Long-term impact of lipofilling in hybrid breast reconstruction: retrospective analysis of two cohorts

Lipofilling has recently gained popularity as a tool in primary treatment of breast cancer, and its association with two-stage implant breast reconstruction is considered as standard treatment in many centers. However, no data are available about the long-term results of the association of lipofilling in combination with expander-implant reconstruction. A retrospective analysis was conducted on patients treated between January 2010 and December 2014. Two groups were compared. Group 1 had a standard expander-implant two-stage reconstruction. Group 2 underwent hybrid breast reconstruction (HBR). Patient characteristics, hospitalization, outcomes, reoperation details, outpatient visits, and evaluation questionnaires were taken into consideration. Intergroup comparison was performed using Wilcoxon Mann-Whitney U test and Pearson’s chi-square test or Fisher’s exact test for categorical variables. Two hundred fourteen patients were evaluated: 130 patients in group 1 and 84 patients in group 2. Group 2 showed significant benefits over group 1 in terms of capsular contracture rate, breast pain, and displacement/rotation of the implant (p = 0.005). The HBR protocol is associated with lower rate of capsular contracture, less breast pain at long follow-up times, and lower overall rates of revision surgery compared to standard expander-implant reconstruction. A specific cost analysis will help further clarify the advantages of this protocol over a standard procedure. Level of Evidence: Level III, risk/prognostic, therapeutic study.


Introduction
Lipofilling is one of the most investigated topics in Plastic Surgery in recent decades, although the use of adipose tissue for reconstruction purposes can be traced back to studies published a century ago. Neuber and Guthrie, respectively, first described the use of adipose tissue to restore body shape after tuberculosis and to correct depressed areas after mastoid surgery [1,2]. Thanks to the introduction of liposuction described by Fisher in 1974, 20 years later, the term lipofilling is now in common use [3]; this is also due to the pivotal work by Roddi et al. on correction of progressive hemifacial atrophy [4]. Since then, lipofilling has gained widespread popularity, and its use in breast surgery has become a standard procedure [5]. Fat grafting has been proposed as a tool to improve the outcome for capsular contracture after breast implant reconstruction [6][7][8][9]. Even in breast implant augmentation, lipofilling now has a well-recognized role [10,11]. A large number of studies have investigated the technique of fat grafting, its safety in breast surgery, the variations offered by the different types of harvesting techniques, optimal donor site, device employed, elaboration of adipose tissue, recipient site preparation, and timing of surgery [12][13][14][15][16][17][18][19][20]. Many reviews are available comparing dozens of studies, analyzing every step of the lipofilling procedure, in order to find a standard method to ensure the best graft-taking rate [21][22][23].
Fat grafting leads to improved outcomes in breast implant surgery [24,25]. Its use on secondary cases can show a beneficial effect after the complication has been assessed in detail, and the introduction of further surgical steps (lipofilling) at the time of expansion can reduce unfavorable outcomes.
The aim of our study is to evaluate long-term results of a "hybrid breast reconstruction" (HBR) protocol [26], wherein fat grafting is performed during the course of the expander/ implant reconstruction. The main object of the study is to determine the rate of capsular contracture. Secondary outcomes were analysis of postoperative complications, reoperations performed or refused by the patient, and the assessment of the esthetic result.

Materials and methods
A retrospective study was conducted taking into consideration all surgical breast reconstructive procedures with two-stage expander/implant reconstruction. The population was divided into two cohorts. Group 1 includes patients who underwent expander/implant standard reconstruction and represents the control group. Group 2 includes patients who underwent the HBR protocol. Criteria for exclusion were chronic liver disease, coagulopathies, and/or anticoagulant therapy, patients who underwent mammary reconstruction after trauma or congenital disease, and patients lost to follow-up.
All procedures performed were in accordance with the ethical standards of the Institutional Committee and with the 1964 Helsinki declaration and its later amendments. Written informed consent was obtained from all patients. This study is in line with the STROCSS criteria [27].

