Comparing the Outcome of Different Biologically Derived Acellular Dermal Matrices in Implant-based Immediate Breast Reconstruction: A Meta-analysis of the Literatures.

BACKGROUND: Acellular dermal matrices (ADMs) have been used extensively in implant-based breast reconstruction. It was reported that due to the different sources and processing methods, the outcomes of ADMs in implant-based breast reconstructions are expected to differ. We designed this study to statistically analyze and discuss the outcome of 3 commonly used ADMs, Alloderm, Strattice, and Surgimend in implant-based breast reconstruction.
METHODS: Comprehensive review of the literatures searched on electronic databases was done to identify studies published between 2006 and 2017 comparing the outcome of ADMs. Pooled random effect estimates for each complication and 95% confidence interval (CI) were calculated. One-way analysis of variance and Bonferroni test were used to compare statistical significance between and within groups, respectively. Multiple linear regression was done to include confounding factors and R statistic program for forest plot.
RESULTS: Twenty-one studies met the inclusion with a total of 1,659, 999, and 912 breasts reconstructions in Alloderm, Strattice, and Surgimend, respectively. Seven complications extracted including major and minor infection, seroma, implant loss, hematoma, capsular contracture, and localized erythema. Pooled total complication rates were 23.82% (95% CI, 21.18-26.47%) in Strattice, 17.98% (95% CI, 15.49-20.47%) in Surgimend, 16.21% (95% CI, 14.44-17.99%) in Alloderm. Seroma rate was the highest in Strattice group (8.61%; 95% CI, 6.87-10.35%). There was no statistical significance between and within groups.
CONCLUSION: Although Strattice exhibited a higher overall pooled complication rate compared with Alloderm and Surgimend, the incidence of individual complication varies between studies. A cost analysis of different ADMs may aid in choosing the type of ADMs to be used.

INTRODUCTION

Acellular dermal matrices (ADMs) are biomaterials used extensively in the last decade in implant-based breast reconstructions with successful outcomes and acceptable complication rates.[1,2] The approach of using this technique is attributable to the ability to perform skin-sparing or nipple-sparing mastectomies.[3] First introduced in 2001, Salzberg et al.[3,4] has successfully introduced its use to provide implant coverage at the inferolateral pole of the breast in implant-based breast reconstructions.

It was reported that due to the different source and processing methods, the outcomes in the use in implant-based breast reconstruction are expected to differ.[4] In this study, we will be discussing and analyzing 3 of the most commonly used ADMs in implant-based breast reconstructions, which are Alloderm (LifeCell Corp., Branchburg, N.J.), Strattice (LifeCell Corp., Branchburg, N.J.), and Surgimend PRS (TEI, Biosciences, Inc., Boston, Mass.).

AlloDerm, the most commonly and widely investigated ADM, is a human-derived ADM commonly used in the United States and has not received the European Conformity (CE) marking for licensed used in Europe.[5] Alloderm is an ADM with reduced antigenicity that is produced from the epidermis human cadaveric skin. Alloderm was described as an ADM that is able to incorporate with host tissue via new matrix formed from specialized stem cells, which allowed for tissue regeneration.[6] Up until 2013, Alloderm, then known as Regenerative Tissue Matrix, was not terminally sterilized[6] contrary to Strattice and Surgimend. The Ready-to-Use counterpart was terminally sterilized.

Strattice is a noncrosslinked porcine-derived xenogeneic ADM. It is derived from porcine fetal dermis.[5,7] It became available at the end of 2008 and was licensed to use in Europe. Due to the limited availability in human cadaveric skin, Strattice has its advantages.[7] The structure and collagen arrangement of Strattice is almost similar to human cadaveric ADMs. It has undergone processing to maintain the integrity of the extracellular matrix allowing Strattice to act as a scaffold for cellular regeneration and neovascularization with hopes of reduced xenogeneic rejection response.[5,8]

SurgiMend, similar to Strattice, is a xenogeneic, noncrosslinked ADM, which is a fetal bovine-derived dermal collagen.[9,10] Studies have shown that it is rich in collagen types I and II[5] and described by manufacturers and study stating that it is rich is collagen type III.[6,11] These may facilitate tissue regeneration by impeding scarring.

