Physical and biological properties of a new synthetic amino acid copolymer used as wound dressing R. Eloy INSERM Unite' 37, 18, avenue du Doyen Lkpine F69500 Bron, France
A. Brack Centre de Biophysique Moleculaire, CNRS, 1, avenue de la Recherche Scientifique F45071 Orleans CMex 2. France
N. Dorme Centre de Recherche Delalande, 10, rue des Carrie'res F92500 Rued Malmaison, France A.M. Cornillac BlOMATECH S. A. Z. 1. de Z'lslon Rue Pasteur F38670, Chasse Sur Rhone, France Anew synthetic amino acid copolymer has been evaluated as wound covering. It is permeable to water vapor in the region of 4.1 kg/m2/24 h, it does not allow microbial proliferation after in vitro inoculation, it is impermeable to bacteria, and is stable and flexible. In vivo experiments were designed to provide qualitative and quantitative evaluation on its possible use as a skin substitute in full-thickness skin excision in the guinea pig. Two excisions, approximately 12-14 cm2 were performed on each side of the spine, leaving the panniculus carnosus. One site was treated with the membrane, the other with gauze. Each animal served as its own control. Photographs with a fixed focal-length camera were taken in identical conditions for all wounds immediately and 7,14, and
21 days after excision. They were analyzed by planimetry. Histological studies were performed at 7,14, and 21 days. The rate of healing between 0 and 21 days of the wounds treated with the copolymer membrane was significantly accelerated in comparison with wounds covered with a dry dressing ( p < 0.05). This increased epithelialization rate was confirmed by histology, which also suggested a reduction of the inflammatory response of t h e wound. In vivo biodegradation studies were also performed by subcutaneous implantation in the rat followed at 15, 30, 60, and 90 days by histology and physicochemical analyses. The results demonstrate that the membrane is not biodegradable.
IN TRODUC TION
During the past decade the desirable properties of an effective burn wound membrane dressing have been well defined and a variety of candidate membranes have been evaluated. The properties needed for a clinically *INERPAN, Laboratoires DELALANDE La Defense 2,16, rue Henri Regnault 92400 COURBEVOIE France. Journal of Biomedical Materials Research, Vol. 26, 695-712 (1992) CCC 0021-9304/92/060695-18$4.00 0 1992 John Wiley & Sons, Inc.
ELOY ET AL.
696
effective skin substitute are as follows:'r2absence of antigenicity, good tissue compatibility, absence of local and systemic toxicity, water vapor transmission similar to normal skin, i.e., which allows a proper balance in the repairing wound to prevent either hydration or desiccation of the repairing tissue, impermeability to exogenous microorganisms, rapid adherence to wound surface, surface structure that does not inhibit ingrowth or proliferation of healing tissue, flexibility and pliability to allow conformation to wound surface, prevention of bacterial proliferation and reduction of bacterial density of the wound, readily available material, ability to survive sterilization. In addition the ideal burn wound dressing will reduce, or suppress pain in the burn wound site. In fact, this ideal definition is somewhat underestimated since: (1) The quantity of fluid lost from a wound is not clearly defined, either by exudation or by water evaporation. (2) The ideal water vapor transmission rate of the dressing is not known, but it should not desiccate the wound. (3) The claim of a water vapor transmission rate equal to that of the normal skin is valid for long-term skin substitutes rather than for short-term dressings3 The wound surface should be kept just damp enough to obtain the benefits of accelerated healing, but there should be no accumulation of fluid between the wound and the dressing because of the risk of infection. The concept of moist wound healing had emerged as early as 1963; when it was demonstrated that wounds reepithelialize more rapidly under moist conditions. We report here our findings with a new nonbiodegradable amino acid copolymer membrane* whose characteristics promote wound healing as compared to gauze dressing in the guinea pig model. The aim of the present protocol was to determine: (1) The rate and quality of healing of dermoepidermic wounds in the guinea pig in contact with a dressing applied at the time of lesioning. (2) The potential biodegradation of the membrane in a subcutaneous implantation site in a rat model using physicochemical characterization of the membrane and histological criteria. MATERIAL A N D METHODS
Synthesis and characteristics of copoly (leucine and methyl glutamate) membrane Copolypeptides of the type:
NH - CH -COX--,
1
CH3
CH3
-NH-
CH -CO'-'--
I CH2 I CH2 I
O=C-OCH3
x
=
molar fraction of leucine
n = average degree of polymerization
697
SYNTHETIC AMINO ACID COPOLYMER DRESSING
were prepared by copolymerization of L-leucine (Leu) and gammamethyl-Lglutamate (Glu (OMe)). The polymerization was monitored for randomness by taking aliquots during the reaction process. Amino acid composition of the polymer was checked by HPLC after complete hydrolysis. The amino acid balance was also obtained from the NMR spectrum (Fig. 1). The optimal rotation was found to lie in a range of -45 & 5" mL dm-lg-'. The number-average molecular weight M, was estimated using the relationship: (v). = 1.06 Mts5(5), where (7) represents the intrinsic viscosity and M o stands for the average molecular weight of the amino acid residue. M, must be higher than 60,000 daltons in order to obtain a membrane with the required mechanical properties. The polydispersity of the samples was determined by gel permeation chromatography. Values around 3.5 were obtained.
/ 'MS
Leu 0.9 9
J1.70
2P 1.78
iL chemical
shift
ppm
Figure 1. Proton NMR spectrum of copoly (leucine and methyl glutamate).
ELOY ET AL.
698
The conformation of the polypeptide in solution and in the solid state is that of an alpha-helix6 as shown by circular dichroism studies (Fig. 2). The membrane of copoly (leucine and methyl glutamate) is impregnated by immersion in a liquid containing 75% of polyethylene glycol 600 and 25% of 0.9% sodium chloride solution. The flexible membrane which is obtained is cut, placed on a support film (which has to be removed before use), put into sachets, and sterilized by gamma rays at a dose of 25-30 kilograys. The membrane obtained had no visible pores crossing through it and the relief of the side facing the support was slightly flattened (Fig. 3). Physicochemical characterization of the membrane
Permeability of water vapor The procedure, derived from a monograph of the European Pharmacopoeia, consisted in measuring the weight loss of a bottle filled with 20 mL
w a v e l e n g t h nm
Figure 2. Circular dichroism spectrum of copoly (leucine and methyl glutamate).
SYNTHETIC AMINO ACID COPOLYMER DRESSING
(4 Figure 3. Scanning electron micrographs of the membrane, (a) side exposed to air (1 cm = 125 p m ) , (b) side facing the support film (1 cm = 125 pm), (c) cross section (1 cm = 250 p m ) .
699
700
ELOY ET AL.
water on the cap of which a punch 28 mm in diameter was closed with a 36mm diameter sample of the membrane. The bottle was placed for 18 hours in an oven at 37°C. The permeability to water vapor (kg/m2/24 h) is expressed by the following:
A x 18 A: surface of the opening of the cap (mm’), WTl: mass of the bottle before placing in the oven (g), Wlz: mass of the bottle after removal from the oven (g). Permeability to liquids The permeability to water and to antiseptic solutions was examined using an Amicon type 8050 ultrafiltration cell. Antiseptics tested were Dakin’s solution, 5% chlorhexidine diluted 1/100, iodated povidone. A surface of 46 mm diameter of the membrane was placed in the cell reservoir. After 1 hour the mass of the liquid contained in the outflow tube was measured. The permeability to a liquid is expressed by the following (WTi - WT2) X 24 X lo3 A A = surface of contact between liquid and the sample of the membrane in mm2
Test of mechanical resistance and elongation capacity This was performed on equipment consisting of type DFGz dynamometer, type LTCM motorized chassis, RS 232 interface, Epson H x 20 microcomputer.
