26.6–73)|
IQR, interquartile range.
Mean Aesthetic Scores and p Value
|Group||Volume||Shape||Placement of the Breast Mound||IMF||Skin Texture||Scar Location||Total Score||Global Score|
IMF, inframammary fold.
Although it is a validated technique, there are still concerns regarding the use of fat transfer for breast reconstruction, namely, its safety, postoperative complications, and cancer recurrence. There are many studies analyzing oncologic safety of fat transfer in breast reconstruction, none of which could demonstrate causality between fat transfer and cancer recurrence.17,18 We followed international guidelines with regard to postoperative imaging, and the cases of recurrences were zero. Fat transfer infiltration techniques have evolved since its presentation to deal with its postoperative complications, such as oil cysts, calcifications and their interference with radiological imaging. Nevertheless, in their study, Rubin et al. demonstrated that when compared with reduction mammaplasty, fat transfer produces fewer radiographic abnormalities with a more favorable breast imaging reporting and less aggressive follow-up recommendations by breast radiologists.19
Del Vecchio and Del Vecchio defined the graft-to-capacity ratio as the volume of grafted fat in relation to the volume of the recipient site; it seems to be a relevant variable in percentage volume maintenance outcomes and may be useful in preoperative volumetric planning.20 Some authors support the use of recipient-site external expansion in fat grafting procedures using devices such as the BRAVA system (Brava LLC, Miami, Fla.),21,22 which nowadays find a role not only in cosmetic procedures but also in reconstructive surgical procedures, including breast reconstruction, but we were not able to get enough compliance from our patients to use these types of devices.
Our indications for fat transfer reconstruction after nipple-sparing mastectomy come from a combination of a small enough breast size with a sufficient adipose tissue donor area to begin breast reconstruction. Particularly in slender patients, to avoid any postoperative irregularity at the donor area, we prefer harvesting smaller quantities of fat tissue from various sites during the same operative setting by performing a dry liposuction.
Because breast reconstruction occurs over the course of a year in various steps, weight gain may occur due to hormonal therapy or improved psychologic well-being, which can be particularly helpful to both the recipient and the donor areas.
Other authors have attempted to evaluate the effect of cannula size on harvested adipose cell viability.9,23,24,25,26 This topic has been the object of debate, and consensus is yet to be reached.26 We believe that there still is little evidence supporting the use of larger cannulas for adipose tissue injection, and the size of the needles does not seem to affect cell viability, at least when using 14-gauge, 16-gauge, and 20-gauge needles.9,25 The cannulas we use are considered closest in diameter to 16-gauge needles (1.19 mm) and sufficient for harvesting larger fat particles used for breast reconstruction.27 We find that injecting smaller volumes of adipose tissues (“microfractions,” as described by Coleman) through 1-mm cannulas creates microchannels that might facilitate engraftment of adipose tissue in the recipient site. This maximizes surface area contact between adipocytes and surrounding tissues.28 Regarding fat injection, we use an infiltration protocol creating only three access points to infiltrate the whole area. When injecting, we change length of cannulas and directions by intersecting the previously created tunnels and distribute the fat in multiple levels and in a three-dimensional fashion. This technique was developed with the goal of both reducing the number of new scars and obtaining a homogenous distribution of adipose tissue across all quadrants (Fig. 2). In addition, the simplified lipostructure technique helps us to inject adipose tissue faster and with a more precise delivery, possibly improving engraftment and reducing risk of local asymmetries, in accordance with structural fat grafting principles.11,25 None of the patients from the population in our study who underwent this specific reconstruction technique complained about donor site defects.
In our previous study, we reported that fat reabsorption rate was higher in patients who underwent radiotherapy.5 In the current study, we decided to increase our sample size and, although certain guidelines discourage performing fat transfer in smokers, we extended our protocol to include this group of patients.29,30 Özalp and Çakmakoğlu analyzed in vivo the effects of smoking on facial fat transfer surgery and concluded that cigarette smoking causes low fat survival rates and impairs the improvement of skin quality.31 We observed that smoking negatively influences fat graft survival; therefore, more sessions are needed for this group of patients too, but the extent of its negative impact is lesser than that for radiation exposure, and we consider this knowledge when delivering informed consent. The fact that the smoker group contained only a small sample population of 14 patients (for a total of 21 breasts) is a limitation of our study.
