Secondary Logo

Journal Logo


A Gelatin Coated Collagen-Polyglycolic Acid Composite Membrane as a Dural Substitute

Matsumoto, Kazuya; Nakamura, Tatsuo; Fukuda, Seijun; Sekine, Takashi; Ueda, Hiroki; Shimizu, Yasuhiko

Author Information
  • Free


Although many experimental and clinical studies have been performed to identify a suitable material for repairing defects of the dura mater, no ideal dural substitute is currently available. Formerly, lyophilized human dura mater was widely used; however, its use is now contraindicated because of the risk of infection with prion diseases such as Creutzfeldt-Jakob disease. 1, 2 Therefore, nonbiodegradable materials such as expanded polytetrafluoroethylene (ePTFE) are used instead 3; however, nonbiodegradable materials have the potential to cause delayed inflammatory reactions. 4

Consequently, we have designed a novel biodegradable composite membrane as a dural substitute. In this study, we examined the microstructure and mechanical properties of our membrane as a surgical material, and evaluated its potential by implantation experiments in dogs.

Materials and Methods

Preparation of the Composite Membrane

Collagen extracted from the skin of Japanese pigs by enzymatic treatment was used. This consists mainly of type I collagen and has very low antigenicity because the telopeptide of the collagen molecule is removed by the enzymatic treatment.

A 1% w/v collagen hydrochloride solution was placed in a cast with a depth of 18 mm, and a polyglycolic acid (PGA) mesh sheet (Figure 1) that had been exposed to a plasma discharge at 9 kV for 5 min was then immersed in the collagen solution to a depth of 9 mm, and the cast and contents were frozen at −20°C. A 2.5% gelatin solution, whipped by stirring at 5,000 rpm in a refrigerated homogenizer for 10 min, was poured onto the surface to a depth of 4 mm, and the membrane was frozen again at −20°C. It was then freeze-dried for 72 h at 6.7 × 10−2 Pa and pressed at 500 kg/cm2. Finally, the membrane was heated to 140°C under 1 × 101 Pa for 24 h to cross-link 5 the collagen fibers and gelatin sponge (Eyele Vacuum Oven VOS-300SD; Tokyo Rikakikai, Ltd., Tokyo, Japan). The thickness of the final membrane was approximately 0.8 mm (Figure 2). For in vivo implantation, the membrane was sterilized with ethylene oxide gas before surgery.

Figure 1
Figure 1:
Macroscopic appearance of the polyglycolic acid (PGA) mesh sheet.
Figure 2
Figure 2:
Macroscopic appearance of our composite membrane. The gelatin coated surface (left) and the collagen layer surface without the gelatin coating (right).

Observation of the Microstructure

The cross-sectional microstructure of the membrane was observed by scanning electron microscopy (SEM).

Measurements of Mechanical Strength

The mechanical strength of the membrane was measured with the following two methods 6, 7 : under dry or wet (immersed in saline at 37°C for 1 min) conditions. The test specimen was 10 mm wide and 20 mm long. One measurement was the tearing stress test (Figure 3A). Both edges of the specimen were fastened to the chucks of the test machine, and the maximum stress (N/cm) was determined. The other measurement was the one-point suspension test (Figure 3B). The sample was fixed with a stitch placed 5 mm in from one edge, the opposite edge was fastened to the chuck of the test machine, and the maximum stretching strength (N) was measured.

Figure 3
Figure 3:
Method for measurement of mechanical properties by tearing stress test (A) and one-point suspension test (B).

Peak strength was also measured using a digital push-pull gauge (model 1356, model 9200; Aikoh Engineering, Ltd., Nagoya, Japan) at a cross-head speed of 5 mm/min. Each measurement was repeated five times, and the mean value and SD were calculated.

In Vivo Implantation Test

Six adult beagle dogs weighing 8.0 to 14.0 kg were used. The dogs were anesthetized by intramuscular administration of ketamine hydrochloride at a dose of 15 mg/kg. Anesthesia was maintained by inhalation of 70% N2O (30% 02) and 1.0% halothane through an endotracheal tube. A unilateral frontoparietal incision was made, and the temporal muscle was retracted. A 40 × 30 mm frontoparietal craniotomy was performed, and a 25 × 15 mm piece of dura mater was carefully excised without causing arachnoidal or cortical damage (Figure 4, left). Afterward, a piece of composite membrane that was 5 mm larger than the dural defect on each side was placed in position so that the gelatin coated surface faced the cerebral cortex. The membrane was secured by interrupted polypropylene sutures (5-0 Prolene; Ethicon, Inc., Somerville, NJ) (Figure 4, right). The cranial bone defect was left open, and the temporal muscle, fascia, and skin were closed. All surgical experiments were performed according to the Guidelines of the Animal Experiment Committee, Kyoto University (1983).

Figure 4
Figure 4:
Macroscopic views of the treated site before (left) and after (right) implantation of our dural substitute.