Artificial con tam inat ion test This test was conducted according to the French Pharmacopoeia, using five different strains (Staphylococcus aureus, Escherichia coli, Pseudomonas aeruginosa, Clostridium sporogenes, Candida albicans). Squares of the membrane, 35 x 35 mm, were placed in Petri dishes and in-oculated with 100,000 of each of the microorganisms. The organisms were counted on day 7 and on day 14.
Occlusiveness test Samples of the membrane were cut into 10 x 10 cm squares and each square was placed on a Mueller Hinton Agar plate (DIFCO) seeded with Pseudomonas aeruginosa strain.
SYNTHETIC AMINO ACID COPOLYMER DRESSING
701
Using specific reagents which promote the production of fluorescent green pigment (Pyoverdine) and blue-green pigment (Pyocyanine) and placed on the membrane for 7 days at 25”C, it could be assessed whether the membrane was or was not permeable to this germ.
Stability testing Using organoleptic characterization, loss on drying, infrared absorption for identification of copoly (leucine and methylglutamate), test for small peptides and polymer assay, this study was carried out in selected packaging at 20°C on one batch over 1 year, and on two batches during 2 years.
Study of wound healing
Experimental procedure Female Dunkin-Hartley albino guinea pigs were used (300-350 g). Under ketamine chlorhydrate anesthesia, two wounds of 3.5 x 3.5 cm were performed on each side of the spine by dermoepidermic excisions. The panniculus carnosus was resected, after control and dissection under a Zeiss operating microscope (magnification x10 to x 40). The wounds were covered immediately after photography with the membrane, on one side and sterile gauze on the other. The dressings were placed alternately on the right or left following the order of the animals. Each dressing was covered with a sterile compress secured by an hypoallergenic elastic adhesive bandage. Animals were housed in individual cages after identification. At the 7th, 14th, and 21st postoperative (p.0.) day, 3 animals were anesthetized, the dressings were removed, macroscopic observations and photographs of each wound with a reference graduated scale were made. After this, a lethal dose of anesthetic (sodium pento barbital, 800 mg/animal) was administered and the samples for histological observation were taken and fixed in Bouin’s fluid. Animal care strictly conformed to the Guidelines of the National Institute of Health and Medical Research (decree No. 87-848, October 19th, 1987).
Morphometric analyses Wounds were photographed at days 0, 7, 14, and 21 in standardized conditions of lighting, distance, and aperture kept constant throughout the whole experiment. A transfer of nonepidermized zone was made by choosing a projection of a photographic image such that the graduated scale of each slide represents the real size. The surface area of each transfer was analyzed by planimetry on a MopKontron AM02 analyser, by following the contour of each wound using an optoelectronic pencil.
ELOY ET AL.
702
Histological processing At days 14 and 21 of sacrifice, the wounds were removed in their entirety and sectioned in the median zone perpendicularly to the head-tail axis. Specimens were fixed in Bouin's fluid for 3 to 10 days. After washing and dehydration, samples were embedded in methylmethacrylate. Sections were cut using an automatic Reichert Jung Polycut microtome. Sections, 3 p m thick, were deplastified, stained with Masson's trichrome, and examined under Polyvar Reichert Jung, a light microscope, equipped with a tracing board for analysis of morphometric data (MOP AM02 KONTRON).
In vivo biodegradation of the membrane In this study, the copolymer membrane was implanted subcutaneously in 20 Wistar rats of 200-220 g distributed in four experimental groups, each consisting of five animals. They were anesthetized by Ip injection of 0.3 mL/100 g of Equitesine. Three cutaneous incisions were made on each side of the spine and patches of the membrane made using a circular metallic patch of 14 mm diameter, were inserted in the six implantation sites. The skin was then sutured and after 15, 30, 62, and 90 days the animals were sacrificed by groups of 5. Four of the six implants of each animal were used for chemical analysis; the remaining two were fixed and one of them was used for histological analysis. After retrieval, the implants were rinsed in sterile physiological serum and two of them were processed as described before for histology and microscopy examination. Physicochemical analysis was performed by basic hydrolysis, 24 hours at 120"C, of the polymer followed by GLC chromatography (Intersmat IGC 16). RESULTS
Physicochemical properties of the membrane
Permeability to water vapor The mean water vapor permeability 4.1 k 0.14 kg/m2/24 h.
(k
SEM) obtained from 21 assays was
Permeability to liquids The mean permeability values following (in kg/m2/24 h): water: 18.9 & 1.5 Dakin's solution: 15.4 2 2.5
(k
SEM) for the different liquids were the
( n = 17) ( n = 15)
703
SYNTHETIC AMINO ACID COPOLYMER DRESSING
5% chlorhexidine: 16.5 f 3.1 ( n = 15) lodated Povidone: 4.3 f 0.3 ( n = 7) They show that the membrane is permeable to these solutions.
Mechanical resistance and elongation capacity The mean mechanical resistance was 0.25 f 0.01 kg (n = 23), whereas the mean elongation was 23.3 2 1.6 cm ( n = 23). These characteristics confer a good flexibility to the membrane.
Artificial contamination test The samples exposed to the organisms during 7 and 14 days no longer contained any of the microorganisms inoculated.
Occlusiveness No color developed on the membrane after 7 days, suggesting that Pseudomonas aeruginosa does not cross the membrane.
Stability test at room temperature All tests performed showed that two batches were stable after storage for 2 years at 20°C, 1 batch tested only during 1 year was stable at 20°C (Table I). TABLE I Stability at Room Temperature: Batch C 21
Test
Initial Analysis
Results after 6 months
Results after 1 year
Results after 2 years
Organolept ic characteristics
Colorless, transparent film
Colorless, transparent film
Colorless, transparent film
Colorless, transparent film
Loss on drying (%)
54.4
53.3 m = 51.5 49.8
55 rn = 54.5 54
50 rn = 51 52
Polymer identification
Complies
Complies
Complies
Complies
2.62
2.9 rn = 2.9 2.9
2.8 m = 2.8 2.7 2.8 2.8
2.6 rn = 2.65 2.7
Test negative
Test negative
Test negative
Test negative
Polymer assay (%)
Test for small peptides M =
mean value
ELOY ET AL.
704
Macroscopic findings on skin healing At the 7th postoperative day, generally the membrane adhered to the wound which was not deep and only slightly hemorrhagic, whereas the wounds covered with gauze were generally deeper, more hemorrhagic, and the gauze adhered closely to the wound. At the 14th postoperative day, the membrane still adhered to the wound and had a cardboard aspect. These wounds were less deep, less extensive than those covered with gauze which were still hemorrhagic. At the 21st postoperative day, the membrane no longer adhered and the three wounds treated with it were completely or almost completely healed, whereas the wounds covered with gauze were still sizable and often hemorrhagic.