Patients and methods
A retrospective chart review was performed for 84 consecutive postmastectomy patients who underwent secondary breast reconstruction from January 2010 to December  f-i Postoperative result after 24 months from the second surgical step of expander/implant exchange procedure. The patient had one lipofilling session during the latter surgery 2014 at our institute. Our study population (group 2) was aged between 35 and 65 years old, who underwent the same surgical procedure (unilateral mastectomy, immediate reconstruction with a tissue expander, and replacement with definitive implant), with similar follow-up and the same type of therapeutic approach in terms of complementary therapies.
After unilateral mastectomy and immediate reconstruction with an expander, these patients were submitted to one to three serial deflation-lipofilling procedures. Positioning of a definitive prosthesis was performed during the last lipofilling session. Patients were assessed for the thickness of mastectomy flaps and the presence of radiotherapy. For mastectomy skin flap thickness < 0.5 cm and/or those subjected to radiotherapy, patients were submitted to two lipofilling sessions before replacement of the expander. Smoking habit was not taken into account for patient selection.
For this study, a group of 130 subjects who underwent the standard expander/implant reconstruction was selected as the control population (group 1).
All patients underwent surgery by the senior author of this study (P.C.P). Patients were examined on postoperative days 3, 7, and 14, and follow-up visits were scheduled at 1 month and every 6 months postoperatively. Clinical evaluation of the reconstruction and patient satisfaction were recorded. Assessment of capsular contracture was based on Baker's modified classification [28].

Preoperative planning and HBR protocol
The first surgical step was the same for all patients as they underwent mastectomy and reconstruction with a breast tissue expander. Mastectomy skin flap thickness was determined at 3 months postoperative at the end of the outpatient expansion procedures by a pinch test (by a skinfold caliper) and ultrasound [29]. Five regions of the breast were identified with a template, and measurements were recorded in the same area in all participants with both measurement devices (see Fig. 1).

Pinch caliper measurements
In the middle box of each template, a skin caliper (Gima, Gessate, Italy) was used to measure subcutaneous fat thickness. One measurement for each was taken in each middle box, for a total of 5 measurements. The clinician (S.C.) taking the skinfold caliper measurement was right-hand dominant, so with her left thumb and forefinger she lifted the skin away from the muscle and pinched it together. With her right hand, she placed the jaws of the skinfold caliper inferior to the pinch and completely released her thumb from the handle. The reading was recorded after the first rapid fall on the dial.
The fat thickness recorded by the skinfold caliper was divided in half to represent the subcutaneous fat layer over the muscle.
Ultrasound imaging Ultrasound imaging (GE Heathcare Logiq F8, Waukesha, WI) was used to measure subcutaneous fat thickness in each middle box. An 8-MHz linear array was used while capturing the image in B-mode. Ultrasound gel was applied liberally to the center of the template before placing the transducer on the skin. After a clear image was identified, the image was saved. The images were labeled with the participant number and site (upper, central, left, right, below) of measurement. On a subsequent day, the clinician (S.C.) who obtained the images opened the saved images and measured subcutaneous fat thickness using the measurement tool.
All the patients who were randomly selected to have HBR protocol accepted it, and signed informed consent was obtained from all.
Depending on the skin thickness and previous radiotherapy, an average of 1 to 3 deflation-lipofilling sessions were programmed (see Fig. 2a-i). If a single lipofilling procedure was necessary, it was performed at the time of expander/ implant exchange procedure. If multiple deflation-lipofilling sessions were scheduled, they were performed after final filling of the tissue expander. The fat was positioned at the level of the mastectomy flap, between skin and peri-prosthetic capsule. Deflation was conducted to be 10 cc superior to the total amount of fat injected.
If we noticed the skin over the breast area was too stretched after the fat grafting session, we would remove a greater quantity of saline solution from the breast expander. Deflationlipofilling sessions were scheduled every 3 months.
In all patients, anatomical textured silicone gel implants were placed in the subpectoral pocket. All patients received perioperative antibiotic prophylaxis (cephazolin 2 g, intravenous injection, 30 min before the beginning of surgery).

