Advertisement
Research Article| Volume 56, ISSUE 8, P775-783, December 2003

Prefabricated skin flaps in a rat model based on a dermal replacement matrix Permacol™

      Abstract

      The reconstruction of three-dimensional structures is one of the most challenging problems in plastic surgery. While shaped frameworks incorporated within soft tissue envelopes often lose their fine detail, the integration of a blood supply to a stable biomaterial scaffold allows skin graft cover and emphasis of fine detail. The aim of this study was to evaluate whether a porcine derived isocyanate crosslinked collagen tissue matrix (Permacol™) could be vascularised from an isolated vascular pedicle and subsequently sustain an overlying skin graft as a pedicled prefabricated flap. A vascular induction technique was employed using the epigastric pedicles in Sprague-Dawley rats to vascularise 2×2 cm2 pieces of Permacol™ of standard 0.75 and 1.5 mm thickness Permacol™ and laser treated Permacol™ of 0.4 and 0.75 mm thickness. Flaps were able to support overlying skin grafts as early as 2 weeks following vessel implantation and laser treated Permacol™ had an increased vascularity compared to standard Permacol™ with no evidence of degradation. The study showed that Permacol™ could be a useful matrix for use in tissue engineered flaps.

      Keywords

      The reconstruction of three-dimensional tissue defects such as those of the ear and nose remain one of the most challenging problems in plastic surgery. A shaped framework (bone, cartilage, or alloplast) incorporated into a soft tissue envelope (e.g. skin grafted fascial flap) is a commonly used technique whose results are usually unsatisfactory because of diffusion and obliteration of most of the fine detail of whatever scaffold is used beneath a thick tissue envelope. The integration of a blood supply to a stable biomaterial scaffold without remodelling or significant alteration in its shape and size would allow it to be covered with a thin skin graft so retaining the fine details of the scaffold and producing an improved aesthetic result.
      Several attempts have been made in experimental models to base such constructs on synthetic materials with only partial success. For example, Walton and colleagues demonstrated the feasibility of vascularising sheets of polytetrafluoroethylene from isolated vascular pedicles in the rabbit model.
      • Walton R.L
      • Brown R.E
      • Zhang L
      Creation of a vascularized alloplastic unit for composite reconstruction.
      • Walton R.L
      • Chick L.R
      • Brown R.E
      Development of vascularized allografts and neocomposite autografts.
      • Walton R.L
      • Brown R.E
      Tissue engineering of biomaterials for composite reconstruction: an experimental model.
      Although these constructs were able to support overlying skin grafts and could be transferred as free flaps, implant extrusion related to instability at the tissue–alloplast interface proved to be a problem. Similarly, Horl and co-workers produced a composite flap using polytetrafluoroethylene, a thin layer of fatty gliding tissue and a full thickness skin graft. This composite flap could also be transplanted as a free flap in a rabbit model.
      • Horl H.W
      • Kloppel M
      • Biemer E
      Experimental study using porous, vascularizable polytetrafluoroethylene in preformed free flaps.
      In a pilot clinical trial, Costa and colleagues attempted to reconstruct the nose and ear using prefabricated constructs based on a silicone tube framework inserted into a radial forearm flap.
      • Costa H
      • Cunha C
      • Guimaraes I
      • Comba S
      • Malta A
      • Lopes A
      Prefabricated flaps for the head and neck: a preliminary report.
      The use of collagen in flap prefabrication has focused mainly on its use as a carrier for growth factors such as bone morphogenetic protein
      • Kusumoto K
      • Bessho K
      • Fujimura K
      • Akioka J
      • Ogawa Y
      • Iizuka T
      Prefabricated muscle flap including bone induced by recombinant human bone morphogenetic protein-2: an experimental study of ectopic osteoinduction in a rat latissimus dorsi muscle flap.
      • Kuriloff D.B
      • Fayad J.N
      Tracheal autograft prefabrication using microfibrillar collagen and bone morphogenetic protein.
      or as a temporary matrix for tissue generation.
      • Tanaka Y
      • Tsutsumi A
      • Crowe D.M
      • Tajima S
      • Morrison W.A
      Generation of an autologous tissue (matrix) flap by combining an arteriovenous shunt loop with artificial skin in rats: preliminary report.
      The present study was designed to investigate the feasibility of producing a pedicled prefabricated flap in a rat model based on the permanent non-resorbable collagen implant Permacol™. It was hypothesized that a collagen-based biomaterial would provide a stable biomaterial–tissue interface, while promoting vascular integration.