Morphometric findings The results of the analysis of surface area are reproduced in Table 11. The means of the wound surface covered with the gauze, expressed as a percentage of the wound's initial surface area on days 7, 14, and 21, were 56%, 33%, and 1576, respectively. On the same days, the wounds treated with the polymer membrane had values of 4976,1276, and 2%. TABLE I1 Surface of Wounds Polymeric Membrane
Day of Sacrifice
Guinea Pig Number
D 7
1 2 3 Mean SD 4 5 6 Mean SD 7 8 9 Mean SD
D 14
D 21
P S S' % ! M
P
S
S'
%
P
S
S'
%
R L R
988 992 982 987 5 793 837 953 861 93 940 1140 1171 1084 125
534 476 432 481 51 113 93 101 102 10 13 25 21 20 6
54 47 44 49 6 14 11 11 12 26 1 2 2 2 0
L R L
1017 1953 1010 1026 23 987 1003 1052 989 71 930 868 1169 989 159
514 518 688 573 99 298 367 298 321 40 841 518 688 573 99
51 49 58 56 11 33 37 28 33 4 19 11 16 15 4
L
R L -
R L R
= position of wound (R = right, L = left)
initial surface of wound in mm2 residual surface of wound in mmz = percentage S ' / S = mean SD = standard deviation = =
Gauze
-
-
R L R -
L R L
SYNTHETIC AMINO ACID COPOLYMER DRESSING
705
S f at ist icaI analysis The mean values of the differences of wound surface area, observed for each animal between day 7 and day 0, day 15 and day 0, and day 21 and day 0 for the two types of wound dressing were analyzed using the comparison of regression slopes obtained for each type of wound. The mean values -C standard deviation of the slope in the case of wounds treated with the polymer membrane was 39.6 5 5.26 (v = 0.94), whereas for the dry dressing the value was 27.44 ? 5.84 (Y = 0.87). The slope was significantly steeper in the polymer membrane treated wounds in relation to those treated with a dry dressing ( p < 0.05). This shows that there was a significant acceleration of the healing process between days 0 and 21 of the guinea pig wound treated with the polymer membrane in comparison with a wound covered with a dry dressing.
Histological findings At day 7, the wounds covered with the polymer membrane showed a scar bud characterized by a large fibrin deposit, the epithelialization process had started in two among three animals and was almost complete in one animal (Table 111). On the wounds covered with gauze, no epithelium was detectable, the dermis was very rich in polynuclear and macrophage inflammatory cells. TABLE 111 Morphometric Data Obtained from Histological Examination of Wounds (Sections realized in the median part of the wound) Polymeric Membrane Day of Sacrifice ~
Guinea Pig Number ~
L
T
Gauze
D
L
T
D
* *
* *
~
1 2 3
12.6 11.2 0
0.5 -C 0.1 1 f 0.3 2 f 0.2
>12 >11 1.8
13.2
1.9 f 0.5 2.6 f 0.1 0.7 f 0.2
D 14
4 5 6
3.1 4.1 4.1
1.2 t 0.2 2 f 0.2 0.9 f 0.2
5.9 6.4 6.7
7.7 16.9 10.3
0.5 f 0.2 0.5 f 0.1 0.5 f 0.1
10.3 18.6 12.1
D 21
7 8 9
0 0 0
2.2 f 0.2 2 ? 0.1 2 f 0.1
7.7 3.6 6.8
6.1 4.4 7.6
1.2 f 0.3 1.6 f 0.3 1.4 2 0.3
6.5 9.7 7.7
D 7
213
L = nonepithelialized length in mm, T = average thickness in mm of the scar (layer above the panniculus cornosus) (mean f standard deviation of at least 20 measurements per section) (derm + epiderm in normal tissue = T = 2 ? 0.4), D = distance between epidermic adnexa in mm. *Measurements could not be made because the two edges of the wound were not present on the section.
706
ELOY ET AL.
At the 14th postoperative day (Fig. 4), wounds covered with the polymer membrane showed an epithelium which covered a large area of the scar zone as seen in Table 111. The newly formed dermis was not so thick as that of the healthy tissue, but in the epithelialized zones fibroblasts were taking the place of inflammatory cells. On the wounds treated with gauze, the epithelialization process was very slight, the dermis was highly vascularized and infiltrated by macrophages, giant cells localized around gauze fiber debris or foreign bodies. At the 21st postoperative day (Fig. 5), the epithelialization of the polymer membrane treated wounds was complete (Table 111). This epithelium was well organized and covered on its internal side by a horny layer, whereas the dermis was rich in fibroblasts, all oriented along the same axis parallel to the skin surface. Formation of epidermal adnexa could be observed toward the edges of the scar. Some residual inflammatory cells and giant cells could be observed, and the thickness of scar tissue was identical to that of a healthy tissue. At this stage, wounds covered with gauze were still not epithelialized (Table 111) and layers of fibrin, hemorrhagic and strongly inflammatory dermis were still present, with macrophages and giant cells.
In vivo degradation of the membrane implanted subcutaneously in rats: histological evaluation (Fig. 6) At the 15th postimplantation day, a well organized fibrocellular capsule in which fibroblasts were predominant, surrounded the membrane. The membrane was generally not altered, only small defects were detected at the edges of some implants, probably related to initial irregularities rather than to a process of degradation. At the 30th postimplantation day, the tissular reaction around the implants was unmodified while macrophage cell density was decreased, and fibroblast density increased. At the 62nd postimplantation day, no degradation process was detectable on histological sections of the implant. There was no increase in the number of giant cells and no increase of membrane defects or enlargement was observed. At the 90th postimplantation day, the inflammatory component was still decreasing and neovascularization was reduced; some macrophages and giant cells were still visible in small foci at the poles of the implant, which could present at these places slight irregularities.
In vivo biodegradation of the membrane: physicochemical analyses Each animal had been statistically implanted with the same amount of the membrane of about 96 mg with a relative standard deviation of 2.6% to 3.4%.
SYNTHETIC AMINO ACID COPOLYMER DRESSING
Figure 4. Histology of wounds after 14 days of treatment. (A,B) Wound covered with the membrane after 14 days of contact. The middle of the wound is not epithelialized. The dermis is thick but still inflammatory. (C,D) Control wound covered with gauze after 14 days of contact. The middle of the wound is not epithelialized. The thickness of the scar tissue is reduced. The dermis is infiltrated by inflammatory cells and giant cells. Bar = 100 pm.
707
708
ELOY ET AL.
Figure 5. Histology of wounds after 21 days of treatment. (A,B) Wound covered with the membrane after 21 days of contact. The scar zone is completely epithelialized. The dermis is rich in fibroblasts with a beginning of orientation. There are still some large mononuclear cells. (C,D) Wound covered with gauze after 21 days of contact. The wound has not yet healed, there remains a nonepithelialized and hemorrhagic zone. The dermis is inflammatory, characterized by its richness in macrophages giant cells and neovessels. Bar = 100 pm.
SYNTHETIC AMINO ACID COPOLYMER DRESSING
Figure 6. Biodegradation study of the polymeric membrane implanted subcutaneously in rats. (A,B) After 15 days of implantation the implant is surrounded by a fibrocellular capsule in which fibroblasts are predominant. There are some macrophages located at the pole of the implant. The membrane is not altered by cellular penetration. (C,D) After 90 days of implantation no degradation process is detectable. The small focus of inflammatory cells, visible in (C) in one pole of the implant, is probably related to an initial irregularity of the membrane. The implant is surrounded by a fibrocellular tissue. Bar = 100 pm.
