Introduction

Two-stage tissue expander-based modality has been a leading option for breast reconstruction worldwide. Its popularity is attributable to several advantages including its relatively easy and uncomplicated nature, an acceptably low complication rate, and reliable and aesthetically satisfactory outcomes [1,2]. To further reduce postoperative complications and to improve final outcomes, constant efforts have been made on the technical specifications in several aspects. A traditional subject of debate over the use of Acellular Dermal Matrix (ADM) versus total muscular coverage [3,4], and a recently heated issue of prepectoral versus subpectoral placement of the tissue expander [5,6], are representative topics. The optimal intraoperative inflation volume has been also the subject of controversy. Traditionally, intraoperatively inflating the tissue expander to the tolerance of the overlying mastectomy skin flap has been considered the best way to preserve the original breast footprint and prevent skin shrinkage, leading to good aesthetic outcomes [7]. Moreover, maximum expansion has been thought to decrease the frequency of outpatient visits for inflation and the amount of postoperative expansion, which could reduce the patient discomfort. However, reconstructive surgeons have expressed their concern that a large tissue expansion volume could put pressure to the overlying mastectomy skin flap, thus compromising its perfusion, which may be already threatened after a total mastectomy [8]. This could eventually result in the development of mastectomy flap necrosis, which could lead to severe complications including reconstruction failure at worst.

Several clinical studies have investigated potential effects of intraoperative filling volume on complications rates after tissue expander-based breast reconstruction. However, the results of these studies have been inconsistent [9-12]. Influences of the intraoperative inflation volume on postoperative outcomes remains still unclear. We have similar impression that a greater intraoperative tissue expander volume could indeed increase the risk of postoperative complications including wound problem. Therefore, we have adopted a revised protocol for use in some patients, involving an intentionally reduced intraoperative tissue expansion volume, termed ‘submaximal inflation’. This study aimed to evaluate the postoperative outcomes of patients treated with ‘submaximal inflation’ relative to those of patients with the conventional inflation protocol, and to investigate the influences of intraoperative inflation volume on development of complications

Patients and Methods

Study Population

Patients who underwent two-stage tissue expander-based breast reconstruction immediately following a total mastectomy at a tertiary referral center between 2016 and 2018 were identified using a prospective database. Patients receiving neo-adjuvant chemotherapy were excluded from the study, as they could have relatively higher risks of developing postoperative complications, which could act as a confounder. Two types of tissue expanders had been used during the study period: Siltex-microtextured expanders (Mentor Worldwide, Santa Barbara, Calif.) and Biocell-macrotextured expanders (Allergan, Inc, Irvine, Calif.). The former was predominantly used in the authors’ institution. Thus, cases using the latter were excluded to reduce the potential confounding effects of tissue expander texture on postoperative outcomes. Four attending reconstructive surgeons conducted operations during the study period. Surgeon 2, 3, and 4 used the conventional intraoperative inflation protocol, in which the tissue expander was inflated as much as the overlying mastectomy skin flap permitted. The perfusion of mastectomy skin flap was assessed based on clinical signs. Those patients were assigned to Cohort 2, which served as the control group for the analyses below. Surgeon 1 (principal investigator) prospectively adopted the submaximal inflation protocol for his consecutive patients. The submaximal inflation protocol was defined as inflating the expander about 10 to 15 percent of the suggested full capacity less than the conventional protocol (Figure 1). Those patients were assigned to Cohort 1. This study obtained approval from the institutional review board of Samsung Medical Center. All patients gave informed consent to the operative procedures and postoperative managements.

Figure 2: Scheme of extent of dead space beneath the pectoralis muscle in tissue expander-based breast reconstruction. (Left) Cases with maximal inflation. Theoretically considerable amount of dead space (blue) can remain in the cranial side of the tissue expander. (Right) Cases with submaximal inflation. The upper dead space (blue) can be relatively reduced with lesser inflation of the tissue expander.

