Comparative Host Response of 2 Human Acellular Dermal Matrices in a Primate Implant Model

Study overview

The experimental protocol was approved by the Institutional Animal Care and Use Committee of the Behavioral Sciences Foundation, St Kitts, Eastern Caribbean. Behavioral Sciences Foundation is accredited with the Canadian Council for Animal Care.

Nine adult male Caribbean vervets (Chlorocebus aethiops) (3—6 kg) were included in the study. Animals were randomly assigned to 1 of 3 treatment groups to receive a subcutaneous TE implant either on the left or right side of the dorsum. Animals were randomized to receive (1) TE only and TE with an overlay of aseptically processed HADM (TE + AD, with AD being AlloDerm Regenerative Tissue Matrix, LifeCell Corporation, Branchburg, New Jersey); (2) TE only and TE with an overlay of a gamma-irradiated HADM (TE + AM, with AM being AlloMax, Bard-Davol, Warwick, Rhode Island); or (3) TE + AD and TE + AM (Table 1). At 9 weeks following implantation, animals were killed and soft tissue around the TE was harvested. The TE-only group was used as a control (surgery, TE implant, skin flap coverage, no HADM).

TE implantation

Animals were fasted for 24 hours prior to the procedure and anesthetized by intramuscular injection of ketamine (10 mg/kg) and xylazine (1.0 mg/kg). Each animal was then placed in ventral recumbency and its upper back shaved and aseptically prepared for surgery. A single 125-mg dose of cefazolin was given intramuscularly before incision. Two horizontal curved incisions (≥6 cm) were made through the skin and subcutaneous tissues on either side of the back below the shoulder blades. Dorsal, rather than ventral, implantation was performed to prevent host postoperative tampering. A subcutaneous pocket was created above the dorsal musculature and deep fascia on either side of the spine to accommodate a 30-mL, 3-cm diameter circular, smooth silicone shell TE (PMT Corp, Chanhassen, Minnesota). For TE-only treatment, a TE was inserted into the subcutaneous pocket and filled with 25-mL sterile saline solution. The port and tubing were ligated and removed, and the TE anchored to the fascia with 2-0 polypropylene sutures around the port remnant. For the TE + AD and TE + AM groups, a 6 × 6 cm² dermal graft sheet (1.04-2.28 mm thick for AD; 0.8-1.8 mm thick for AM) was placed into the subcutaneous pocket and sutured down in a circumferential pattern to the underlying dorsal musculature. A TE was inserted above the muscle and positioned such that it was completely covered by dermal graft. The subcutaneous layer was then closed with polydioxanone sutures in a continuous subcuticular pattern, and skin was closed using nonabsorbable nylon sutures in an interrupted pattern. Animals received a 3-day course of cefazolin (125 mg bid, intramuscularly [IM]) with Banamine (Schering-Plough Animal Health, Kenilworth, New Jersey) 5 mg/kg given as needed. Animals were given flunixin meglumine (2.0—5.0 mg/kg, IM) or buprenorphine (0.01 mg/kg subcutaneously or IM) immediately after surgery, with additional analgesic permitted twice daily for 3 days or longer for persistent pain. All animals underwent biweekly physical examinations to monitor for complications, particularly at the surgical site.

Tissue harvesting

Animals were euthanized by intravenous injection of pentobarbital 9 weeks postimplantation. TE and surrounding tissues were removed en bloc, inclusive of the implanted dermal graft material, host tissues above the implanted TE, and the thoracic muscle below the TE. The entire explanted surgical pocket was then carefully divided along a sagittal plane into thirds, with the lateral one-third flash frozen in dry ice for later immunochemical analysis and the remaining two-thirds divided for fixation in 10% formalin (lateral segment) pending histologic and immunohistochemical analysis or immediate biomechanical testing (central unfixed segment).

