- Email: [email protected]
Porous-Type Tracheal Prosthesis Sealed With Collagen Sponge
Porous-Type Tracheal Prosthesis Sealed With Collagen Sponge Masayoshi Teramachi, MD, Tatsuo Nakamura, MD, Yasumichi Yamamoto, MD, Tetsuya Kiyotani, MD, Yukinobu Takimoto, MD, and Yasuhiko Shimizu, MD Department of Artificial Organs, Research Center for Biomedical Engineering, Kyoto University, Kyoto, Japan
Background. Reconstruction of a long section of the trachea is clinically problematic. Tracheal reconstructions using prostheses have met with limited success due to local infection, hemorrhage, luminal stenosis and prosthesis dislocation. Methods. We have designed a porous type of tracheal prosthesis in which the mesh is sealed with collagen sponge. We used this prosthesis (50 mm in length) to reconstruct the cervical trachea in 10 mongrel dogs and evaluated its efficacy. Results. One dog died due to an accident with anesthesia at 6 weeks and 1 of suffocation at 10 weeks. The other 8 dogs had an uneventful postoperative course
until they were killed between 6 and 24 months after implantation. At sacrifice, all the prostheses had become completely incorporated into the host. Microscopic examination revealed advanced formation of a new epithelial lining in 1 dog at 6 months, and a confluent epithelial lining was observed in another dog at 12 months. Central stenosis was not significant in any of the animals. Conclusions. This tracheal prosthesis gives good results in canine tracheal reconstruction, and appears very promising for the clinical repair of tracheal defects.
T
Material and Methods
racheal reconstruction, for which primary end-to-end anastomosis cannot be used, has remained an unsolved clinical problem. Tracheal reconstructions using a prosthesis have met with limited success due to local infection, anastomotic leakage in the early stages after implantation, hemorrhage, granuloma formation, or eventual stenosis and dislocation of the prosthesis in the later stages [1–3]. We attribute these problems to the poor biocompatibility of prosthetic materials, and have been applying amorphous collagen on synthetic polymers to enhance the biocompatibility of the prosthesis. Collagen is the most abundant protein in mammals, and has been used for a variety of biomedical applications because of its good biocompatibility. Recently, it has become clear that collagen matrix not only functions as the scaffolding for cells but also plays active roles in the development, migration, and proliferation of cells as an extracellular matrix of the tissue [4]. We have designed a porous type of tracheal prosthesis in which the mesh was sealed with collagen sponge instead of amorphous collagen. In the present study, we evaluated the results of animal experiments with a length of 50 mm of our collagen-sponge–sealed prosthesis in a canine tracheal replacement model with long-term observation up to 24 months.
Accepted for publication March 31, 1997. Address reprint requests to Dr Teramachi, Research Center for Biomedical Engineering, Kyoto University, 53 Kawahara-cho, Shogoin, Sakyo-ku, Kyoto 606, Japan (e-mail: [email protected]).
© 1997 by The Society of Thoracic Surgeons Published by Elsevier Science Inc
(Ann Thorac Surg 1997;64:965–9) © 1997 by The Society of Thoracic Surgeons
Collagen Solution Collagen was extracted from pig skin and treated with a method similar to that described in our previous study [5]. It was supplied by Nippon Meat Packers Inc, Ibaraki, Japan. This purified collagen, which consisted of type I collagen (70% to 80%) and type III collagen, was dissolved in hydrochloric acid solution with a pH of 3.0, so that its concentration became 1.3%.