709
ELOY ET AL.
710
The amount of polymer expressed in mg per implant was 2.69 k 0.175 (mean ? SD of 9 determinations). The amount of polymer in test implants is given in Table IV and the results show that polymer amounts in the membrane recovered after 15, 30, 62, 90 days of subcutaneous implantation in rat did not differ significantly from polymer amount in blank implant. The same characteristic infrared absorption bands were observed in test and blank implants.
DISCUSSION
Laminates composed of polypeptides have already been suggested as a wound covering about 15 years ago but, as viewed by Aiba et a1.7 Feijen et al.; Anderson et al.," the physical and chemical characteristics of the polypeptides were poorly understood. They were suggested as potential for the scaffolding of cell adhesion and/or growth. A copoly (leucine methylglutamate) membrane with specific properties was developed in order to comply with certain requirements of an "optimized wound covering." Since a limited initial adherence to defects can reduce pain, as well as control infection, and this optimizes the rate of healing,'" a relative elasticity and adhesion to the wound were searched for. The membrane adheres over moist areas and as the healing process develops adhesion of the membrane to the granulation tissue is progressively replaced by epithelialization and elimination of the membrane. The membrane which is not biodegradable, has never been found inside the wound, by histology. The second specification for this membrane is a relative permeability to water vapor. Normal skin allows water vapor to pass at a rate of approximately 8.5 g/m2/hour while retaining proteins and electrolytes. Although the optimal water vapor transmission of burn dressings has not been established, an excess of permeability can lead to water loss and drying of tissues, whereas absence of permeability induces formation of blisters. The recently developed wound dressing, derived polyurethane membrane, have a permeability ranging between 500 g/m2/24 h for OpSite and 5000 g/m2/24 h for Omiderm." The present polypeptide membrane, by its permeability to water and by the addition of a humectant like PEG, appears to maintain during the first days of dressing an adequate hydration of the membrane, thus reducing the risk of extensive eschar formation. As reported previously for polypropylene glycol'
+
TABLE IV Amount of Polymer (mg/implant) after Increasing Periods of Implantation Before Implantation 2.69
+ 0.175
After 15 Days of Implantation
After 30 Days of Implantation
After 62 Days of Implantation
2.73 f 0.166
2.87 f 0.158
2.70 k 0.398
After YO Days of Implantation 2.65
* 0.336
Results are expressed as the mean f SD in at least 5 different experimental animals
SYNTHETIC AMINO ACID COPOLYMER DRESSING
711
it could be expected that water electrolytes and proteins should reach equilibrium in the liquid component of the membrane with the wound exsudate. This concept providing a controlled adequate microenvironment to the wound is based on the evidence that wound healing is enhanced by occlusive dressings (for review see ref. 12). The effect is not limited to the dermis since occlusion in guinea pig in our study caused a great reduction of the inflammatory response. Especially neutrophils, but also macrophages, appeared earlier but were more rapidly replaced by numerous fibroblasts and collagenous material. An attractive but still not demonstrated hypothesis for the faster epithelialization under occlusive dressings is that they retain wound fluids which contain a variety of factors for connective tissue formation and epidermal migration, in contact with the wound. In this view the polymeric membrane may be considered as a pharmacological agent, absorbing to some extent factors present in the wound but able to keep them available for the different steps of healing. The durability of the membrane may also contribute to this effect by suppressing the need and the drawbacks of a new "non-biologically filled dressing. After a preliminary ~tudy,'""~ which reported the potential interest of the biomaterial, the quantitative data generated in this study suggest that the dermoepithelial repair process is enhanced in the presence of the polymeric membrane considering the following: the excision procedure and size were uniformly standardized and randomized; each animal served as its own control; the control and treated sites were photographed in standard conditions and the only variable was the rate of wound healing; the degree of epithelial proliferation was determined by planimetry for each wound; epithelialization of excised wounds was significantly higher than for control gauze dressings. The polymer fulfils biocompatibility testing: it is not cytotoxic on mouse L 929 fibroblasts, it is nonhemolytic in nitro, it does not induce hypersensitivity in the guinea pig, it is nonantigenic and well tolerated in the intracutaneous test in the rabbit. It is nonmutagenic in ~ i f r 0 . l ~ " ~ Preliminary clinical studies including skin graft donor sites, mesh grafts, skin defects of various origin, partial and full thickness burns treated with the membrane confirm these experimental data. Work is in progress to compare the epithelialization rate and characteristics of different experimental wound models in rats, treated comparatively with other. References 1. R.M. Nalbadian, R.L. Henry, K.W. Balko, D.V. Adams, and N.R. Neuman, "Pluronic F-127 Gel preparation as an artificial skin in the treatment of third-degree burns in pigs," J. Biomed. Mater. Res., 21, 1135-1148 (1987). 2. B . A . Pruitt and N.S. Levine, "Characteristics and uses of biologic dressings and skin substitutes," Arch. Surg., 119, 314-322 (1984). 3. K. J. Quinn, J. M., Courtney, J. H. Evans, J. D. S. Gaylor, and W. H. Reid, "Principles of burn dressing," Biomaterials, 6, 369-377 (1985).
ELOY ET AL.
712
8. 9.
10.
11.
12.
13. 14.
C. D. Hinman and H. Maibach, ”Effect of air exposure and occlusion on experimental human skin wounds,” Nature, 200, 377-378 (1963). A. Brack and Y. Trudelle, “Molecular weight determination of polypeptides from viscosity measurements,” Polym. Com. 26, 369-370 (1985). A. Brack and G. Spach, ”Multiconformational synthetic polypeptide,” 1. A m . Chem. Soc., 103,6319-6329 (1981). S . Aiba, N. Minoura, Y. Fujiwara, S. Yamada, and T. Nakagawa, “Laminates composed of polypeptides and elastomers as a burn wound covering physicochemical properties,” Biomuterials, 6, 290-296 (1985). J. Feijen, W. L. Sederel, and T. Beugeling, ”Synthetic polypeptides as biomaterials,” Proc. Euro. Soc. Artif. Organs, 1, 37-42 (1975). J. M. Anderson, D.E Gibbons, R. L. Martin, A. Miltner, and R. Woods, ”The potential of poly a amino acids as biomaterials,” J. Biomed. Muter. Res., 5, 197-207 (1974). M. J. Tavis, J. Thornton, R. Danet, and R. H. Barlett, ”Current status of skin substitutes,” Surg Clin. of North Am., 58, 1233-1248 (1978). D. Behar, M. Juszynski, N. Ben Hur, J. Golan, A Eldad, J. Tuchman, N. Sterenberg, and B. Rudensky, “Omiderm a new synthetic wound covering: physical properties and drug permeability studies,” J. Biomed. Mater. Res., 20, 731-738 (1986). V. Falanga, ”Occlusive wound dressing. Why, When, Which?,” Arch. Dermatol., 124, 872-877 (1988). R. Eloy and N. Weill, ”Etude prhclinique du PA 286, substitut cutan6 temporaire,” XPme CongrPs de la Societe Franqaise d’Etude et de Traitement des Briilures, 28-30 Septembre 1989, Metz, France. R. Eloy and N. Weill, “Preclinical studies of PA 186, a temporary skin substitute.” Td Congress of the European Burns Association, October 4-7, 1989. Prague, Czechoslovakia.