      1. Materials and methods

      Female Sprague-Dawley rats weighing 250–350 g were used in all experiments. All experimental surgical procedures were performed under License from the UK Home Office (Animals Scientific Procedures Act 1986). Rats were anaesthetised by giving an intra-muscular injection of Hypnorm® (0.3 ml/kg) followed by an intraperitoneal injection of diazepam (2.5 mg/kg). Euthanasia was performed by an intracardiac injection of 1 ml pentobarbitone (Lethobarb®).

      1.1 Permacol™

      Permacol™ was supplied by Tissue Science Laboratories as 5×5 cm2 sheets of 0.4, 0.75 and 1.5 mm thickness. Treatment of 0.4 and 0.75 mm thickness Permacol™ with the diamond CO2 laser Permacol™ was performed by Exitech Ltd. Laser treated Permacol™ was sterilised at Tissue Science Laboratories. Four different types of Permacol™ matrix were used for the construction of prefabricated flaps: (a) standard 1.5 mm thickness; (b) standard 0.75 mm thickness; (c) laser treated 0.75 mm thickness with a median pore diameter of 124.33 μm (range 114.59–161.63 μm); and (d) laser treated 0.4 mm thickness material with a median pore diameter of 68.52 μm (range 48.32–97.79 μm). Polyethylene backing material (0.2 mm thickness) was used to isolate all constructs from the surrounding tissues.

      1.2 Histological examination

      Tissue specimens were fixed in 4% paraformaldehyde in phosphate buffered saline. After fixation, tissues were embedded in paraffin and 5 μm thick sections were cut and stained with haematoxylin and eosin or Masson's trichrome.

      1.3 Quantification of percentage vascularity

      For each slide specimen, eight randomly chosen fields of matrix were viewed under ×20 magnification through a superimposed standard stereology grid. Vascularity was defined on the presence of structures exhibiting typical vascular walls and containing red blood cells. Vessels intersecting the lines of the grid were counted and this was used to calculate the percentage vascularity.

      1.4 Experimental design

      Ninety-six female Sprague-Dawley rats (weighing 250–350 g) divided into four groups of 24 were used in this study. Each rat was anaesthetised and had two pieces of Permacol™ inserted into one groin sandwiching the epigastric pedicle (Fig. 1(A)–(C)) . Rats in Group 1 received standard 1.5 mm thickness Permacol™, rats in Group 2 received standard 0.75 mm thickness Permacol™, rats in Group 3 received laser treated 0.75 mm thickness Permacol™, and rats in Group 4 received laser treated 0.4 mm thickness Permacol™. Constructs were isolated from the surrounding tissues using polyethylene sheets (Fig. 1(D)).
      Figure thumbnail gr1
      Fig. 1Diagrammatic representation of the stages of prefabrication: (A) isolation of the superficial epigastric vessels; (B) placement of a polyethylene sheet together with a piece of Permacol™ behind the vessels; (C) placement of a second piece of Permacol™ in front of the vessels (sandwiching the superficial epigastric vessels between them); and (D) isolation of the Permacol™ construct from the surrounding tissues with a second piece of polyethylene sheet.
      Four rats in each group were killed at each of the time points (2, 4, 6 and 8 weeks) and their constructs excised and processed for paraffin sections. An assessment of the percentage vascularity of these constructs was performed.
      At each time point (2, 4, 6 and 8 weeks), the remaining two rats in each group underwent a further operation. The distal part of the epigastric pedicle was ligated and a split thickness skin graft (harvested from the flank of the rat) was sutured to the upper surface of the construct using 7/0 mersilk® (Fig. 2(B)) . Skin grafted constructs were isolated using polyethylene sheet and buried once again in the groin. After a further week, the rats were killed and the skin grafted constructs were fixed and processed for paraffin sections. A histological assessment of skin graft viability was performed.
      Figure thumbnail gr2
      Fig. 2Process of prefabrication continued: (A) opening of the polyethylene sheet at 2, 4, 6 or 8 weeks; (B) skin graft application to the upper surface of the construct and ligation of the distal part of the vascular pedicle; and (C) the pedicled prefabricated Permacol™ flap.
      One skin grafted 0.75 mm laser treated Permacol™ construct became infected and was omitted from the study.