Questionnaire
Patients were asked to fill in the postoperative BREAST-Q reconstruction questionnaire [30][31][32] to evaluate the outcomes. The BREAST-Q reconstruction module has the following scales: satisfaction with breasts, satisfaction with implants, psychosocial well-being, sexual well-being, physical well-being. We administered the questionnaire electronically 2 years after surgery.
We compared patient questionnaire scores (group 1 vs group 2) to determine if there was a significant improvement in esthetic outcomes between the groups.

Statistical analysis
Intergroup comparison was performed using a Mann-Whitney U test and Pearson's chi-square test or Fisher's exact test for categorical variables. A p < 0.05 was accepted as the level of statistical significance.
The chi-squared test is used to compare the distribution of the Breast-Q items in group 1 and 2. p < 0.01 was accepted as the level of statistical significance.
All patients in group 2 had at least one lipofilling session at the time of removal of the expander. Moreover, nine patients had a total of two lipofilling sessions and six patients had a total of three lipofilling sessions.
Thirty-four patients (26%) from group 1 underwent lipofilling as a secondary procedure. There was one session for 29 patients, two sessions for four patients, and three sessions in one patient.
The average quantity of adipose tissue infiltrated for both groups was 88.27 cc (85.26 cc for group 1 and 89.5 cc for group 2).
There was a significant difference between groups (p > 0.05) regarding the total amount of adipose tissue infiltrated: there was more fat infiltration in group 2 (Fig. 3b).
Postoperative complications (Table 2) included 6 hematomas for group 1 (p = 0.045), 7 seromas (5 for group 1 and 2 for group 2), 5 implant infections (4 for group 1 and 1 for group 2), and one case of breast prosthesis rupture (group 1). There were no significant differences between groups regarding asymmetry (p > 0.05). Incidence of capsular contracture in two groups was the most important endpoint of our study for which a significant difference between groups was found (p = 0.004). Twenty-eight patients developed capsular contracture in group 1 (21.53%) compared to 6 in group 2 (7.14%).
The difference in the onset of breast pain between groups was statistically significant (p = 0.013): it involved 18 patients in group 1 and 3 patients in group 2.
Outcomes were similar regarding the reoperation rate: as shown in Table 2, there were no significant differences between groups regarding the number of prosthesis replacements (p = 0.127) and prosthesis removals (p = 0.253).
For group 2, we considered fat grafting procedures not included in HBR protocol, but which were conducted as secondary procedures. We did not observe any significant difference between the experimental and control group for secondary fat grafting procedures (p = 0.431). Refusal of reoperation between groups was statistically significant.
To understand the difference in costs regarding the reconstructive protocol in groups 1 and 2, we assessed surgical time in the operating room, number of hospitalization days, number of interventions required for each patient, and number of outpatient department visits.
With regard to surgical time, the hybrid breast reconstruction was longer than the surgical procedure in the control population (p < 0.0001). This indicates that HBR was more expensive compared with the replacement of the breast expander with a permanent implant standard procedure (Fig. 3c).
However, there were no significant differences between groups regarding the number of definitive prostheses and breast tissue expanders used, number of admissions to   (Table 3). Eighty-six patients of group 1 and sixty-four of group 2 completed BREAST-Q surveys, with a response rate of 66% for group 1 and 78% for group 2.
The answers of the patients to the BREAST-Q and the statistical analysis are shown in Table 4 and Fig. 4a, b. After further analysis of the questionnaire, we observed that patients in group 2 obtained significantly better postoperative results than patients from group 1 (control) regarding the following items: the reconstructed breast softness, symmetry (breasts of equal size relative to the other), reconstructed breast look and touch, amount of implant rippling perceived by the patients, and physical well-being about chest and upper body.