Manufacturing

The manufacturing steps behind an ADM largely govern its subsequent functional properties in situ. One ADM can vary from another by either its source and/or processing of the tissue. Independent of the source, the tissue undergoes a decellularization process in which the extracellular matrix is isolated. The method of isolation varies between different types of ADMs but can be largely categorized into mechanical and more commonly chemical or biological processing.

The primary purpose of decellularization is to reduce the immunogenicity of the scaffold material so that it is host-compatible. In general, decellularization helps to augment the reconstructive capabilities of surrounding tissues.[12,13] The commonest methods of decellularization are chemical and biological including the use of trypsin/triton,[14] sodium hydroxide (NaOH),[15] sodium dodecyl sulfate, and sodium deoxycholate.

Following decellularization, the dermal scaffolds then undergo terminal sterilization process by various methods including ultraviolet radiation, gamma radiation, and supercritical fluid techniques including use of CO2.[16,17] Before shipment, the matrix is stored in a hydrated form or lyophilized to dry; then rehydrated for usage.[18]

METHODS

We have conducted a meticulous literature search on the databases Embase, Pubmed, and Medline, looking at literatures published from 2006 to 2017 in all 3 databases on the use of ADMs in implant-based immediate breast reconstruction. Search terms that were used include acellular dermal matrix, Acellular dermal matrices, ADMs, Breast reconstruction, breast implantation, breast implants, strattice, surgimend, alloderm, porcine, bovine, human ADM, complications, outcome, properties, collagen, seroma, infections, capsular contracture, hematoma, implant loss, explantation localized inflammation, localized erythema, and red breast syndrome.

Data Extraction

Primary outcome of interest for this meta-analysis was incidence of postoperative complications. Seven common complications associated with the use of ADMs in implant-based breast reconstruction were identified in the literatures, which were major infections classified as infections including cellulitis that were required readmission to theatre, minor infections classified as infections that were treated with oral or intravenous antibiotics that resolved without further complications, seroma, hematoma, implant loss, localized inflammation. Localized inflammation is erythema of the overlying skin in the absence of cellulitis or erysipelas or other skin infections and was not classified under complications in most studies. Figure 1 showed the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flow chart for the selection of articles.

Fig. 1.

Study attrition diagram.

The group of authors have also listed a series of inclusive criteria for the study, including all original studies on alloplastic/implant-based breast reconstruction with the use of ADMs, different biologically derived ADMs used in breast reconstruction, Alloderm, Strattice, and Surgimend. Additional data that were extracted were first authors; study institution; publication year; follow-up (mean/median); types of ADMs used, Alloderm, Strattice, and Surgimend; data for procedural characteristics; patient’s body mass index (BMI); confounding risk factors such as smoking, diabetes, neoadjuvant and adjuvant chemotherapy, and radiotherapy.

Exclusion criteria were review articles, discussions, published abstracts, case reports, articles written in non-English language, and articles published before 2006.

Statistical Analysis

Pooled random effect estimates for each postoperative complication and 95% confidence interval (95% CI) were calculated using Microsoft Excel. Using IBM SPSS Statistics for Windows (Version 22.0. Armonk, NY: IBM Corp.), one-way analysis of variance (ANOVA) and Bonferroni test were used to compare statistical significance between and within 3 groups, respectively. Multiple linear regression was done to include confounding factors. Using the complication rate and 95% CI, findings were presented on a forest plot using R Statistics (R Core Team (2013). R: A language and environment for statistical computing. R Foundation for Statistical Computing, Vienna, Austria.).

Risk of Bias Assessment

Cochrane risk of bias tool was used for risk bias assessment (Table 1). Studies were assessed for performance outcome bias, selective reporting bias, attrition bias, and funding bias. Low, fair, and high outcome levels were ranked in each study based on the level of biasness.

Table 1.