Prosthesis The bulk structure of the prosthesis was the same as we used in the previous study [5, 6]. It consisted of a fine Marlex mesh (CR Bard Inc, Billerica, MA) cylinder, 18 to 22 mm in inner diameter and 50 mm in length, which was reinforced with a continuous polypropylene spiral. The pore size of the mesh was about 260 mm. A polypropylene spiral, the diameter of which was 1 mm, was attached to the external surface of the middle 40 mm of the mesh cylinder by thermal melt-bonding, and further fixed with 6-0 Prolene sutures (Ethicon Inc, Somerville, NJ) (Fig 1). The mesh cylinder was first exposed to corona discharge at 9 kV for 15 minutes to make the polymer surface hydrophilic. Then it was placed in a Teflon tube with an inner diameter of 30 mm. Then another 6 mm tube was inserted in their center with its axis parallel to the outer tube to make a space in the mesh cylinder. The 1.3% collagen solution, which had been stirred to make it foamy, was poured into the space between the inner and outer tubes and then freeze-dried. In this process the collagen became a porous sponge with a pore size range of 100 to 500 mm. This mesh cylinder sealed with collagen 0003-4975/97/$17.00 PII S0003-4975(97)00755-8
966
TERAMACHI ET AL COLLAGEN SPONGE-SEALED TRACHEAL PROSTHESIS
Ann Thorac Surg 1997;64:965–9
Fig 1. Tracheal prosthesis. (Upper left) The mesh cylinder reinforced with a continuous polypropylene spiral is the bulk structure of this prosthesis. (Lower left) Outer view of the prosthesis, after it has been sealed with collagen sponge. Views of cross-section (upper right) and longitudinal section (lower right) of the prosthesis are shown. The collagen sponge seals both the inner and outer surfaces of the mesh cylinder.
sponge was heated at 105°C under vacuum for 12 hours to generate moderate crosslinkage in the collagen molecules.
bital. The prosthesis was resected en bloc with both the distal and proximal sites of the native trachea, and after macroscopic examination, the specimens were prepared
Animal Experiment Ten mongrel dogs, weighing 12 to 19 kg, were anesthetized with an intramuscular injection of ketamine hydrochloride (10 mg/kg) and xylazine hydrochloride (4 mg/kg), and anesthesia was maintained with intravenous injections of sodium pentobarbital (7 mg/kg). Under spontaneous respiration, the cervical trachea was exposed through a midline incision in the neck. A 45-mm-long segment of the trachea including six to nine tracheal cartilage rings was resected circumferentially in the lower cervical trachea. A tracheal prosthesis was well soaked with canine blood, and then both cut ends of the native trachea were invaginated into the prosthesis and anastomosed using interrupted suturing with 3-0 Vicryl (Ethicon Inc). After the distal anastomosis was completed, a silicone tube (1 mm thick, outer diameter 4 mm less than the inner diameter of the prosthesis) was inserted into the prosthesis lumen and fixed with a 4-0 PDS suture (Ethicon Inc) (Fig 2). This silicone tube was designed as a stent for promoting transport of secretory products and protecting the prosthesis from local infection until it was covered with host tissue. The proximal stump of the trachea was then invaginated into the other end of the prosthesis and anastomosed in the same manner. After both anastomoses were completed, the cervical incision was closed by layers. Ampicillin was given intravenously at a dose of 1 g on the day of the operation and daily in an oral dose of 250 mg for 8 weeks. Bronchoscopic examinations were performed periodically under general anesthesia. The silicone tube was removed 8 weeks after the reconstruction using the forceps of a bronchofiberscope. For histologic examinations, the animals were killed at intervals of 6 to 24 months with an overdose injection of sodium pentobar-
Fig 2. (Top) Photograph after the tracheal was replaced with a tracheal prosthesis. (Bottom) Schema of the tracheal replacement with a prosthesis: a tracheal prosthesis is invaginated outside of the cut ends of the dog’s trachea. A silicone tube is inserted into the lumen of the prosthesis.
Ann Thorac Surg 1997;64:965–9
967
TERAMACHI ET AL COLLAGEN SPONGE-SEALED TRACHEAL PROSTHESIS
Table 1. Tracheal Reconstructions Removed Rings
Stenosis
Ulceration
Prognosis
Survival (mos)
1
9
2
Small
2
8
2
2
83%
Killed
6
44%
Killed
6
3
8
2
4
6
2
2
45%
Killed
24
Small
10%
Dieda
5
8
2
2
100%
Killed
12
6
9
7
8
1
2
,10%
Killed
12
2
2
42%
Killed
8
6
9
2
Large
,10%
Killed
12
9
8
1
2
,10%
Killed
12
10
8
1
2
0%
Diedb
Dog No.
a
Died of an anesthetic accident.
b
Epithelialization
1.5
2.5
Died of suffocation.
for light microscopic and scanning electron microscopic examinations. All animals received humane care in compliance with the “Guide for the Care and Use of Laboratory Animals” published by the National Institutes of Health (NIH publication 85-23, revised 1985).