Two dogs each were killed at 2 and 4 months after surgery, whereas two dogs were observed over 6 months. The dogs were deeply anesthetized with 30 mg/kg intramuscular ketamine hydrochloride and then killed by an overdose of intravenous pentobarbital sodium. The treated area was excised as a single block; the histologic specimen was fixed in 10% formaldehyde solution for 48 h and embedded in paraffin. Sections 1 × 10−6 m thick were cut and stained with hematoxylin-eosin and Masson’s trichrome and then examined under a light microscope.


Microstructure of the Membrane

Figure 5 illustrates the cross-sectional scanning electron microscopic appearance of the composite membrane. The PGA mesh is sandwiched in the layers of collagen sponge and is not exposed to the surface of the membrane. SEM shows that the compressed collagen sponge consists of sheets of thin collagen layers, approximately 6 to 10 layers per 100 μm. One surface of the collagen-PGA membrane is covered with a gelatin sponge layer approximately 250 to 350 × 10−6 m thick.

Figure 5
Figure 5:
Cross-sectional scanning electron microscopic view of the composite membrane. White arrows indicate the PGA mesh.

Mechanical Strength of the Membrane

The results of the mechanical strength measurements are shown in Figure 6. The tear stress was 45.6 ± 5.5 N/cm under dry conditions and 20.2 ± 6.2 N/cm under wet conditions. In the one-point suspension test, the values were 21.0 ± 3.5 N under dry conditions and 12.9 ± 2.8 N under wet conditions.

Figure 6
Figure 6:
The results of the tearing stress test (A) and the one-point suspension test (B).

In Vivo Implantation Test

All animals survived until they were killed or until the end of the observation period. No infections, cerebrospinal fluid leakage, convulsions, or other significant complications were observed in any of the dogs. At 2 months after surgery, the implanted dural substitute was partially absorbed and a thin fibrous membrane had formed at the implant site. At 4 months after surgery, the implanted membrane was completely absorbed and no residual membrane was detectable (Figures 7 and 8). White glossy fibrous membrane resembling host dura mater had regenerated in the implanted area and was firmly united with the host dura. Although adhesion to the cerebral cortex was observed at the circumference where the sutures were exposed, it was easily separable without causing any cortical damage, and there was no adhesion to the cerebral cortex in the central area of the regenerated neomembrane. On the opposite side, moderate adhesion of the regenerated membrane to the temporal muscle was observed.

Figure 7
Figure 7:
Macroscopic appearance of the treated site 4 months after implantation. Arrows indicate the regenerated membrane. No significant adhesion to the cerebral cortex is evident.
Figure 8
Figure 8:
Macroscopic view of a cross-section of the tissue specimen after 4 months. The portion between the two arrows is the regenerated membrane.

Microscopic examination at 2 months after implantation showed that the implanted membrane had partially degraded and that a collagenous membrane approximately 150 to 200 μm thick had regenerated as a continuous sheet below the degraded implant (Figure 9). At 4 months after implantation, the implanted membrane had disappeared completely and was fully replaced by a 450 to 600 μm thick collagenous membrane (Figure 10). This regenerated membranous tissue consisted of dense well-oriented collagen fiber layers containing some small capillaries. No significant inflammation or foreign body reaction was observed in the neomembrane or the cerebral cortex.

Figure 9
Figure 9:
Microscopic view of the regenerated membrane after 2 months (40×, Masson’s trichrome stain). Arrows indicate the residue of the implanted membrane.
Figure 10
Figure 10:
Microscopic view of the regenerated membrane after 4 months (40×, Masson’s trichrome stain). Arrows indicate the regenerated membrane.


An ideal substitute for the dura mater should satisfy the following criteria. First, the materials of the dural substitute must be safe and biocompatible. Second, it must have adequate mechanical properties for use as a surgical membrane. Third, it must prevent cerebrospinal fluid leakage and cortical adhesion. Lastly, it is desirable that the substitute be absorbed after regeneration of the host dura mater.

Our composite membrane is composed of collagen, a polyglycolic acid (PGA) mesh sheet, and gelatin sponge, and these three components are all biodegradable. In the animal experiment, these components were absorbed by 4 months after implantation and a firm neomembrane had regenerated. Collagen is a major component of the extracellular matrix, and it is well known that collagen promotes cellular proliferation and tissue healing, making collagen widely used in surgical prostheses. We have also used extracted collagen for prostheses such as an artificial esophagus, 8 an artificial trachea, 9 and an artificial nerve conduit, 10 and reported its superior ability to promote regeneration of host tissue and organs. In this composite membrane, we also used collagen as a scaffold to promote regeneration of the host dura mater. The collagen used in this study was an atelopeptide collagen of low antigenicity extracted from the skin of Japanese pigs. We selected this source because it is easy to prepare at low cost and because prion diseases have never been reported in these animals. The PGA mesh was used as reinforcement against cutting by sutures, because compressed collagen sponge itself is fragile to suture, especially under wet conditions. Gelatin was used to prevent tissue adhesion until the host dura mater had regenerated. Our preliminary study of artificial peritoneum in rabbits indicated that building a freeze-dried gelatin layer on a membrane surface is effective in preventing tissue adhesion (unpublished observation). In this study, the gelatin layer functioned well to prevent adhesion, because the regenerated membrane did not adhere to the cerebral cortex, which faced the gelatin layer, whereas moderate adhesion was seen at the temporal muscle, which faced the collagen layer. To cross-link the collagen fibers and gelatin, we chose dehydrothermal treatment, because this method avoids the problem of residual cytotoxic agents that may remain after chemical treatment.