Risk Bias Study of the Literatures

RESULTS

A total of 58 publications were identified in the initial search on PubMed, Ovid, Medline, and Embase. Using predefined exclusion and inclusion selection criteria, 2 literatures were found to be duplicate, 16 literatures were not related to ADMs, 3 abstracts, 1 letter, 1 case report, and 14 reviews were excluded resulting in 21[1-5,7-11,19-29] studies that were eligible for this meta-analysis. Four of these studies were comparative studies of ADMs; therefore, we have stratified these data as individual studies[5,9,10,19] making a total of 25 studies. All of these studies satisfied criteria for the study of ADMs used in implant-based immediate breast reconstruction.

Eight Alloderm studies, 11 Strattice studies, and 6 Surgimend studies were identified for our analysis. As described in Table 1, in the studies that were assessed using Cochrane Risk of Bias Tool, 12 studies had high risk, 5 studies had low risk, and 4 studies had fair risk of bias.

Patient demographics, risk factors, and indication for surgery were pooled together in Table 2. The total number of implant-based immediate breast reconstructions were 1,659 breasts, 999 breasts, and 912 breasts involving the use of Alloderm, Strattice, and Surgimend, respectively. Majority of the cases were indicated for invasive mastectomies, followed by prophylactic mastectomies with only a total of 1 revision and 7 delayed mastectomies as reflected in Table 2.

Table 2.

Patient Demographics, Risk Factors, and Indications for Surgery between ADM Groups

From the literatures that were investigated, the patients age range were 31.9–58.5 years. Only 6 studies reported the mean mastectomy weight and axillary node clearance surgery. All of the studies reported patients who had adjuvant chemotherapy and radiotherapy. Patient’s comorbidities, history of chemotherapy, radiotherapy, and axillary node clearance surgery may affect postsurgical outcomes, and this will be discussed further in discussion.

Six common complications associated with the use of ADMs in implant-based breast reconstruction were identified in the literatures. Pooled complication rates are listed in Table 3. This showed a higher overall complication rate (23.82%) in the Strattice group as compared with Surgimend (17.98%) and Alloderm (16.21%), which had the best overall outcome. Incidence of individual complications of each group were listed in Tables 5–7. The most common complication was major infections in Alloderm (3.80%; CI, 2.88–4.72%) and seroma in Surgimend (4.61%; CI, 3.24–5.97%) and Strattice (8.61%; CI, 6.87–10.35%). Strattice was also found to have the highest incidence of localized inflammation at 3.3% (95% CI, 2.2–4.41%) as seen in Table 8.

Table 3.

One-way ANOVA between ADM Groups

Table 5.

Surgimend Group Pooled Complication

Table 7.

Alloderm Group Pooled Complication

Table 8.

Incidence of Localized Inflammation between ADMs

One-way analysis of variance test was computed on SPSS, and it was found that there were no significant statistical differences between all 3 groups. Bonferroni analysis was then done to compare the P value between groups, and there were no significant differences as depicted in Table 4. Further analysis of the outcome data by taking account for confounding factors, a multiple linear regression analysis was carried out. It was found that confounding factors did not significantly affect the outcome associated with the use of ADMs with the exception of minor infection rates in predictor group (d), which had a regression significance of P = 0.034 as seen in Table 9.

Table 4.

Bonferroni Statistical Analysis within Groups using SPSS

Table 9.

Regression Analysis of Outcome and Confounding Effects on All ADM Groups using SPSS

The occurrence of major infections was plotted in Figure 2, and it was statistically assumed that in 95% of these intervals intersection was at about 5.5%. Evgeniou et al.[22] had much smaller number of cases; hence a larger CI had a much deviated and higher complication rate (23.8%; 95% CI, 5.59–42.02%) associated with the use of Strattice. In Figure 3, minor infection rates were significantly higher in study by Lardi et al.[23] with the use of Strattice at 11.5% in 200 breasts (95% CI, 7.08–15.92%) and Liu et al.[24] at 8.5% in 165 breasts (95% CI, 4.23–12.73%). Salzberg et al.[3] reports a very small number of complication at 0.2% in 466 breast reconstructions (95% CI, −0.21% to 0.63%).

Fig. 2.

Forest plot estimating the proportion of the incidence of major infection among Alloderm, Surgimend, and Strattice groups.

Fig. 3.