Received September 17,1990 Accepted October 31,1991
A. Brack Centre de Biophysique Moleculaire, CNRS, 1, avenue de la Recherche Scientifique F45071 Orleans CMex 2. France
N. Dorme Centre de Recherche Delalande, 10, rue des Carrie'res F92500 Rued Malmaison, France A.M. Cornillac BlOMATECH S. A. Z. 1. de Z'lslon Rue Pasteur F38670, Chasse Sur Rhone, France Anew synthetic amino acid copolymer has been evaluated as wound covering. It is permeable to water vapor in the region of 4.1 kg/m2/24 h, it does not allow microbial proliferation after in vitro inoculation, it is impermeable to bacteria, and is stable and flexible. In vivo experiments were designed to provide qualitative and quantitative evaluation on its possible use as a skin substitute in full-thickness skin excision in the guinea pig. Two excisions, approximately 12-14 cm2 were performed on each side of the spine, leaving the panniculus carnosus. One site was treated with the membrane, the other with gauze. Each animal served as its own control. Photographs with a fixed focal-length camera were taken in identical conditions for all wounds immediately and 7,14, and
21 days after excision. They were analyzed by planimetry. Histological studies were performed at 7,14, and 21 days. The rate of healing between 0 and 21 days of the wounds treated with the copolymer membrane was significantly accelerated in comparison with wounds covered with a dry dressing ( p < 0.05). This increased epithelialization rate was confirmed by histology, which also suggested a reduction of the inflammatory response of t h e wound. In vivo biodegradation studies were also performed by subcutaneous implantation in the rat followed at 15, 30, 60, and 90 days by histology and physicochemical analyses. The results demonstrate that the membrane is not biodegradable.
IN TRODUC TION
During the past decade the desirable properties of an effective burn wound membrane dressing have been well defined and a variety of candidate membranes have been evaluated. The properties needed for a clinically *INERPAN, Laboratoires DELALANDE La Defense 2,16, rue Henri Regnault 92400 COURBEVOIE France. Journal of Biomedical Materials Research, Vol. 26, 695-712 (1992) CCC 0021-9304/92/060695-18$4.00 0 1992 John Wiley & Sons, Inc.
ELOY ET AL.
696
effective skin substitute are as follows:'r2absence of antigenicity, good tissue compatibility, absence of local and systemic toxicity, water vapor transmission similar to normal skin, i.e., which allows a proper balance in the repairing wound to prevent either hydration or desiccation of the repairing tissue, impermeability to exogenous microorganisms, rapid adherence to wound surface, surface structure that does not inhibit ingrowth or proliferation of healing tissue, flexibility and pliability to allow conformation to wound surface, prevention of bacterial proliferation and reduction of bacterial density of the wound, readily available material, ability to survive sterilization. In addition the ideal burn wound dressing will reduce, or suppress pain in the burn wound site. In fact, this ideal definition is somewhat underestimated since: (1) The quantity of fluid lost from a wound is not clearly defined, either by exudation or by water evaporation. (2) The ideal water vapor transmission rate of the dressing is not known, but it should not desiccate the wound. (3) The claim of a water vapor transmission rate equal to that of the normal skin is valid for long-term skin substitutes rather than for short-term dressings3 The wound surface should be kept just damp enough to obtain the benefits of accelerated healing, but there should be no accumulation of fluid between the wound and the dressing because of the risk of infection. The concept of moist wound healing had emerged as early as 1963; when it was demonstrated that wounds reepithelialize more rapidly under moist conditions. We report here our findings with a new nonbiodegradable amino acid copolymer membrane* whose characteristics promote wound healing as compared to gauze dressing in the guinea pig model. The aim of the present protocol was to determine: (1) The rate and quality of healing of dermoepidermic wounds in the guinea pig in contact with a dressing applied at the time of lesioning. (2) The potential biodegradation of the membrane in a subcutaneous implantation site in a rat model using physicochemical characterization of the membrane and histological criteria. MATERIAL A N D METHODS
Synthesis and characteristics of copoly (leucine and methyl glutamate) membrane Copolypeptides of the type:
NH - CH -COX--,
1
CH3
CH3
-NH-
CH -CO'-'--
I CH2 I CH2 I
O=C-OCH3
x
=
molar fraction of leucine
n = average degree of polymerization
697
SYNTHETIC AMINO ACID COPOLYMER DRESSING
were prepared by copolymerization of L-leucine (Leu) and gammamethyl-Lglutamate (Glu (OMe)). The polymerization was monitored for randomness by taking aliquots during the reaction process. Amino acid composition of the polymer was checked by HPLC after complete hydrolysis. The amino acid balance was also obtained from the NMR spectrum (Fig. 1). The optimal rotation was found to lie in a range of -45 & 5" mL dm-lg-'. The number-average molecular weight M, was estimated using the relationship: (v). = 1.06 Mts5(5), where (7) represents the intrinsic viscosity and M o stands for the average molecular weight of the amino acid residue. M, must be higher than 60,000 daltons in order to obtain a membrane with the required mechanical properties. The polydispersity of the samples was determined by gel permeation chromatography. Values around 3.5 were obtained.
/ 'MS
Leu 0.9 9
J1.70
2P 1.78
iL chemical
shift
ppm
Figure 1. Proton NMR spectrum of copoly (leucine and methyl glutamate).
ELOY ET AL.
698
The conformation of the polypeptide in solution and in the solid state is that of an alpha-helix6 as shown by circular dichroism studies (Fig. 2). The membrane of copoly (leucine and methyl glutamate) is impregnated by immersion in a liquid containing 75% of polyethylene glycol 600 and 25% of 0.9% sodium chloride solution. The flexible membrane which is obtained is cut, placed on a support film (which has to be removed before use), put into sachets, and sterilized by gamma rays at a dose of 25-30 kilograys. The membrane obtained had no visible pores crossing through it and the relief of the side facing the support was slightly flattened (Fig. 3). Physicochemical characterization of the membrane
Permeability of water vapor The procedure, derived from a monograph of the European Pharmacopoeia, consisted in measuring the weight loss of a bottle filled with 20 mL
w a v e l e n g t h nm
Figure 2. Circular dichroism spectrum of copoly (leucine and methyl glutamate).
SYNTHETIC AMINO ACID COPOLYMER DRESSING
(4 Figure 3. Scanning electron micrographs of the membrane, (a) side exposed to air (1 cm = 125 p m ) , (b) side facing the support film (1 cm = 125 pm), (c) cross section (1 cm = 250 p m ) .
699
700
ELOY ET AL.
water on the cap of which a punch 28 mm in diameter was closed with a 36mm diameter sample of the membrane. The bottle was placed for 18 hours in an oven at 37°C. The permeability to water vapor (kg/m2/24 h) is expressed by the following:
A x 18 A: surface of the opening of the cap (mm’), WTl: mass of the bottle before placing in the oven (g), Wlz: mass of the bottle after removal from the oven (g). Permeability to liquids The permeability to water and to antiseptic solutions was examined using an Amicon type 8050 ultrafiltration cell. Antiseptics tested were Dakin’s solution, 5% chlorhexidine diluted 1/100, iodated povidone. A surface of 46 mm diameter of the membrane was placed in the cell reservoir. After 1 hour the mass of the liquid contained in the outflow tube was measured. The permeability to a liquid is expressed by the following (WTi - WT2) X 24 X lo3 A A = surface of contact between liquid and the sample of the membrane in mm2
Test of mechanical resistance and elongation capacity This was performed on equipment consisting of type DFGz dynamometer, type LTCM motorized chassis, RS 232 interface, Epson H x 20 microcomputer.