Procedures

Breast cancer patients who were evaluated as having relatively early-stage tumors during their preoperative evaluation were referred to reconstructive surgeons for immediate reconstruction. There were no specific matchings between ablative breast surgeons and reconstructive surgeons, and patients were referred to reconstructive surgeons based on the availability of operation schedule. The tissue expander insertion following total mastectomy was conducted as usual [13]. The selection of tissue expander size was mainly based on breast width and expected breast volume. Tissue expanders were inserted into the subpectoral pocket. Its lateral aspect was covered with either ADM or with serratus anterior fascia according to the attending surgeon’s preferences. Surgeon 2 rarely used ADM, while surgeon 4 used ADM for all his cases. Surgeons 1 and 3 used ADM selectively, according to intraoperative conditions including the remaining mastectomy skin flap and the extent of damage to the pectoralis major muscle and serratus anterior fascia. Other procedures were relatively standardized between the four reconstructive surgeons. Two drains were inserted during surgery: one into the subpectoral space and the other into the prepectoral space. A mild compressive dressing of fluffed gauze was then applied. Postoperatively, drains were removed when the daily output from the drains was less than 40 cc for two consecutive days. Antibiotics were administered until the drains were removed.

Outcome Measures

Patient demographics and operation-related characteristics were prospectively recorded on a database by ancillary doctors. Postoperative complications were noted by attending surgeons in their outpatient clinics and recorded in electronic medical records. The information in the database was updated monthly by ancillary doctors. Primary outcomes were the rates of complications following the 1^st stage operation (tissue expander insertion) including seroma, hematoma, infection, wound dehiscence, mastectomy flap necrosis, wound revision, reoperation, and premature removal of the tissue expander. Reoperation was defined as a return to the operating room for any reason, such as wound repair or hematoma evacuation. Secondary outcomes were postoperative inflation-related outcomes including the time to 1^st postoperative inflation, postoperative inflation frequency, mean single-visit inflation volume, postoperative inflation volume, final inflation volume, and final inflation ratio, which was defined as the ratio of the final inflation volume to the maximum capacity of tissue expander.

Statistical Analysis

The postoperative complication rates and inflation courses of Cohorts 1 and 2 were compared. We used the Pearson chi-square test or Fisher’s exact test to analyze categorical variables and the Student’s t-test or Mann-Whitney test for continuous variables. To evaluate the influence of variables on the complications, univariate and multivariate logistic regression analyses were conducted. A backward selection model was applied for the multivariate analyses. A p value of less than 0.05 was considered statistically significant. All statistical analyses were performed using SPSS version 20.0 (IBM Corporation, Armonk, NY, USA).

Results

Based on the above inclusion and exclusion criteria, a total of 734 patients representing 780 cases were included in the final analysis. There were 46 patients with bilateral reconstruction. According to the intraoperative inflation protocol, patients were categorized into two cohorts; cohort 1 (submaximal inflation protocol) containing 384 cases and cohort 2 (conventional inflation protocol) containing 396 cases. Table 1 lists the baseline characteristics of the two cohorts. The cohorts had relatively similar characteristics in terms of age, Body Mass Index (BMI), co-morbidities (diabetes and hypertension), smoking status, mastectomy types, and mastectomy specimen weights. A larger tissue expander was used in the Cohort 2, probably due to the greater weight of resected mastectomy specimens. As expected, the intraoperative inflation volume and its ratio were remarkably different between the cohorts. The average intraoperative inflation volume in Cohort 1 was approximately 50 cc smaller than in Cohort 2. The rate of ADM use was also significantly different between the two cohorts. Usually ADM use allows for a greater intraoperative tissue expansion volume,³ which was also observed in this study (Table 2). Given that, this difference in the rate of ADM use is thus a potential confounding variable that influenced both complication rates and postoperative inflation courses. Therefore, we decided the further analysis with divided by the use of ADM or not.

BMI: Body Mass Index; Wt: Weight; ADM: Acellular Dermal Matrix

Table 1: Patient demographics.

Table 2: Differences in intraoperative inflation according to the use of ADM.