      1.5 Statistics

      1.5.1 Vascularity study

      For the purposes of analysis, both standard 1.5 and standard 0.75 mm thickness Permacol™ constructs were combined into one group and this was compared against a laser treated group representing the combined laser treated 0.4 and 0.75 mm thickness Permacol™ constructs.
      The initial study was a comparison of the percentage vascularity of standard Permacol™ constructs compared to laser treated Permacol™ constructs at each of the different time points (2, 4, 6, and 8 weeks). The standard deviations of the mean percentage vascularities of the standard and laser treated groups were found to be vastly different and so a Welch test was used to make the comparison.
      The second study examined the increase in percentage vascularity of standard and laser treated Permacol™ constructs over the 8-week period. A paired t-test was used to compare the percentage vascularities: at 2 and 4 weeks; 4 and 8 weeks; and 2 and 8 weeks.

      1.5.2 Skin graft viability

      Skin graft viability was first assessed macroscopically and then graded microscopically as either histologically viable or non-viable. This scheme is classed as categorical and was analysed using Fisher's exact test. Comparison was made between skin grafts applied to the four types of Permacol™ construct: laser treated 0.4 mm thickness Permacol™; laser treated 0.75 mm thickness Permacol™, standard 0.75 mm thickness Permacol™; and standard 1.5 mm thickness Permacol™.

      2. Results

      2.1 Macroscopic appearance

      All polyethylene covered constructs were invested in a layer of fibrous tissue at the retrieval operation (Fig. 3) . The macroscopic appearance of the epigastric vascular pedicle inside a standard Permacol™ construct is shown intraoperatively (Fig. 4A) and then at 2 weeks post-operatively (Fig. 4B). At 4 weeks, blood vessels were seen to pass both around the periphery of the standard constructs (Fig. 5A) and directly through hair follicle remnants present in the Permacol™ matrix (Fig. 5B). After 6 weeks standard Permacol™ appeared to be fully vascularised (Fig. 6A) . A skin graft applied to the outer surface of a Permacol™ sandwich was still viable 1 week after its application to the outside of the Permacol™ construct (Fig. 6B).
      Figure thumbnail gr3
      Fig. 3Macroscopic appearance of fibrous tissue encapsulating the polyethylene enclosed constructs.
      Figure thumbnail gr4
      Fig. 4Photomicrograph of epigastric artery and vein on the surface of Permacol™ as seen intraoperatively (A) and 2 weeks postoperatively (B). Significant vascular outgrowth is seen from the epigastric pedicle onto inner surface of Permacol™ construct.
      Figure thumbnail gr5
      Fig. 5Photomicrographs showing the outer surface of a standard Permacol™ construct 4 weeks after implantation of the epigastric pedicle. Blood vessels are seen to pass both (A) around the periphery of the construct and (B) through hair follicle remnants present within the matrix.
      Figure thumbnail gr6
      Fig. 6Photomicrographs showing the outer surface of a standard Permacol™ construct: (A) 6 weeks after implantation of the vascular pedicle and (B) 1 week after application of a split thickness skin graft to its outer surface.

      2.2 Comparison of the vascularity of laser treated Permacol™ constructs vs. standard Permacol™ constructs over an 8-week period

      Laser treated Permacol™ constructs showed a significantly greater percentage vascularity compared to standard Permacol™ constructs at all time points studied: 2 weeks (P<0.001); 4 weeks (P<0.001); 6 weeks (P<0.001) and 8 weeks (P<0.001) (Fig. 7) . The macroscopic appearance of a laser treated Permacol™ construct is shown in Fig. 8(A). Fig. 8(B) demonstrates the histological appearance of vascular ingrowth into the pores within a laser treated Permacol™ construct.
      Figure thumbnail gr7
      Fig. 7Chart showing the percentage vascularity of: Laser treated 0.4 mm thickness Permacol™; Laser treated 0.75 mm Permacol™; Standard 0.75 mm thickness Permacol™ constructs; and Standard 1.5 mm thickness Permacol™ constructs 2, 4, 6 and 8 weeks after implantation of the epigastric vascular pedicle within the construct. Results shown are means±standard error of the mean of n=4 constructs at each time point. Laser treated Permacol™ constructs have a significantly greater percentage vascularity vs. standard Permacol™ constructs at all time points studied ***P<0.001.
      Figure thumbnail gr8
      Fig. 8Photomicrographs of laser treated Permacol™ construct showing: (A) the macroscopic appearance and (B) the microscopic appearance of blood vessels within the laser pores of the matrix. Histological section stained with Masson's trichrome (original magnification ×20).