Artificial con tam inat ion test This test was conducted according to the French Pharmacopoeia, using five different strains (Staphylococcus aureus, Escherichia coli, Pseudomonas aeruginosa, Clostridium sporogenes, Candida albicans). Squares of the membrane, 35 x 35 mm, were placed in Petri dishes and in-oculated with 100,000 of each of the microorganisms. The organisms were counted on day 7 and on day 14.
Occlusiveness test Samples of the membrane were cut into 10 x 10 cm squares and each square was placed on a Mueller Hinton Agar plate (DIFCO) seeded with Pseudomonas aeruginosa strain.
SYNTHETIC AMINO ACID COPOLYMER DRESSING
701
Using specific reagents which promote the production of fluorescent green pigment (Pyoverdine) and blue-green pigment (Pyocyanine) and placed on the membrane for 7 days at 25”C, it could be assessed whether the membrane was or was not permeable to this germ.
Stability testing Using organoleptic characterization, loss on drying, infrared absorption for identification of copoly (leucine and methylglutamate), test for small peptides and polymer assay, this study was carried out in selected packaging at 20°C on one batch over 1 year, and on two batches during 2 years.
Study of wound healing
Experimental procedure Female Dunkin-Hartley albino guinea pigs were used (300-350 g). Under ketamine chlorhydrate anesthesia, two wounds of 3.5 x 3.5 cm were performed on each side of the spine by dermoepidermic excisions. The panniculus carnosus was resected, after control and dissection under a Zeiss operating microscope (magnification x10 to x 40). The wounds were covered immediately after photography with the membrane, on one side and sterile gauze on the other. The dressings were placed alternately on the right or left following the order of the animals. Each dressing was covered with a sterile compress secured by an hypoallergenic elastic adhesive bandage. Animals were housed in individual cages after identification. At the 7th, 14th, and 21st postoperative (p.0.) day, 3 animals were anesthetized, the dressings were removed, macroscopic observations and photographs of each wound with a reference graduated scale were made. After this, a lethal dose of anesthetic (sodium pento barbital, 800 mg/animal) was administered and the samples for histological observation were taken and fixed in Bouin’s fluid. Animal care strictly conformed to the Guidelines of the National Institute of Health and Medical Research (decree No. 87-848, October 19th, 1987).
Morphometric analyses Wounds were photographed at days 0, 7, 14, and 21 in standardized conditions of lighting, distance, and aperture kept constant throughout the whole experiment. A transfer of nonepidermized zone was made by choosing a projection of a photographic image such that the graduated scale of each slide represents the real size. The surface area of each transfer was analyzed by planimetry on a MopKontron AM02 analyser, by following the contour of each wound using an optoelectronic pencil.
ELOY ET AL.
702
Histological processing At days 14 and 21 of sacrifice, the wounds were removed in their entirety and sectioned in the median zone perpendicularly to the head-tail axis. Specimens were fixed in Bouin's fluid for 3 to 10 days. After washing and dehydration, samples were embedded in methylmethacrylate. Sections were cut using an automatic Reichert Jung Polycut microtome. Sections, 3 p m thick, were deplastified, stained with Masson's trichrome, and examined under Polyvar Reichert Jung, a light microscope, equipped with a tracing board for analysis of morphometric data (MOP AM02 KONTRON).
In vivo biodegradation of the membrane In this study, the copolymer membrane was implanted subcutaneously in 20 Wistar rats of 200-220 g distributed in four experimental groups, each consisting of five animals. They were anesthetized by Ip injection of 0.3 mL/100 g of Equitesine. Three cutaneous incisions were made on each side of the spine and patches of the membrane made using a circular metallic patch of 14 mm diameter, were inserted in the six implantation sites. The skin was then sutured and after 15, 30, 62, and 90 days the animals were sacrificed by groups of 5. Four of the six implants of each animal were used for chemical analysis; the remaining two were fixed and one of them was used for histological analysis. After retrieval, the implants were rinsed in sterile physiological serum and two of them were processed as described before for histology and microscopy examination. Physicochemical analysis was performed by basic hydrolysis, 24 hours at 120"C, of the polymer followed by GLC chromatography (Intersmat IGC 16). RESULTS
Physicochemical properties of the membrane
Permeability to water vapor The mean water vapor permeability 4.1 k 0.14 kg/m2/24 h.
(k
SEM) obtained from 21 assays was
Permeability to liquids The mean permeability values following (in kg/m2/24 h): water: 18.9 & 1.5 Dakin's solution: 15.4 2 2.5
(k
SEM) for the different liquids were the
( n = 17) ( n = 15)
703
SYNTHETIC AMINO ACID COPOLYMER DRESSING
5% chlorhexidine: 16.5 f 3.1 ( n = 15) lodated Povidone: 4.3 f 0.3 ( n = 7) They show that the membrane is permeable to these solutions.
Mechanical resistance and elongation capacity The mean mechanical resistance was 0.25 f 0.01 kg (n = 23), whereas the mean elongation was 23.3 2 1.6 cm ( n = 23). These characteristics confer a good flexibility to the membrane.
Artificial contamination test The samples exposed to the organisms during 7 and 14 days no longer contained any of the microorganisms inoculated.
Occlusiveness No color developed on the membrane after 7 days, suggesting that Pseudomonas aeruginosa does not cross the membrane.
Stability test at room temperature All tests performed showed that two batches were stable after storage for 2 years at 20°C, 1 batch tested only during 1 year was stable at 20°C (Table I). TABLE I Stability at Room Temperature: Batch C 21
Test
Initial Analysis
Results after 6 months
Results after 1 year
Results after 2 years
Organolept ic characteristics
Colorless, transparent film
Colorless, transparent film
Colorless, transparent film
Colorless, transparent film
Loss on drying (%)
54.4
53.3 m = 51.5 49.8
55 rn = 54.5 54
50 rn = 51 52
Polymer identification
Complies
Complies
Complies
Complies
2.62
2.9 rn = 2.9 2.9
2.8 m = 2.8 2.7 2.8 2.8
2.6 rn = 2.65 2.7
Test negative
Test negative
Test negative
Test negative
Polymer assay (%)
Test for small peptides M =
mean value
ELOY ET AL.
704
Macroscopic findings on skin healing At the 7th postoperative day, generally the membrane adhered to the wound which was not deep and only slightly hemorrhagic, whereas the wounds covered with gauze were generally deeper, more hemorrhagic, and the gauze adhered closely to the wound. At the 14th postoperative day, the membrane still adhered to the wound and had a cardboard aspect. These wounds were less deep, less extensive than those covered with gauze which were still hemorrhagic. At the 21st postoperative day, the membrane no longer adhered and the three wounds treated with it were completely or almost completely healed, whereas the wounds covered with gauze were still sizable and often hemorrhagic.