Results Neither operative death nor dehiscence of the tracheal prosthesis occurred in any animal. Two dogs died within 3 months of tracheal reconstruction. One died of an anesthetic accident during bronchoscopic examination at 6 weeks, and the other of suffocation by sputum at 10 weeks. Eight dogs were killed between 6 and 24 months; 3 dogs at 6 months, 4 at 12 months, and 1 at 24 months (Table 1).
Bronchoscopic Examination Before the removal of the silicone stent, a circumferential stenosis was recognized in 3 dogs just proximal to the stent 4 weeks after reconstruction, and 1 of these dogs died of suffocation at 10 weeks. This animal was the only one in this study that died of complications directly related to the prosthesis. When the silicone tube was removed 8 weeks after reconstruction, the luminal surface of the tracheal prosthesis was almost completely covered with host tissue. The luminal surface was lustrous and was still reddish at this time. Significant stenosis was not observed except in 3 dogs in which circumferential stenosis occurred at the proximal site before the removal of the silicone stent. A large area of the mesh was exposed to the tracheal lumen (we call this “ulceration”) in another dog even at 12 months.
to which the prosthesis had adhered. In 6 dogs, the inner diameters of the prostheses remained at more than twothirds of the diameter of the adjacent native trachea. The luminal surface was almost entirely covered with the host tissue. In 3 dogs, however, ulceration was observed (Fig 3). Light microscopic examination showed that connective tissue, including small vessels, had invaded into the mesh pores and that a layer of connective tissue had covered the prosthesis. The collagen sponge of the prosthesis was totally absorbed even at 6 weeks. Neoepithelium extended over the connective tissue layer from both tracheal stumps (Fig 4). Varying degrees of epithelial lining were seen to have formed on the prosthesis surface (see Table 1). Although ciliated columnar epithelium was observed near the anastomoses (Fig 5A), only squamous epithelium was recognized at the center of the prosthesis (Fig 5B). Scanning electron microscopic examination showed that the regenerated epithelial cells near the anastomoses had cilia similar to those of normal tracheal epithelium.
Comment Reconstruction of circumferential tracheal defects is performed after tissue extirpation in cancer or resection of
Histologic Evaluation In 8 dogs that survived more than 3 months after reconstruction, the postoperative courses were uneventful. The prostheses were incorporated into the native trachea and adhered to the surrounding organs, especially to the esophagus. However, no erosion was observed on the vessels
Fig 3. Macroscopic finding of the inner surface of the extirpated prosthesis of dog 8, which had the largest ulceration of all the dogs at 12 months. The inner surface showed a lustrous appearance with a large ulceration (arrowheads).
968
TERAMACHI ET AL COLLAGEN SPONGE-SEALED TRACHEAL PROSTHESIS
Fig 4. Microscopic finding of the mucosa regenerated on the prosthesis at 12 months: Neomucosa with epithelial lining was regenerated on the mesh of the prosthesis. (P 5 polypropylene mesh.) (Hematoxylin and eosin; original magnification, 310.)