      2.3 Study of the increases in percentage vascularity of laser treated Permacol™ constructs over an 8-week period

      There was no significant increase in the percentage vascularity of laser treated Permacol™ constructs beyond that seen after 2 weeks of implantation (Table 1) .
      Table 1Increase in the percentage vascularity of the combined laser treated 0.4 and 0.75 mm thickness Permacol™ constructs from: 2 to 4 weeks; 4 to 8 weeks; and 2 to 8 weeks after implantation of the epigastric vascular pedicle within the construct. Results shown are means±standard deviation for n=12 constructs. No significant increase in percentage vascularity is seen in laser treated Permacol™ constructs beyond 2 weeks of implantation
      Time intervalsMean difference in percentage vascularity (%)P-value
      Week 2–week 40.64±1.310.21
      Week 4–week 80.16±1.100.67
      Week 2–week 80.48±0.930.19

      2.4 Study of the vascularity of standard Permacol™ constructs over an 8-week period

      Standard Permacol™ constructs showed a significant progressive increase in percentage vascularity from 2 to 4 weeks (P=0.001), 4 weeks to 8 weeks (P=0.005) and 2 to 8 weeks (P<0.001) (Table 2) .
      Table 2Increase in the percentage vascularity of the combined standard 0.75 and 1.5 mm thickness Permacol™ constructs from: 2 to 4 weeks; 4 to 8 weeks; and 2 to 8 weeks after implantation of the epigastric vascular pedicle within the construct. Results shown are means±standard deviation for n=12 constructs. Standard Permacol™ constructs showed a significant progressive increase in percentage vascularity from 2 to 4 weeks (P=0.001), 4 weeks to 8 weeks (P=0.005) and 2 to 8 weeks (P<0.001)
      Time pointsMean difference in percentage vascularity (%)P-value
      Week 2–week 40.53±0.290.001
      Week 4–week 80.28±0.200.005
      Week 2–week 80.81±0.12<0.001

      2.5 Analysis of skin grafted Permacol™ constructs

      There was no significant difference in the viability of skin grafts applied to standard and laser treated Permacol™ constructs (Table 3) .
      Table 3Viability of skin grafts applied to 0.4 mm thickness laser treated Permacol™; 0.75 mm thickness laser treated Permacol™; 0.75 mm thickness standard Permacol™; and 1.5 mm thickness standard Permacol™ constructs. For each type of construct there are a maximum of eight skin grafts (representing two constructs grafted at each of four time points). Viability was assessed 1 week after grafting and values represent the number of viable and non-viable skin grafts
      0.4 laser treated Permacol™0.75 laser treated Permacol™0.75 standard Permacol™1.5 standard Permacol™P-value
      Number of skin graftsViable87860.23
      Non-viable012

      3. Discussion

      In 1987, Walton and colleagues
      • Walton R.L
      • Brown R.E
      • Zhang L
      Creation of a vascularized alloplastic unit for composite reconstruction.
      • Walton R.L
      • Chick L.R
      • Brown R.E
      Development of vascularized allografts and neocomposite autografts.
      • Walton R.L
      • Brown R.E
      Tissue engineering of biomaterials for composite reconstruction: an experimental model.
      demonstrated the feasibility of using an isolated vascular pedicle to vascularise synthetic biomaterials (polytetrafluoroethylene and expanded polytetrafluoroethylene) and produce a synthetic flap capable of supporting an overlying skin graft. They demonstrated that such a flap could be transferred as a free flap 6 weeks after vessel implantation in a rabbit model. Costa and colleagues
      • Costa H
      • Cunha C
      • Guimaraes I
      • Comba S
      • Malta A
      • Lopes A
      Prefabricated flaps for the head and neck: a preliminary report.
      then prefabricated a silicone tube framework into a radial forearm flap for total reconstruction of the nose and ear.
      The use of collagen as a biomaterial for the construction of prefabricated flaps has previously focused on its use as a carrier for growth factors such as bone morphogenetic protein
      • Kusumoto K
      • Bessho K
      • Fujimura K
      • Akioka J
      • Ogawa Y
      • Iizuka T
      Prefabricated muscle flap including bone induced by recombinant human bone morphogenetic protein-2: an experimental study of ectopic osteoinduction in a rat latissimus dorsi muscle flap.
      • Kuriloff D.B
      • Fayad J.N
      Tracheal autograft prefabrication using microfibrillar collagen and bone morphogenetic protein.
      or as a temporary scaffold for tissue generation.
      • Tanaka Y
      • Tsutsumi A
      • Crowe D.M
      • Tajima S
      • Morrison W.A
      Generation of an autologous tissue (matrix) flap by combining an arteriovenous shunt loop with artificial skin in rats: preliminary report.
      A recent study performed by Tark allowed a 2-week period of vascularisation of a subcutaneously implanted dermal matrix (AlloDerm®) before turning this over as a random pattern prefabricated flap for the support of sheets of cultured autologous keratinocytes in a pig model.
      • Tark K.C
      • Chung S
      • Shin K.S
      • Park B.Y
      Skin flap prefabrication using acellular dermal matrix and cultured keratinocytes in a porcine model.