Discussion
A number of studies have been carried out on the introduction of lipofilling as a preliminary step in breast reconstruction for patients undergoing mastectomy and radiotherapy [33,34]. Several authors have proposed the use of lipofilling on expanders, limiting this indication to patients who received postmastectomy radiotherapy (PMRT) [35]. In 2008, Sarfati et al. published an analysis of 28 patients who underwent multiple lipofilling sessions to prepare the thoracic region for breast implant reconstruction [36]. Panettiere et al., in 2009, were the first to propose fat grafting as a surgical tool for improving outcomes of breast implant reconstruction after radiotherapy in a case-control study. Improved esthetic and functional result were reported, together with a reduced complication rate, in the study group [36]. Between 2010 and 2012, Salgarello et al. published their LIPOBED protocol [34,37].
In 2010, Serra-Renom et al. [38] introduced two fat grafting sessions during two-stage breast implant reconstruction. No statistical analysis was conducted, but they reported a lower rate of capsular contraction.
In 2013, Ribuffo et al. published the Cagliari University Hospital (CUH) protocol for breast reconstruction after PMRT. This was a case control study with no report of complications in the study group [35,39].
The concept of lipofilling has evolved as a primary tool in breast reconstruction. Deflation-lipofilling sessions are conducted to prevent complications during the reconstructive step, limiting the onset of postoperative complications. As other authors promote the use of polyurethane implants as a first step in breast reconstruction to reduce complication rates [40,41] and not only as a rescue tool after a failure, in the same way, the present study investigated the beneficial effect of fat grafting as a part of the standard two-stage breast implant reconstruction for all postmastectomy patients in order to reduce overall capsular contracture and improve the esthetic result. This is the first report of 84 patients, a case-control study, with a mean follow-up time of 34.87 months. Statistical analysis showed that fat grafting may have a role in reducing the capsular contracture rate. One to three more surgeries per patient should be costeffective for a Plastic Surgery Department. Operative time at the moment of expander/implant exchange has been reported to be significantly longer for HBR. However, the number of hospitalizations, surgical procedures, and outpatient sessions did not show any significant differences.
The beneficial effect of deflation-lipofilling procedures, despite adjunctive costs, was evaluated by Sommeling et al. and Stillaert et al. [42,43]. The latter study included the removal of the expander at the end of the programmed reconstruction, without any implant positioning, in order to guarantee a total autologous reconstruction. The same conclusion was reached by the study by Fabiocchi et al. [44].
The study by Sommeling et al. was detailed and endoscopic imaging was used to show the capsular vascular plexus, which is related to increased survival of adipose tissue. However, that study involved a limited number of patients, without a control group and with 9 to 29 months of follow-up.
Nava et al. reported on a large series of 350 patients treated with lipofilling and implants. He reported detailed algorithms for the surgical pathway to follow regarding the patient BMI. Although in that study there is no study description, statistical analysis, or detailed results [45].
Since Rigotti's work on the effect of adipose tissue on radiated breast [46], many attempts have been made to further confirm the advantages of fat grafting procedures. Indeed, the The explanation for the lower rate of capsular contracture provided by fat grafting has not yet been answered. Won-Serk et al. underlined how adipose stem cells lead to an increase in fibroblasts and their stimulation with increased collagen production of type 1 and fibronectin. Their results showed a beneficial effect of adipose stem cell treatment on wound healing in mice and the absence of a significant number of inflammatory cells in the dermis of wounds [47]. Reduction in the number of cells with chronic inflammation cells may be one of the factors contributing to the reduced capsular contracture rate.
By using a validated survey tool (BREAST-Q), in this study, we show that patient of group 2 were significantly more satisfied with the appearance of their breast and showed an improved psychosocial, sexual, and physical well-being than patients exposed to prosthetic reconstruction alone.
As already reported by other authors [31,48,49], fat grafting provides a significant analgesic effect in patients who underwent an implant-based breast reconstruction.
The current study is not exempt from limitations and mainly represented by the relatively small size of the examined sample, short follow-up, and clinical and radiological evaluations that are highly physician-dependent, and its retrospective design.

Conclusions
The HBR protocol is associated with a lower rate of capsular contracture, less breast pain at long follow-up times, and a lower overall rate of revision surgery compared to standard expander-implant reconstruction. A specific cost analysis could further clarify the advantages of this protocol vs. a standard procedure.