Tissue assessments

Paraffin-embedded tissue samples were sectioned and underwent routine hematoxylin and eosin (H+E) staining and Verhoeff-Van Gieson (VVG) staining for elastin. For immunohistochemical labeling, tissue sections were deparaffinized and rehydrated, and proteinase K applied for antigen retrieval. Monoclonal antibodies to α-SMA, CD68, MMP-1, TIMP-1, and procollagen-1 were applied using the appropriate dilutions (Table 2). Detection was achieved using an appropriate secondary anti-immunoglobulin G (Table 2) conjugated with horseradish peroxidase and labeling was visualized with diaminobenzidine. Sections were evaluated for localization and intensity of staining for each immunohistochemical marker, scored by 2 independent histopathologists who were blinded to the nature of the individual samples, and the scores for each biopsy averaged. Immunohistochemical detection of the selected markers was used to determine the presence of inflammation (CD68), degradation (MMP-1/TIMP-1), characterization of capsule formation (α-SMA), and fibrosis (procollagen-1). HαE- and VVG-stained slides were evaluated for the presence of capsule and graft resorption as indicated by heterogeneous distribution of dermal elastin, respectively. Biopsies were scored for these factors as either present or absent. Relative presence and staining for each individual immunohistochemical marker was compared to the appropriate positive control, with a score of “-” indicating 0% staining, “+/-” indicating 5% to 20% of positive control, “+” indicating 20% to 40% of positive control, “++” indicating 40% to 60% of positive control, “+++” indicating 60% to 80% of positive control, and “++++” indicating 80% to 100% or equivalent to positive control as indicated in Table 2.

For biochemical analysis, flash-frozen samples were extracted and analyzed for inflammatory cytokines using a multiplex immunoassay array for biochemical markers of inflammation and fibrosis, including interleukins (ILs) 1, 2, 4, 6, 8, and 10, interferon (INF)-γ, regulated upon activation normal T cell expressed and secreted (RANTES), tumor necrosis factor (TNF)-α, and vascular endothelial growth factor (Bio-Rad, Hercules, California, cat M50-000007A). Briefly, frozen samples were pulverized in a tissue cryohomogenizer, followed by treatment with cell lysis kit solution (Bio-Rad, cat 171-304011) according to manufacturer's instructions. Samples were shaken for 15 minutes followed by centrifugation and the supernatant collected. For total protein content, an aliquot of supernatant was mixed with Bradford protein assay reagent (Bio-Rad, cat 500-0006), added to a 96-well plate and read at 595 nm. Additional supernatant aliquots were added to a 96-well plate incubated and rinsed according to manufacturer's instructions. On the basis of the array results showing consistently detectable concentrations of IL-6, INF-γ, and TNF-α in tissue explants, we validated the results using a 3-plex enzyme-linked immunosorbent assay (BioRad).

Uniaxial tensile loading was conducted on freshly dissected dermal graft samples for TE + AD and TE + AM groups. Briefly, a 1 × 3 cm² strip was placed in pneumatic side action grips with rubber faces and tested at a controlled strain rate of 1.65 per minute until failure (Instron 5860; Norwood, Massachusetts). Outcome measures included ultimate strength and elastic modulus. Data were processed using Instron Bluehill software and Microsoft Excel.

Data analysis

All implant sites were evaluated, with the exception of those where the TE extruded prior to scheduled explant date. Density and qualitative features of routine histologic staining and immunolabeling were described and scored independently by 2 independent histologists who were blinded to the nature of the individual samples, taking into account the entirety of 3 biopsy sections each, when viewed at 40×, 100×, and 200× magnifications. Scores were averaged to obtain the overall score for each sample, and the qualitative scores were compared across TE + AM and TE + AD groups using a Mann-Whitney test. For immunochemical evaluation, cytokine concentration for each sample was normalized to the total protein concentration and results reported in picogram analyte per milligram total protein. Immunochemical testing results from all evaluable specimens dissected from the overlying subcutaneous tissues were compared between the treatment groups using a 2-way analysis of variance model, following transformation to the normal distribution for biochemical results, with group defined as a factor; significance was defined as P ≤ .05. For mechanical testing results, a student t test was used to determine significance between individual groups.

Postoperative recovery

Two of the animals experienced extrusion of TEs before killing (1 animal [07657] with implant extruded from a TE + AD site at week 4.5, while the other side [TE + AM] was intact and 1 animal [07655] with both TEs extruded [TE-only and TE + AD] at week 1.5). The taut nature of the skin and observed necrotic tissue overlying these 3 extruded TEs early in the experiment indicated pressure necrosis due to overinflation of these particular TEs in correspondingly tight subcutaneous pockets. These surgical sites were not included in the analysis. No other surgical sites were observed to have pressure necrosis-induced complications.