Morphometric findings The results of the analysis of surface area are reproduced in Table 11. The means of the wound surface covered with the gauze, expressed as a percentage of the wound's initial surface area on days 7, 14, and 21, were 56%, 33%, and 1576, respectively. On the same days, the wounds treated with the polymer membrane had values of 4976,1276, and 2%. TABLE I1 Surface of Wounds Polymeric Membrane
Day of Sacrifice
Guinea Pig Number
D 7
1 2 3 Mean SD 4 5 6 Mean SD 7 8 9 Mean SD
D 14
D 21
P S S' % ! M
P
S
S'
%
P
S
S'
%
R L R
988 992 982 987 5 793 837 953 861 93 940 1140 1171 1084 125
534 476 432 481 51 113 93 101 102 10 13 25 21 20 6
54 47 44 49 6 14 11 11 12 26 1 2 2 2 0
L R L
1017 1953 1010 1026 23 987 1003 1052 989 71 930 868 1169 989 159
514 518 688 573 99 298 367 298 321 40 841 518 688 573 99
51 49 58 56 11 33 37 28 33 4 19 11 16 15 4
L
R L -
R L R
= position of wound (R = right, L = left)
initial surface of wound in mm2 residual surface of wound in mmz = percentage S ' / S = mean SD = standard deviation = =
Gauze
-
-
R L R -
L R L
SYNTHETIC AMINO ACID COPOLYMER DRESSING
705
S f at ist icaI analysis The mean values of the differences of wound surface area, observed for each animal between day 7 and day 0, day 15 and day 0, and day 21 and day 0 for the two types of wound dressing were analyzed using the comparison of regression slopes obtained for each type of wound. The mean values -C standard deviation of the slope in the case of wounds treated with the polymer membrane was 39.6 5 5.26 (v = 0.94), whereas for the dry dressing the value was 27.44 ? 5.84 (Y = 0.87). The slope was significantly steeper in the polymer membrane treated wounds in relation to those treated with a dry dressing ( p < 0.05). This shows that there was a significant acceleration of the healing process between days 0 and 21 of the guinea pig wound treated with the polymer membrane in comparison with a wound covered with a dry dressing.
Histological findings At day 7, the wounds covered with the polymer membrane showed a scar bud characterized by a large fibrin deposit, the epithelialization process had started in two among three animals and was almost complete in one animal (Table 111). On the wounds covered with gauze, no epithelium was detectable, the dermis was very rich in polynuclear and macrophage inflammatory cells. TABLE 111 Morphometric Data Obtained from Histological Examination of Wounds (Sections realized in the median part of the wound) Polymeric Membrane Day of Sacrifice ~
Guinea Pig Number ~
L
T
Gauze
D
L
T
D
* *
* *
~
1 2 3
12.6 11.2 0
0.5 -C 0.1 1 f 0.3 2 f 0.2
>12 >11 1.8
13.2
1.9 f 0.5 2.6 f 0.1 0.7 f 0.2
D 14
4 5 6
3.1 4.1 4.1
1.2 t 0.2 2 f 0.2 0.9 f 0.2
5.9 6.4 6.7
7.7 16.9 10.3
0.5 f 0.2 0.5 f 0.1 0.5 f 0.1
10.3 18.6 12.1
D 21
7 8 9
0 0 0
2.2 f 0.2 2 ? 0.1 2 f 0.1
7.7 3.6 6.8
6.1 4.4 7.6
1.2 f 0.3 1.6 f 0.3 1.4 2 0.3
6.5 9.7 7.7
D 7
213
L = nonepithelialized length in mm, T = average thickness in mm of the scar (layer above the panniculus cornosus) (mean f standard deviation of at least 20 measurements per section) (derm + epiderm in normal tissue = T = 2 ? 0.4), D = distance between epidermic adnexa in mm. *Measurements could not be made because the two edges of the wound were not present on the section.
706
ELOY ET AL.
At the 14th postoperative day (Fig. 4), wounds covered with the polymer membrane showed an epithelium which covered a large area of the scar zone as seen in Table 111. The newly formed dermis was not so thick as that of the healthy tissue, but in the epithelialized zones fibroblasts were taking the place of inflammatory cells. On the wounds treated with gauze, the epithelialization process was very slight, the dermis was highly vascularized and infiltrated by macrophages, giant cells localized around gauze fiber debris or foreign bodies. At the 21st postoperative day (Fig. 5), the epithelialization of the polymer membrane treated wounds was complete (Table 111). This epithelium was well organized and covered on its internal side by a horny layer, whereas the dermis was rich in fibroblasts, all oriented along the same axis parallel to the skin surface. Formation of epidermal adnexa could be observed toward the edges of the scar. Some residual inflammatory cells and giant cells could be observed, and the thickness of scar tissue was identical to that of a healthy tissue. At this stage, wounds covered with gauze were still not epithelialized (Table 111) and layers of fibrin, hemorrhagic and strongly inflammatory dermis were still present, with macrophages and giant cells.
In vivo degradation of the membrane implanted subcutaneously in rats: histological evaluation (Fig. 6) At the 15th postimplantation day, a well organized fibrocellular capsule in which fibroblasts were predominant, surrounded the membrane. The membrane was generally not altered, only small defects were detected at the edges of some implants, probably related to initial irregularities rather than to a process of degradation. At the 30th postimplantation day, the tissular reaction around the implants was unmodified while macrophage cell density was decreased, and fibroblast density increased. At the 62nd postimplantation day, no degradation process was detectable on histological sections of the implant. There was no increase in the number of giant cells and no increase of membrane defects or enlargement was observed. At the 90th postimplantation day, the inflammatory component was still decreasing and neovascularization was reduced; some macrophages and giant cells were still visible in small foci at the poles of the implant, which could present at these places slight irregularities.
In vivo biodegradation of the membrane: physicochemical analyses Each animal had been statistically implanted with the same amount of the membrane of about 96 mg with a relative standard deviation of 2.6% to 3.4%.
SYNTHETIC AMINO ACID COPOLYMER DRESSING
Figure 4. Histology of wounds after 14 days of treatment. (A,B) Wound covered with the membrane after 14 days of contact. The middle of the wound is not epithelialized. The dermis is thick but still inflammatory. (C,D) Control wound covered with gauze after 14 days of contact. The middle of the wound is not epithelialized. The thickness of the scar tissue is reduced. The dermis is infiltrated by inflammatory cells and giant cells. Bar = 100 pm.
707
708
ELOY ET AL.
Figure 5. Histology of wounds after 21 days of treatment. (A,B) Wound covered with the membrane after 21 days of contact. The scar zone is completely epithelialized. The dermis is rich in fibroblasts with a beginning of orientation. There are still some large mononuclear cells. (C,D) Wound covered with gauze after 21 days of contact. The wound has not yet healed, there remains a nonepithelialized and hemorrhagic zone. The dermis is inflammatory, characterized by its richness in macrophages giant cells and neovessels. Bar = 100 pm.
SYNTHETIC AMINO ACID COPOLYMER DRESSING
Figure 6. Biodegradation study of the polymeric membrane implanted subcutaneously in rats. (A,B) After 15 days of implantation the implant is surrounded by a fibrocellular capsule in which fibroblasts are predominant. There are some macrophages located at the pole of the implant. The membrane is not altered by cellular penetration. (C,D) After 90 days of implantation no degradation process is detectable. The small focus of inflammatory cells, visible in (C) in one pole of the implant, is probably related to an initial irregularity of the membrane. The implant is surrounded by a fibrocellular tissue. Bar = 100 pm.
709
ELOY ET AL.