Forest plot estimating the proportion of the incidence of minor infections among Alloderm, Surgimend, and Strattice groups.

Seroma rate associated with the use of Strattice had the highest (8.61%) pooled complication. In Figure 4, Strattice studies, Dikmans et al.,[7] Hille-Betz et al.,[25] and Evgeniou et al.[22] reported a complication rate of 20.9%, 20.4%, and 19%, respectively. In Alloderm group, pooled complication rates were the lowest at 3.07%. Glasberg and Light,[19] Butterfield,[10] and Gdalevitch et al.[1] reported an individual seroma rate of 21.4%, 15.7%, and 10.4%, respectively. Surgimend group has a pooled seroma rate of 4.61%. A comparative study[10] reported a lower seroma rate at 8.5% (95% CI, 5.62–11.47%) in 351 cases compared with Alloderm, which had a rate of 15.7% (95% CI, 8.17–23.29%) in 89 cases.

Fig. 4.

Forest plot estimating the proportion of the incidence of seroma among Alloderm, Surgimend, and Strattice groups.

Implant loss rate associated with the use of Strattice was the highest (5.61%) pooled complication. According to Figure 5, there was significantly higher implant loss rate in Strattice studies by Evgeniou et al.[22] Dikmans et al.,[7] and Lardi et al.[23] 23.8% (95% CI, 5.59–42.02%), 11.8% (95% CI, 5.79–17.83%), and 12.5% (95% CI, 7.92–17.08%), respectively. The rate of hematoma was relatively low and consistent with a pooled complication rate of 2.10%, 2.11%, and 1.21% in Strattice, Alloderm, and Surgimend, respectively in Figure 6.

Fig. 5.

Forest plot estimating the proportion of the incidence of implant loss among Alloderm, Surgimend, and Strattice groups.

Fig. 6.

Forest plot estimating the proportion of the incidence of hematoma among Alloderm, Surgimend, and Strattice groups.

The incidence rates for localized inflammation or erythema were irregular in the use of Strattice, Alloderm, and Surgimend as seen in Figure 7. The occurrence of capsular contracture in Table 3 showed that Alloderm was the highest at 1.21% (95% CI, 0.68–1.73%) compared with Strattice (0.80%; 95% CI, 0.25–1.35%). There were no reported cases of capsular contracture associated with the use of Surgimend in the searched articles.

Fig. 7.

Forest plot estimating the proportion of the incidence of localized inflammation among Alloderm, Surgimend, and Strattice groups.

DISCUSSION

ADM-assisted breast reconstruction provides better esthetic outcome when compared with the traditional subpectoral implant placement, in terms of creating a better inframammary fold definition and allowing for lower rate of capsular contracture.[30] In this meta-analysis, the performance of ADMs within in vivo models and how well this compared with human studies were also investigated.

A known complication in implant-based breast reconstruction following skin-sparing mastectomy is skin necrosis. It ranges from simple epidermolysis to a full-thickness flap necrosis, and if severe enough, lead to implant exposure, and subsequent implant loss.[3] Although this outcome was not included in our study because ADMs do not directly cause skin/flap necrosis, this will be discussed further. Studies report various methods to improve cellular behavior through surface modifications of ADMs including chemical modification with L-arginine in bovine ADM[31] and modification of porcine acellular dermal matrix PADM via dopamine self-polymerization/collagen immobilization.[32] A recent in vivo rat study further showed chemical cross-linking Permacol, which is porcine-derived, with hexamethylene diisocyanate caused an increase in cellular density and penetration at 12 months postimplantation compared with noncrosslinked implants. Furthermore, it was suggested that a thorough assessment of postmastectomy skin flap viability is crucial to reduce the incidence of skin necrosis.[23,27]