We found that the differences in the number of needed sessions (p < 0.001) and the total required fat transfer volume (p = 0.002) were both significantly in favor of the control group, indicating that both irradiated patients and smokers have a lower engraftment ratio. Mean total volume of injected fat was 578.3 cc in the radiotherapy group, 450.3 cc in smokers, and 413.6 cc in the control group. An aesthetic evaluation of all patients was also performed along with the analysis of the clinical results. The total subscale score showed a statistically significant difference (p = 0.004) among groups, whereas the difference was not significant for the global score (p = 0.145). All subscales obtained high score evaluations in both groups (p > 0.05), whereas the skin texture subscale showed lower score evaluations in groups A and B than in group C (p < 0.003).
We also confirmed another conclusion drawn from the previous study, which is that radiotherapy (group A) negatively affects the skin envelope quality, therefore reducing its elasticity, especially during the first two fat transfer sessions and affecting initial volume of fat transfer (p = 0.003) compared to groups B and C. Irradiated patients unsurprisingly sustained more damage to the mastectomy skin, causing texture to be classified as category 0 or 1. Mean aesthetic score for skin texture was found to be 1.00 in radiotherapy patients (p < 0.003). These results are expectedly lower than those in the smoking group (1.93) and control group (1.98) because of the known damage caused by irradiation to soft tissues. Nevertheless, we realized that these findings were only valid for the first two fat transfer sessions; in fact, from the third fat transfer session onward, even in radiotherapy patients, skin quality improved, allowing for the injection of around a third of the original mastectomy weight. Several authors attribute the regenerative properties of fat transfer on irradiated tissues to the elevated content of adipose-derived stem cells in the injected fat tissue32; their potential in reversing damage caused by irradiation should be explored, perhaps by encouraging an earlier approach to fat transfer in irradiated patients, shorter than 6 months.23 However, apart from the need for more evidence before such a decision can be substantiated, we decided to keep the protocol begun long before 2015 in patients who have undergone radiotherapy, not only to allow the situation to settle and leave time for the irradiated tissues to heal before beginning fat transfer but also to avoid introducing further bias in the method.
Skin envelope quality and elasticity are reduced by irradiation and can lead to nipple malposition and poor aesthetic outcome. Those situations are usually challenging and are addressed as case-by-case situations. We implement fat transfer with scar release “Rigottomies” during the same operative settings, in an attempt to preferentially correct specific areas with the most fibrosis and irradiation damage in order to restore proper nipple position and symmetry. This approach, however, requires careful personalized preoperative planning and it usually takes several fat transfer sessions before proper symmetry can be recovered.
We systematically used the same incision lines in nipple-sparing mastectomy patients, by placing the preoperative markings in the inframammary fold, between the 6-o’clock position and the 9-o’clock position (to the right side) and between the 6-o’clock position and the 3-o’clock position (to the left side). Compared to the classic inframammary fold incision, we slightly tilt the incision laterally to allow the general surgeon to perform the mastectomy and the lymphadenectomy within the same operative access, avoiding additional incisions to the axilla and without impairing the plastic surgeon for the reconstruction. Having the incision laterally on the inframammary fold allows us to better hide the scar by making it less noticeable, and it helps avoid skin breast envelope retractions.33 To evaluate the reconstructed breasts, we used a subscale analysis of postprocedure digital photographs. All subscales, total scores, and global scores had substantial reliability between the observers, showing the possibility of obtaining pleasant aesthetic results using fat transfer even in irradiated patients and smokers (Fig. 3 and 4).
From our cumulative 11-year experience, we found that total breast reconstruction with fat transfer can still be achieved in nipple-sparing mastectomy patients with small to moderate breast size. Although further knowledge is required, in this prospective study, we confirm the efficacy of our defined fat transfer protocol for both irradiated and nonirradiated nipple-sparing mastectomy patients and propose to extend its indication to smokers as well, offering it with comparable clinical and aesthetic results.