710
The amount of polymer expressed in mg per implant was 2.69 k 0.175 (mean ? SD of 9 determinations). The amount of polymer in test implants is given in Table IV and the results show that polymer amounts in the membrane recovered after 15, 30, 62, 90 days of subcutaneous implantation in rat did not differ significantly from polymer amount in blank implant. The same characteristic infrared absorption bands were observed in test and blank implants.
DISCUSSION
Laminates composed of polypeptides have already been suggested as a wound covering about 15 years ago but, as viewed by Aiba et a1.7 Feijen et al.; Anderson et al.," the physical and chemical characteristics of the polypeptides were poorly understood. They were suggested as potential for the scaffolding of cell adhesion and/or growth. A copoly (leucine methylglutamate) membrane with specific properties was developed in order to comply with certain requirements of an "optimized wound covering." Since a limited initial adherence to defects can reduce pain, as well as control infection, and this optimizes the rate of healing,'" a relative elasticity and adhesion to the wound were searched for. The membrane adheres over moist areas and as the healing process develops adhesion of the membrane to the granulation tissue is progressively replaced by epithelialization and elimination of the membrane. The membrane which is not biodegradable, has never been found inside the wound, by histology. The second specification for this membrane is a relative permeability to water vapor. Normal skin allows water vapor to pass at a rate of approximately 8.5 g/m2/hour while retaining proteins and electrolytes. Although the optimal water vapor transmission of burn dressings has not been established, an excess of permeability can lead to water loss and drying of tissues, whereas absence of permeability induces formation of blisters. The recently developed wound dressing, derived polyurethane membrane, have a permeability ranging between 500 g/m2/24 h for OpSite and 5000 g/m2/24 h for Omiderm." The present polypeptide membrane, by its permeability to water and by the addition of a humectant like PEG, appears to maintain during the first days of dressing an adequate hydration of the membrane, thus reducing the risk of extensive eschar formation. As reported previously for polypropylene glycol'
+
TABLE IV Amount of Polymer (mg/implant) after Increasing Periods of Implantation Before Implantation 2.69
+ 0.175
After 15 Days of Implantation
After 30 Days of Implantation
After 62 Days of Implantation
2.73 f 0.166
2.87 f 0.158
2.70 k 0.398
After YO Days of Implantation 2.65
* 0.336
Results are expressed as the mean f SD in at least 5 different experimental animals
SYNTHETIC AMINO ACID COPOLYMER DRESSING
711
it could be expected that water electrolytes and proteins should reach equilibrium in the liquid component of the membrane with the wound exsudate. This concept providing a controlled adequate microenvironment to the wound is based on the evidence that wound healing is enhanced by occlusive dressings (for review see ref. 12). The effect is not limited to the dermis since occlusion in guinea pig in our study caused a great reduction of the inflammatory response. Especially neutrophils, but also macrophages, appeared earlier but were more rapidly replaced by numerous fibroblasts and collagenous material. An attractive but still not demonstrated hypothesis for the faster epithelialization under occlusive dressings is that they retain wound fluids which contain a variety of factors for connective tissue formation and epidermal migration, in contact with the wound. In this view the polymeric membrane may be considered as a pharmacological agent, absorbing to some extent factors present in the wound but able to keep them available for the different steps of healing. The durability of the membrane may also contribute to this effect by suppressing the need and the drawbacks of a new "non-biologically filled dressing. After a preliminary ~tudy,'""~ which reported the potential interest of the biomaterial, the quantitative data generated in this study suggest that the dermoepithelial repair process is enhanced in the presence of the polymeric membrane considering the following: the excision procedure and size were uniformly standardized and randomized; each animal served as its own control; the control and treated sites were photographed in standard conditions and the only variable was the rate of wound healing; the degree of epithelial proliferation was determined by planimetry for each wound; epithelialization of excised wounds was significantly higher than for control gauze dressings. The polymer fulfils biocompatibility testing: it is not cytotoxic on mouse L 929 fibroblasts, it is nonhemolytic in nitro, it does not induce hypersensitivity in the guinea pig, it is nonantigenic and well tolerated in the intracutaneous test in the rabbit. It is nonmutagenic in ~ i f r 0 . l ~ " ~ Preliminary clinical studies including skin graft donor sites, mesh grafts, skin defects of various origin, partial and full thickness burns treated with the membrane confirm these experimental data. Work is in progress to compare the epithelialization rate and characteristics of different experimental wound models in rats, treated comparatively with other. References 1. R.M. Nalbadian, R.L. Henry, K.W. Balko, D.V. Adams, and N.R. Neuman, "Pluronic F-127 Gel preparation as an artificial skin in the treatment of third-degree burns in pigs," J. Biomed. Mater. Res., 21, 1135-1148 (1987). 2. B . A . Pruitt and N.S. Levine, "Characteristics and uses of biologic dressings and skin substitutes," Arch. Surg., 119, 314-322 (1984). 3. K. J. Quinn, J. M., Courtney, J. H. Evans, J. D. S. Gaylor, and W. H. Reid, "Principles of burn dressing," Biomaterials, 6, 369-377 (1985).
ELOY ET AL.
712
8. 9.
10.
11.
12.
13. 14.
C. D. Hinman and H. Maibach, ”Effect of air exposure and occlusion on experimental human skin wounds,” Nature, 200, 377-378 (1963). A. Brack and Y. Trudelle, “Molecular weight determination of polypeptides from viscosity measurements,” Polym. Com. 26, 369-370 (1985). A. Brack and G. Spach, ”Multiconformational synthetic polypeptide,” 1. A m . Chem. Soc., 103,6319-6329 (1981). S . Aiba, N. Minoura, Y. Fujiwara, S. Yamada, and T. Nakagawa, “Laminates composed of polypeptides and elastomers as a burn wound covering physicochemical properties,” Biomuterials, 6, 290-296 (1985). J. Feijen, W. L. Sederel, and T. Beugeling, ”Synthetic polypeptides as biomaterials,” Proc. Euro. Soc. Artif. Organs, 1, 37-42 (1975). J. M. Anderson, D.E Gibbons, R. L. Martin, A. Miltner, and R. Woods, ”The potential of poly a amino acids as biomaterials,” J. Biomed. Muter. Res., 5, 197-207 (1974). M. J. Tavis, J. Thornton, R. Danet, and R. H. Barlett, ”Current status of skin substitutes,” Surg Clin. of North Am., 58, 1233-1248 (1978). D. Behar, M. Juszynski, N. Ben Hur, J. Golan, A Eldad, J. Tuchman, N. Sterenberg, and B. Rudensky, “Omiderm a new synthetic wound covering: physical properties and drug permeability studies,” J. Biomed. Mater. Res., 20, 731-738 (1986). V. Falanga, ”Occlusive wound dressing. Why, When, Which?,” Arch. Dermatol., 124, 872-877 (1988). R. Eloy and N. Weill, ”Etude prhclinique du PA 286, substitut cutan6 temporaire,” XPme CongrPs de la Societe Franqaise d’Etude et de Traitement des Briilures, 28-30 Septembre 1989, Metz, France. R. Eloy and N. Weill, “Preclinical studies of PA 186, a temporary skin substitute.” Td Congress of the European Burns Association, October 4-7, 1989. Prague, Czechoslovakia.
Received September 17,1990 Accepted October 31,1991