      3.1 Vascularisation of permacol™ constructs

      This study showed that the vascularity of laser treated Permacol™ constructs was significantly greater than standard Permacol™ constructs at each corresponding time point over an 8-week period of implantation in the Sprague-Dawley rat model (Fig. 7). On comparing the rate of vascularisation of standard and laser treated Permacol™ constructs it was found that while laser treated constructs appear to be optimally vascularised within 2 weeks (Table 1), the vascularity of standard Permacol™ constructs gradually increases over the 8-week period (Table 2). The greater porosity of laser treated Permacol™ explains the difference in percentage vascularity compared to standard Permacol™. The fact that these porous spaces are filled quickly is consistent with the knowledge that fibrovascular ingrowth is much more rapid into a porous biomaterial. Blood vessels present in the standard Permacol™ appeared to pass around the periphery of the material and through hair follicle remnants present within the matrix.
      This study was a simple histological study of vascularity. Both the upper and lower surfaces of the Permacol™ were isolated using polyethylene sheet although a space was present to allow the entry and exit of the vascular pedicle from the construct. Hence, it is likely that the majority of vessels within the Permacol™ construct arose from the vascular pedicle although it is possible that some grew in from surrounding connective tissues.
      The factor responsible for fibrovascular outgrowth from the epigastric vascular pedicle is considered to be an inflammatory response secondary to the presence of the porcine collagen matrix and polyethylene backing material. The source of endothelial sprouting is likely to be from the vasa vasorum and/or the lumen of the epigastric vessels. Assessment of vascularity was performed histologically and it was therefore not possible to confirm that these capillaries were in communication with the epigastric vessels implanted at the centre of each construct. Future studies should be performed with carbon injection studies,
      • Tanaka Y
      • Tsutsumi A
      • Crowe D.M
      • Tajima S
      • Morrison W.A
      Generation of an autologous tissue (matrix) flap by combining an arteriovenous shunt loop with artificial skin in rats: preliminary report.
      Indian ink preparation
      • Hickey M.J
      • Wilson Y
      • Hurley J.V
      • Morrison W.A
      Mode of vascularization of control and basic fibroblast growth factor-stimulated prefabricated skin flaps.
      or microangiographic contrast techniques
      • Yao S.T
      Vascular implantation into skin flap. Experimental study and clinical application: a preliminary report.
      to demonstrate these connections.