Cases without ADM use

A total of 394 cases, including 298 in Cohort 1 and 96 in Cohort 2 were analyzed. The two cohorts had similar characteristics, except for the type of mastectomy. Cohort 1 had an inflation volume of 64.1 cc with a ratio of 0.17 on average, while Cohort 2 showed a volume of 104.8 cc with a ratio of 0.27, which difference was significant. Over 80 percent of patients in Cohort 1 had an intraoperative filling of less than 25 percent of the maximal capacity of tissue expander (Table 3). The time to 1^st postoperative inflation and the mean one-visit inflation volume were similar between the two cohorts. However, the postoperative inflation frequency was significantly higher in Cohort 1, as patients in Cohort 1 visited the outpatient clinic an average of one more time than patients in Cohort 2. Accordingly, the postoperative inflation volume was significantly greater in Cohort 1; however, the final inflation volume and its ratio were similar between the two cohorts (Table 4). Cohort 1 had a significantly lower rate of overall postoperative complications than Cohort 2. Especially, the rate of seroma was significantly lower in Cohort 1 relative to Cohort 2 (p = 0.003). Cohort 1 also had lower rates of delayed healing and wound dehiscence relative to Cohort 2, although this difference was not significant. Rates of other complications including infection and hematoma did not differ between the cohorts (Table 4). Multivariable analyses revealed that Cohort 1 was associated with a significantly reduced rate of overall complications relative to Cohort 2 (adjusted p = 0.001, odds ratio: 5.458, 95%, confidence intervals: 2.051 - 14.527). Age, diabetes, tissue expander size, and mastectomy type were also significantly associated with postoperative complications. Similarly, rates of seroma and wound dehiscence were significantly affected by inflation cohort after adjusting for other variables, showing a significantly reduced odds ratio in Cohort 1. Cohort 1 showed reduced rates of delayed healing and reoperation in multivariate analyses, although these associations were of borderline significance (Table 5).

Table 3: Comparison of patient- and operation-related characteristics between cohort 1 and 2 in cases not using ADM.

Table 4: Postoperative course of tissue expander inflation in cases not using ADM.

OR: Odds Ratio; CI: Confidence Intervals

Table 5: Multivariable analysis for independent predictors of complications in cases not using ADM.

Cases with ADM use

In total 386 cases using ADM were analyzed. In this subset of cases, Cohort 2 (300 cases) contained many more cases than Cohort 1 (86 cases). The two cohorts showed similar baseline characteristics except for rate of active smokers, which was significantly higher in the Cohort 1. The intraoperative inflation volume and its ratio to maximum inflation volume were also significantly lower in Cohort 1 relative to Cohort 2 (Table 6). Similar to the above analyses, the time to 1^st inflation, mean single-visit inflation volume, and the final inflation volume and its ratio were similar between cohorts. However, Cohort 1 had a significantly higher frequency of outpatient clinic visits for inflation than Cohort 2 (median: 6 vs. 5 visits, p < 0.001). The overall complication rate was not significantly different between the two cohorts. However, Cohort 1 had a significantly lower rate of seroma and wound revision than Cohort 2. Rates of wound-related complications tended to be lower in Cohort 1, though the differences were not significant (Table 7). Multivariate analyses showed Cohort 1 had significantly reduced odds of developing seroma and wound revision compared to Cohort 2. Inflation protocol did not influence the rate of other complications.

Table 6: Comparison of patient- and operation-related characteristics between cohort 1 and 2 in cases using ADM.

N/A: Not Applicable

Table 7: Comparison of complication profiles between cohort 1 and 2 in cases using ADM.

Discussion

With a long-standing, but still unproven presumption that a greater intraoperative inflation volume may increase the rate of postoperative complications, this study examined the effects of the submaximal inflation protocol in a prospective cohort of immediate tissue expander-based breast reconstruction. We also compared their complication profiles and postoperative inflation courses with those of patients treated with the conventional inflation protocol to evaluate whether the inflation protocol could influence outcomes independently. Intraoperative inflation volume is largely influenced by many factors including breast size, redundancy of remnant mastectomy skin flap, use of ADM, and other such factors. Most anecdotal studies have compared the intraoperative inflation volumes of patients who developed complications relative to patients who did not [9-12]. With a case-control design of these studies, it would be difficult to adjust potential interaction between those factors and related confounding effects. In this study, we used a submaximal inflation protocol for one cohort that consisted of consecutive patients treated by one surgeon. This design allowed us to analyze the independent influence of intraoperative tissue expander inflation volume in a less biased manner. As the two cohorts in our study were treated by different surgeons, inter-surgeon variability could bias several aspects including operative techniques and patient characteristics. However, procedures of tissue expander insertion are relatively less complicated and could be conducted consistently with low diversity. In fact, generally homogeneous surgical techniques and postoperative management protocol had been adopted regardless of surgeons. In addition, there was no specific matching between breast surgeons and plastic surgeons as mentioned above. As a result, the two cohorts had generally similar baseline characteristics, lessening the likelihood of bias. One thing that was remarkably different between the cohorts was the rate of ADM usage. This can influence not only the patient tolerance of the intraoperative inflation protocol, but also the postoperative outcomes. We therefore conducted further analyses stratified by ADM use to reduce its potential confounding effects.