Morphology: Histology and immunohistochemistry of implanted HADM

H+E and VVG staining confirmed the presence of both HADM types in the areas overlying the TE. There were, however, qualitative differences between the 2 groups. The soft tissue in the TE + AD group showed little to no sign of graft resorption, illustrated by an even distribution of elastin. Histology indicated the presence of fibroblasts and neovascularization within the AD graft, infiltrating from the matrix-dermis interface and not yet contacting the matrix-TE interface below, indicated by an asterisk (*) in Figure 1I. The immunohistochemistry of the soft tissue in the TE + AD group showed moderate presence of MMP-1 (Fig 1L, Table 3), with relatively little evidence of TIMP-1 (Table 3). Few α-SMA—labeled myofibroblasts were seen (Fig 1K), with only 1 of the 4 TE + AD sites exhibiting a single layer of α-SMA—positive myofibroblasts at the TE-matrix interface (Table 3). Procollagen-1 labeling was observed within the matrix and at the periphery (Table 3). There were few CD68-positive macrophages observed (Fig 1J, Table 3). On the contrary, the immunohistochemical profile of the soft tissue in the TE + AM group was different with regard to intensity and distribution. In 5 of 6 TE + AM implants evaluated, staining showed the presence of HADM with nonuniform thickness (Fig 1E), marked by thinned areas and a dense elastin fiber concentration, uncharacteristic of dermal tissue and suggesting matrix resorption. Areas of robust MMP-1 staining (Fig 1H), coupled with moderate TIMP-1 staining provided strong evidence for a different staining pattern compared to the AD group (Table 3). The α-SMA—positive myofibroblasts at the TE-matrix interface (Fig 1G, Table 3) suggested the presence of robust capsule formation in the TE + AM group. Procollagen-1 staining was negligible to moderate. Overlying tissue from TE + AM capsules demonstrated areas of significant macrophage presence (Fig 1F, Table 3) exhibiting a synovial-like metaplasia at the TE-matrix interface. Qualitative scores were compared across TE + AM and TE + AD groups using a Mann-Whitney test; P values were as follows: α-SMA, P = .04; CD68, P = .15; MMP-1, P = .08; TIMP-1, P = .12); procollagen-1, P = .15.

Figure 1. Representative histology and immunohistochemical staining using serial sections of tissue overlying the implanted tissue expander (TE) interface (indicated by *) for the TE-only (A-D), AlloMax (AM; E-H), and AlloDerm (AD; I-L) groups. Tissues are stained/labeled for hematoxylin and eosin (A, E, I); CD68 epitope for macrophages (B, F, J); α-smooth muscle cell actin (SMA) for myofibroblasts (C, G, K); and matrix metalloproteinase-1 (MMP-1) for matrix collagenase (D, H, I). In the TE-only group, a rim of macrophages can be detected ([, panel B) and presence of myofibroblasts ([, panel C). Tissue around the AM-covered TE had more modest macrophage presence ([, panel F); however, clusters were observed (indicated by arrowhead in F) and myofibroblast staining was strong ([, panel G). The lowest signal for macrophages and myofibroblast was noted in tissues from AD-covered TE. Excessive MMP-1 staining (D, H, L) is suggestive of accelerated ECM degradation in TE-only and clustered activity in AM (not observed in tissue from AD-covered TE).

Immunoassay

Concentrations of IL-1, IL-2, IL-4, IL-8, IL-10, RANTES, and vascular endothelial growth factor in all 3 groups were not detectable within the linear range of the assay. Measurable results were obtained for TNF-α, INF-γ, and IL-6 using the same array. Repeat analysis was performed to validate the array data and a separate 3-plex array for TNF-α, INF-γ, and IL-6 revealed that neither TNF-α nor INF-γ were significantly different among the 3 groups for samples collected from tissues overlying the TEs (Table 4). Analysis of variance revealed statistical significance across the treatment groups for IL-6 results (P = .006). Significantly lower levels of IL-6 were detected in tissues from the TE + AD group as compared to both the TE + AM (P = .004) and TE-only groups (P = .02) as determined by t test between individual groups. Interleukin-6 tissue concentrations were not significantly different between TE + AM and TE-only groups.

Mechanical strength testing

Table 5 shows the mechanical properties of HADMs, both out-of-package and following implantation for 9 weeks. While AM out-of-package tensile strength (6.19 ± 0.79 MPa) was significantly less than that of AD (10.93 ± 4.61 MPa) (P < .01), both HADMs exhibited a change in mechanical properties following implantation, with AD retaining ˜55% mechanical strength in vivo (6.00 ± 2.55 MPa) and AM retaining ˜12% (0.73 ± 0.46 MPa) (P < .001), in line with histology findings indicating significant resorption of AM. The elastic modulus of AD was significantly higher than AM, both out-of-package (48.85 ± 24.43 vs 17.42 ± 2.49 MPa, respectively; P < .001) and following implant (20.48 ± 5.80 vs 3.51 ± 3.43 MPa, respectively; P < .001).