      3.2 Pedicled prefabricated flap

      This study found that skin grafts applied to both standard and laser treated Permacol™ constructs as early as 2 weeks after vessel implantation maintained their viability when seen 1 week after grafting (Table 3). This finding is consistent with that of Tanaka and colleagues who demonstrated the ability of flaps consisting of a vascularised porcine type I collagen sponge (3 mm in thickness) to support an overlying skin graft after 2 weeks of vascular implantation.
      • Tanaka Y
      • Tsutsumi A
      • Crowe D.M
      • Tajima S
      • Morrison W.A
      Generation of an autologous tissue (matrix) flap by combining an arteriovenous shunt loop with artificial skin in rats: preliminary report.
      Histological examination of constructs revealed a good degree of integration between the dermis of the skin grafts and the Permacol™ matrix. There was no evidence of any implant extrusion as is sometimes seen with synthetic biomaterials such as polytetrafluoroethylene (PTFE).
      • Walton R.L
      • Chick L.R
      • Brown R.E
      Development of vascularized allografts and neocomposite autografts.
      There are conflicting reports in the literature as to the optimum time for raising and transferring a prefabricated flap. Walton found a period of 6 weeks sufficient for prefabricated PTFE while Tanaka found a period of 2 weeks sufficient for collagen-based constructs. The coating of alloplastic materials such as polyethylene with immobilised collagen is known to promote soft tissue ingrowth.
      • Kinoshita Y
      • Kuzuhara T
      • Kirigakubo M
      • Kobayashi M
      • Shimura K
      • Ikada Y
      Soft tissue reaction to collagen-immobilized porous polyethylene: subcutaneous implantation in rats for 20 wk.
      This study was performed to establish the feasibility of producing a prefabricated pedicled flap using the biomaterial Permacol™. Future studies should address the size limitations of such flaps and the success of free flap transfer. Angiographic studies should be performed to establish the communications of the vascular network within the Permacol™ with the epigastric vessels. The results presented clearly demonstrate the feasibility of creating a pedicled prefabricated flap from Permacol™ capable of supporting an overlying skin graft. The fact that both standard and laser treated Permacol™ could be vascularised from an isolated vascular pedicle suggests that this biomaterial may be useful in the prefabrication of artificial flaps in patients.

      Acknowledgements

      This work was funded and supported by the Restoration of Appearance and Function Trust (RAFT).

      References

        • Walton R.L
        • Brown R.E
        • Zhang L
        Creation of a vascularized alloplastic unit for composite reconstruction.
        Plast Surg Forum. 1987; 56: 32
        • Walton R.L
        • Chick L.R
        • Brown R.E
        Development of vascularized allografts and neocomposite autografts.
        Plast Surg Forum. 1988; 57: 160
        • Walton R.L
        • Brown R.E
        Tissue engineering of biomaterials for composite reconstruction: an experimental model.
        Ann Plast Surg. 1993; 30: 105-110
        • Horl H.W
        • Kloppel M
        • Biemer E
        Experimental study using porous, vascularizable polytetrafluoroethylene in preformed free flaps.
        Handchir Mikrochir Plast Chir. 1992; 24: 243-249
        • Costa H
        • Cunha C
        • Guimaraes I
        • Comba S
        • Malta A
        • Lopes A
        Prefabricated flaps for the head and neck: a preliminary report.
        Br J Plast Surg. 1993; 46: 223-227
        • Kusumoto K
        • Bessho K
        • Fujimura K
        • Akioka J
        • Ogawa Y
        • Iizuka T
        Prefabricated muscle flap including bone induced by recombinant human bone morphogenetic protein-2: an experimental study of ectopic osteoinduction in a rat latissimus dorsi muscle flap.
        Br J Plast Surg. 1998; 51: 275-280
        • Kuriloff D.B
        • Fayad J.N
        Tracheal autograft prefabrication using microfibrillar collagen and bone morphogenetic protein.
        Arch Otolaryngol Head Neck Surg. 1996; 122: 1385-1389
        • Tanaka Y
        • Tsutsumi A
        • Crowe D.M
        • Tajima S
        • Morrison W.A
        Generation of an autologous tissue (matrix) flap by combining an arteriovenous shunt loop with artificial skin in rats: preliminary report.
        Br J Plast Surg. 2000; 53: 51-57
        • Tark K.C
        • Chung S
        • Shin K.S
        • Park B.Y
        Skin flap prefabrication using acellular dermal matrix and cultured keratinocytes in a porcine model.
        Ann Plast Surg. 2000; 44: 392-397
        • Hickey M.J
        • Wilson Y
        • Hurley J.V
        • Morrison W.A
        Mode of vascularization of control and basic fibroblast growth factor-stimulated prefabricated skin flaps.
        Plast Reconstr Surg. 1998; 101: 1296-1304
        • Yao S.T
        Vascular implantation into skin flap. Experimental study and clinical application: a preliminary report.
        Plast Reconstr Surg. 1981; 68: 404-410
        • Kinoshita Y
        • Kuzuhara T
        • Kirigakubo M
        • Kobayashi M
        • Shimura K
        • Ikada Y
        Soft tissue reaction to collagen-immobilized porous polyethylene: subcutaneous implantation in rats for 20 wk.
        Biomaterials. 1993; 14: 209-215