Whether the submaximal protocol could be associated with reduced risks for wound-related problems was of great interest in the beginning of the current study. We found that rates of delayed healing, including wound dehiscence, were lower in Cohort 1 compared to Cohort 2 independently of ADM usage. Furthermore, these associations remained significant after adjusting for other variables on multivariate analyses. Cohort 1 had significantly reduced odds for wound dehiscence in cases with not using ADM and those for wound revision in cases using ADM than Cohort 2. Considering that wound dehiscence and situations requiring wound revision are usually closely related to compromised perfusion of the mastectomy skin flap, our results suggest that intraoperative submaximal tissue expander inflation might promote a favorable environment for wound healing by allowing proper perfusion of the mastectomy skin flap. While maximal inflation based on clinical decision of experienced surgeons may not usually compromise perfusion of the mastectomy skin flap itself, it might be burdensome for successful wound healing to some extent especially in cases with marginal mastectomy skin flap perfusion.

We have observed a significantly reduced rate of seroma formation in Cohort 1 relative to Cohort 2 after adjusting for other variables regardless of ADM usage. This result was unexpected because canonically the intraoperative maximal tissue expander inflation was thought to reduce dead space beneath the mastectomy skin flap, thereby decreasing the drainage period and risk of seroma formation. In this study, in cases using ADM, Cohort 1 showed a longer drainage period than Cohort 2, while in cases not using ADM, the two cohorts had similar drainage periods. When conducting maximal inflation, the lower part of the pocket is well occupied by the inflated tissue expander, however the upper space beneath the pectoralis major muscle remains vacant. Paradoxically, this upper dead space can enlarge as the tissue expander inflation is increased. In cases with submaximal inflation, the upper dead space could be minimized theoretically because the tissue expander is less inflated (Figure 2). The potential dead space in the subcutaneous pocket caused by redundancy in the mastectomy skin flap during submaximal inflation can be reduced with negative pressure during drainage. Also development of postoperative seroma is known to be multifaceted. Delayed healing may result in sustained fluid collection after drainage removal. As suggested above, the submaximal inflation may accelerate wound healing by not compromising perfusion of the mastectomy skin flap, which can result in lower postoperative seroma rates. Further investigations are required to verify this result.

Figure 1: Intraoperative appearances of representative cases of Cohort 1 with submaximal inflation (Left) and Cohort 2 with the conventional inflation protocol (Right).

In the present study, the submaximal inflation cohort visited the outpatient clinic for inflation more frequently, averaging one more time compared to the conventional protocol cohort. This result was expected and quite plausible. Notable was that mean single-visit inflation volume and final inflation volume were similar between the cohorts regardless of ADM use. Traditionally surgeons have assumed that a less expanded skin flap may shrink and restrict postoperative expansion. However, we observed the opposite in this study, suggesting these concerns may be unnecessary. Before the active application of the submaximal inflation protocol in the clinic, a cost-benefit estimate needs to be conducted. Our results suggest the submaximal inflation protocol will reduce the risks of complications relative to the conventional protocol. However, the submaximal inflation protocol requires more frequent postoperative visits for inflation, which is not a trivial issue considering patient discomfort. Given these concerns, the selective use of this protocol in cases with high complication risks might be appropriate.

The present study has a number of limitations. As mentioned above, composition of reconstructive surgeons were heterogenous between the cohorts, possibly acting as a confounding variable. To resolve this issue, a prospective randomized controlled study involving multiple surgeons would be necessary. Although nearly 800 cases were analyzed in this study, the sample size was still not large enough to conduct a detailed subgroup analysis. Larger multicenter studies are necessary to conduct such analyses. Our study population also has a relatively lower BMI and lower rate of co-morbidities compared to other Western population. This may make it difficult to generalize our results. Finally, we did not evaluate final reconstructive outcomes after the 2^nd stage operation, which is one of the main limitations of this study. We have had impression of no remarkable differences between the cohorts in this regard. However, further detailed assessment on results of the 2^nd stage operation with more objective evaluation tools are necessary. Those outcomes can be influenced by multiple factors, and so well-designed controlled studies are essential.

Conclusions

Our results suggest that the submaximal intraoperative tissue expander inflation could be a safe protocol, resulting in a lower rate of postoperative complications (including seroma and wound-related complications) compared to the conventional maximal inflation protocol. It could allow for sufficient postoperative expansion, although more frequent visit for inflation are necessary. The submaximal inflation protocol could thus provide a reliable foothold upon which to conduct the 2^nd stage operation.