In this study, the mechanical properties of the membrane were measured by two methods. The tear stress test estimates the tensile strength of the membrane itself, whereas the one-point suspension test represents the resistance to cutting with a suture. Because dural substitutes are usually exposed to blood or cerebrospinal fluid when actually in use, the measurements under wet conditions are thought to be more relevant than those under dry conditions. The measured mechanical properties indicated that our composite membrane had adequate mechanical strength for suturing as a surgical membrane, even under wet conditions. In fact, in the animal experiment, we were able to suture this membrane to a dural defect without damage to the membrane or technical difficulty.

Although the use of collagen coated Vicryl mesh as a dural substitute has been reported in some studies, 11, 12 the collagen used was a thin film (25 μm thick). It is reported that PGA sometimes causes inflammatory reactions when the coating collagen layers are destroyed too early. 13 Conversely, the PGA mesh in our membrane was sandwiched between thick collagen layers and was not exposed to the membrane surface. Moreover, we suggest that the collagen sponge used in our membrane is superior to collage film in terms of tissue regeneration, because it may have sufficient porosity to allow cell proliferation, although further study is necessary to clarify this point.

In conclusion, our novel composite membrane shows adequate mechanical properties as a surgical membrane and potential for repairing dural defects. This membrane seems to satisfy the criteria required for a dural substitute. The present results are encouraging for future clinical application.


This work was partially supported by a grant from the Japan Society for the Promotion of Science (JSPS-RFTF 96100203).


1. Thadani V, Penar PL, Partington J, et al: Creutzfeldt-Jacob disease probably acquired from a cadaveric dura mater graft. Case report. J Neurosurg 69: 766–769, 1988.
2. Masullo C, Pocchiari M, Macci G, Alema G, Piazza G, Panzera MA: Transmission of Creutzfeldt-Jacob disease by dural cadaveric graft. J Neurosurg 71: 954–955, 1989.
3. Yamagata S, Goto K, Oda Y, Kikuchi H: Clinical experience with expanded polytetrafluoroethylene sheet used as an artificial dura mater. Neural Med Chir (Tokyo) 33: 582–585, 1993.
4. Cohen AR, Aleksik S, Ransohoff J: Inflammatory reaction to synthetic dural substitute. J Neurosurg 70: 633–635, 1989.
5. Weadock KS, Miller EJ, Keuffel EL, Dunn MG: Effect of physical crosslinking methods on collagen-fiber durability in proteolytic solutions. J Biomed Mater Res 32: 221–226, 1996.
6. Yamamoto Y, Ueda H, Matsumoto K, et al: Production of acellular matrix membranes from human amnion for medical application. Jpn J Artif Organs 26: 219–224, 1997.
7. Matsumoto K, Nakamura T, Shimizu Y, et al: A novel surgical material made from collagen with high mechanical strength: A collagen sandwich membrane. ASAIO J 45: 288–292, 1999.
8. Sekine T, Nakamura T, Matsumoto K, et al: Carinal reconstruction with a Y-shaped collagen-conjugated prosthesis. J Thorac Cardiovasc Surg 119: 1162–1168, 2000.
9. Yamamoto Y, Nakamura T, Shimizu Y, et al: Intrathoracic esophageal replacement in the dog with the use of an artificial esophagus composed of a collagen sponge with a double-layered silicone tube. J Thorac Cardiovasc Surg 118: 276–286, 1999.
10. Matsumoto K, Ohnishi K, Kiyotani T, et al: Peripheral nerve regeneration across an 80-mm gap bridged by a polyglycolic acid(PGA)-collagen tube filled with laminin-coated collagen fibers: A histological and electrophysiological evaluation of regenerated nerves. Brain Res 868: 315–328, 2000.
11. San-Galli F, Darrouzet V, Rival J, Baquey C, Ducassou D, Guerin J: Experimental evaluation of a collagen-coated Vicryl mesh as a dural substitute. Neurosurgery 30: 396–401, 1992.
12. Meddings N, Scott R, Bullock R, French DA, Hide TA, Gorham SD: Collagen Vicryl: A new dural prosthesis. Acta Neurochir (Wien) 117: 53–58, 1992.
13. Gudmundsson G, Sogaard I: Complications to the use of Vicryl-collagen dural substitute. Acta Neurochir (Wien) 132: 145–147, 1995.
Copyright © 2001 by the American Society for Artificial Internal Organs