Capsular contracture is thought to be a local inflammatory response leading to excessive production of collagen by fibroblasts where they are in contact with the implant.[33] The Baker Classification system is a subjective classification system based upon clinical findings in the patient. Studies have proven that the use of ADMs in implant-based breast reconstruction is associated with a lower rate of capsular contracture up to a 20-fold reduction.[34] Basu et al.[35] reported a significant reduction in granulation tissue formation, levels of vascular proliferation, chronic inflammatory changes, fibroblast cellularity and foreign body giant cell inflammatory reaction, when comparing acellular cadaveric dermis sample to native breast capsules. Cross correlation with cytotoxicity studies in animal models reveal similar results. The degree of inflammation caused by human ADMs at 4 weeks in an in vivo rabbit model for incisional hernia repair was not statistically different from that caused by the use of porcine ADMs.[36] In both instance, the degree of inflammation detected by histology was low grade (level 1). A further study looked at the inflammatory response induced by porcine ADMs that were prepared by ultrasonification and freeze-thawing. Inflammatory markers, Interleukin-2 (IL-2) and Interferon gamma (IFN-y), were absent in both the PADM and human acellular dermal matrix group up to 48 hours postimplantation.[37] These are produced by antigen-sensitized T-cells in the context of foreign body rejection,[38] suggesting both were well tolerated. There may be a role for biopsy during a single-stage ADM breast reconstruction to ascertain local tissue inflammation through definitive histological analysis. This is especially necessary as capsular contracture itself can be influenced by other confounding factors.[34,39]

Although pooled data showed that Surgimend has the highest incidence of infection (7.68%), when individualized, Alloderm has the highest rate of major infections (3.8%) when compared with Surgimend (3.51%) and Strattice (2.1%). Although the HADM, Alloderm is classified as aseptic, histologic studies have reported neovascularization and inflammatory cell penetration into both sterile and aseptic ADMs[40] Nahabedian[41] has demonstrated that Alloderm was able to revascularize, recellularize and following tissue integration, able to tolerate mild-to-moderate infections. AlloDerm showed greater microvascular density and soft-tissue ingrowth.[42]

With regard to high incidence of seroma rate in Strattice group, Hille-Betz et al.[25] report that there was no antibiotic irrigation intraoperatively in their study, whereas the study by Dikmans et al.[7] stated that there was a lack of registration on the use of antibiotics. In the study by Butterfield 2012,[10] Surgimend had a much lower seroma rate as compared with Alloderm. The author reported that the number of drains inserted and whether the ADM is fenestrated or unfenestrated should be taken into consideration. Lardi et al.[23] have also suggested that the rate or seroma could potentially be reduced by lowering the drain removal threshold to around < 20 cc/24 h. Multiple studies have suggested that with the use of antibiotics for a period of time, 2 bulb suction drains, and by reducing the dead space between implant/expanded with skin, the incidence of seroma formation could potentially be reduced.[23]

Localized inflammation or erythema is also known as red breast syndrome, whose etiology is still poorly understood. It was thought to be a delayed hypersensitivity reaction to ADMs in breast reconstruction,[43] causing redness to the skin overlying the ADMs. Salzberg et al.[34] have reported that there is no evidence of true rejection response on histological analysis.

Multiple studies suggested that the main factors leading to an increase in complication rate are age older than 50 years, smoking status, mastectomy weight of > 600 g, and BMI > 30.[34,44] A couple of articles also showed that breast irradiation postoperatively is related to a higher rate of complication, including wound dehiscence, higher rate of infections, and possibly capsular contraction.[34,41] However, as discussed, there were no significant difference when adjusting for confounding factors. Inevitably, due to the lack of individual data from each study during extrapolation of data, we are unable to be completely advocate these accuracy as it might lead to a bias in results.

There were some limitations in this study. One of the main weaknesses is the low level of evidence in the studies included. We were only able to include 1 randomized controlled trial and 3 prospective cohort studies. It was also difficult to compare statistical differences in the indication for surgery as there were a few studies that recorded the number of patients instead of the number of breasts. In quite a few studies, we were not able to extract data on patient’s comorbidities such as smoking and diabetic status. We also acknowledge that most of our studies have high risk of bias as reported in Table 1.

From the thorough analysis, Strattice exhibited a slightly higher overall pooled complication rate compared with Alloderm and Surgimend. However, the incidence of individual complication varies between studies. Potential learning curve effects of using ADMs may affect the outcome. A cost analysis and a large prospective study of different ADMs may aid in choosing the type of ADMs to be used.

Table 6.

Strattice Group Pooled Complications