stenotic regions, and an end-to-end anastomosis is now considered to be the method of choice for defects of up to approximately 6 cm in adults [7–9]. However, in the case of larger defects, another method of reconstruction is required. Researchers encouraged by the success of artificial vascular grafts have developed a variety of tracheal prostheses using synthetic biomaterials. However, prosthetic tracheal reconstructions have met with only limited success because of problems like local infection, anastomotic leakage, dislocation, granuloma formation, and stenosis or rupture of adjacent vessels [1–3]. Such potentially fatal complications and the associated miserable results preclude the clinical use of such reconstructions at present [1–3, 10 –13]. Recently, we produced a tracheal prosthesis made from collagen-grafted polypropylene mesh and designed to be completely covered with host tissue. Animal experiments revealed that it provided high biocompatibility and seemed to overcome the major serious complications such as anastomotic leakage or local infection [5, 6]. However, the problems of luminal stenosis and ulceration (ie, the exposure of mesh on the luminal surface) have not yet been solved. Because stenoses have often been seen in the middle part of the mesh-type trachea [6, 10 –12], we speculated that the stenosis was attributable to a lack of epithelial lining and consequently excessive granuloma formation. In addition, we speculated that ulceration might occur in places where the implanted collagen had been absorbed before the tissue fully invaded the pores of the mesh. In our laboratory, we have been investigating the application of collagen composite to the surface of synthetic polymers in an attempt to enhance their biocompatibility. Until recently, collagen was thought to serve mainly as a relatively inert scaffolding to stabilize the physical structure of tissue. However, progress in molecular biology has clarified that the collagen matrix plays a far more active role in regulating the behavior of the cells that contact it by influencing their development, migration, and proliferation [14]. Moreover, it was reported
Ann Thorac Surg 1997;64:965–9
that the three-dimensional structure of the matrix had advantages of not only providing a better template for cell migration, but also enhancing the synthesis of collagen and other proteins by migrated cells [15, 16]. We, therefore, used collagen sponge, which had a threedimensional porous structure, instead of the amorphous collagen used in our previous study. We have already applied collagen sponge as a bulk material for an artificial esophagus [17]. In animal experiments using dogs, we confirmed that collagen sponge on our artificial esophagus promoted tissue self-repair, namely induced fibroblast invasion within 2 weeks and complete epithelial lining with mature submucosal tissue including muscle cells within 4 weeks in 5-cm replacement [18]. Because collagen has a high affinity to platelets, collagen soaked with blood absorbs a large number of platelets. In particular, collagen sponge can contain much more blood than the amorphous collagen used in the previous prosthesis. Platelets release various factors that stimulate wound healing, the most well known of which is plateletderived growth factor. Platelet-derived growth factor
Fig 5. Microscopic finding with a higher magnification near the anastomoses. (A) The luminal surface was covered with ciliated columnar epithelium. (B) Squamous epithelium was only observed near the center. (Both, hematoxylin and eosin staining; original magnification, 350.)
Ann Thorac Surg 1997;64:965–9
liberated from blood clots in the body plays a major role in stimulating cell division during wound healing [19, 20]. Thus, it may be that collagen sponge not only serves as a scaffold for tissue regeneration but also stimulates cell proliferation through the surrounding clot formation. In the present study, we confirmed that collagen sponge, which was soaked with blood, could seal the mesh-type tracheal prosthesis and provide adequate tissue proliferation into the mesh as a template for the formation of connective tissue. No significant central stenosis has occurred at the replaced part in any dogs. Microscopically, confluent epithelization was observed in 1 dog at 12 months. These findings indicate that this prosthesis has high biocompatibility and may inhibit excessive formation of granulation tissue even where epithelial lining is lacking. Stenoses were observed in 3 dogs at the sites of the anastomosis in this study. Anastomotic stenoses were not observed in the previous study in which the silicone stents were removed at 4 weeks [5]. Because the silicone stents were left in place for 8 weeks in this study, the stimulation evoked by the stent, in particular by its edge, might have induced granulation in the tracheal mucosa. Moreover, collagen sponge was easily detached from the mesh cylinder in the prosthesis when handled carelessly, and extreme care was needed during the anastomosis procedure. Therefore, these stenoses might have occurred not as a result of the properties of the prosthesis, but owing to detachment of the sponge by the surgical procedures. With the use of this prosthesis, ulceration was still observed in some cases. Although the incidence and extent of ulceration were less (3 in 8 dogs) than those (9 in 17 dogs) observed in the former studies using amorphous collagen [6], a large ulceration appeared in 1 of 8 dogs. Judging from the fact that ulceration caused no significant complications even after long implantation in our previous studies, a small ulceration on the luminal surface may not be associated with any serious complications. Nevertheless, we consider that any exposure of the mesh should be avoided because it represents a potential source of infection. The present study has revealed that collagen sponge on the tracheal prosthesis provided air-tightness in the early stage, promoted tissue regeneration into the mesh, and overcame stenosis of the prosthesis lumen even in the longer 50-mm replacement. This prosthesis is very promising for the repair of tracheal defects, with further improvements in promoting the epithelial lining.
TERAMACHI ET AL COLLAGEN SPONGE-SEALED TRACHEAL PROSTHESIS
969
References 1. Neville WE, Bolanowski PJ, Kotia GG. Clinical experience with the silicone tracheal prosthesis. J Thorac Cardiovasc Surg 1990;99:604–13. 2. Toomes H, Mickisch G, Vogt-Moykopf I. Experiences with prosthetic reconstruction of the trachea and bifurcation. Thorax 1985;40:32–7. 3. Deslauriers J, Ginsberg RJ, Nelems JM, Pearson FG. Innominate artery rupture—a major complication of tracheal surgery. Ann Thorac Surg 1975;20:671–7. 4. Richardt LF, Tomaselli KJ. Extracellular matrix molecules and their receptors: function in neural development. Annu Rev Neurosci 1990;14:531–70. 5. Okumura N, Nakamura T, Natsume T, Tomihata K, Ikada Y, Shimizu Y. Experimental study on a new tracheal prosthesis made from collagen-conjugated mesh. J Thorac Cardiovasc Surg 1994;108:337– 45. 6. Okumura N, Nakamura T, Takimoto Y, et al. A new tracheal prosthesis made from collagen grafted mesh. ASAIO J 1993; 39:M475–9. 7. Grillo HC. Circumferential resection and reconstruction of the mediastinal and cervical trachea. Ann Surg 1965;162: 374– 88. 8. Mulliken JB, Grillo HC. The limits of tracheal resection with primary anastomosis. J Thorac Cardiovasc Surg 1968;55:418–21. 9. Maassen W, Greschuchna D, Vogt-Moykopf I, Toomes H, Lu¨llig H. Tracheal resection—state of the art. Thorac Cardiovasc Surg 1985;33:2–7. 10. Pearson FG, Henderson RD, Gross AE, Ginsberg RJ, Stone RM. The reconstruction of circumferential tracheal defects with a porous prosthesis. An experimental and clinical study using heavy Marlex mesh. J Thorac Cardiovasc Surg 1968;55:605–16. 11. Jorge RG, Ramos AS-P, de Guevara ACL, Huedo FM, de Vega MG, Romeu FP. Experimental study of a new porous tracheal prosthesis. Ann Thorac Surg 1990;50:281–7. 12. Jacobs JR. Investigations into tracheal prosthetic reconstruction. Laryngoscope 1988;98:1239– 45. 13. Cull DL, Lally KP, Mair EA, Daidone M, Parsons DS. Tracheal reconstruction with polytetrafluoroethylene graft in dogs. Ann Thorac Surg 1990;50:899 –901. 14. Birk DE, Silver FH, Trelstad RL. Matrix assembly. In: Hay ED, ed. Cell biology of extracellular matrix. 2nd ed. New York: Plenum, 1991:221–54. 15. Berthod F, Hayek D, Damour O, Collombel C. Collagen synthesis by fibroblasts cultured within a collagen sponge. Biomaterials 1993;14:749–50. 16. Bellamkonda R, Ranieri JP, Bouche N, Aebischer P. Hydrogelbased three dimensional matrix for neural cells. J Biomed Mater Res 1995;29:663–71. 17. Natsume T, Ike O, Okada T, Takimoto Y, Shimizu Y, Ikada Y. Porous collagen sponge for esophageal replacement. J Biomed Mater Res 1993;27:867–75. 18. Takimoto Y, Okumura N, Nakamura T, Natsume T, Shimizu Y. Long-term follow-up of the experimental replacement of the esophagus with a collagen-silicone composite tube. ASAIO J 1993;39:M736 –9. 19. Cross M, Dexter TM. Growth factors in development, transformation, and tumorigenesis. Cell 1991;64:271– 80. 20. Heldin C-H. Structural and functional studies on plateletderived growth factor. EMBO J 1992;11:4251–9.
Background. Reconstruction of a long section of the trachea is clinically problematic. Tracheal reconstructions using prostheses have met with limited success due to local infection, hemorrhage, luminal stenosis and prosthesis dislocation. Methods. We have designed a porous type of tracheal prosthesis in which the mesh is sealed with collagen sponge. We used this prosthesis (50 mm in length) to reconstruct the cervical trachea in 10 mongrel dogs and evaluated its efficacy. Results. One dog died due to an accident with anesthesia at 6 weeks and 1 of suffocation at 10 weeks. The other 8 dogs had an uneventful postoperative course
until they were killed between 6 and 24 months after implantation. At sacrifice, all the prostheses had become completely incorporated into the host. Microscopic examination revealed advanced formation of a new epithelial lining in 1 dog at 6 months, and a confluent epithelial lining was observed in another dog at 12 months. Central stenosis was not significant in any of the animals. Conclusions. This tracheal prosthesis gives good results in canine tracheal reconstruction, and appears very promising for the clinical repair of tracheal defects.
T
Material and Methods
racheal reconstruction, for which primary end-to-end anastomosis cannot be used, has remained an unsolved clinical problem. Tracheal reconstructions using a prosthesis have met with limited success due to local infection, anastomotic leakage in the early stages after implantation, hemorrhage, granuloma formation, or eventual stenosis and dislocation of the prosthesis in the later stages [1–3]. We attribute these problems to the poor biocompatibility of prosthetic materials, and have been applying amorphous collagen on synthetic polymers to enhance the biocompatibility of the prosthesis. Collagen is the most abundant protein in mammals, and has been used for a variety of biomedical applications because of its good biocompatibility. Recently, it has become clear that collagen matrix not only functions as the scaffolding for cells but also plays active roles in the development, migration, and proliferation of cells as an extracellular matrix of the tissue [4]. We have designed a porous type of tracheal prosthesis in which the mesh was sealed with collagen sponge instead of amorphous collagen. In the present study, we evaluated the results of animal experiments with a length of 50 mm of our collagen-sponge–sealed prosthesis in a canine tracheal replacement model with long-term observation up to 24 months.
Accepted for publication March 31, 1997. Address reprint requests to Dr Teramachi, Research Center for Biomedical Engineering, Kyoto University, 53 Kawahara-cho, Shogoin, Sakyo-ku, Kyoto 606, Japan (e-mail: [email protected]).
© 1997 by The Society of Thoracic Surgeons Published by Elsevier Science Inc
(Ann Thorac Surg 1997;64:965–9) © 1997 by The Society of Thoracic Surgeons
Collagen Solution Collagen was extracted from pig skin and treated with a method similar to that described in our previous study [5]. It was supplied by Nippon Meat Packers Inc, Ibaraki, Japan. This purified collagen, which consisted of type I collagen (70% to 80%) and type III collagen, was dissolved in hydrochloric acid solution with a pH of 3.0, so that its concentration became 1.3%.
Prosthesis The bulk structure of the prosthesis was the same as we used in the previous study [5, 6]. It consisted of a fine Marlex mesh (CR Bard Inc, Billerica, MA) cylinder, 18 to 22 mm in inner diameter and 50 mm in length, which was reinforced with a continuous polypropylene spiral. The pore size of the mesh was about 260 mm. A polypropylene spiral, the diameter of which was 1 mm, was attached to the external surface of the middle 40 mm of the mesh cylinder by thermal melt-bonding, and further fixed with 6-0 Prolene sutures (Ethicon Inc, Somerville, NJ) (Fig 1). The mesh cylinder was first exposed to corona discharge at 9 kV for 15 minutes to make the polymer surface hydrophilic. Then it was placed in a Teflon tube with an inner diameter of 30 mm. Then another 6 mm tube was inserted in their center with its axis parallel to the outer tube to make a space in the mesh cylinder. The 1.3% collagen solution, which had been stirred to make it foamy, was poured into the space between the inner and outer tubes and then freeze-dried. In this process the collagen became a porous sponge with a pore size range of 100 to 500 mm. This mesh cylinder sealed with collagen 0003-4975/97/$17.00 PII S0003-4975(97)00755-8
966
TERAMACHI ET AL COLLAGEN SPONGE-SEALED TRACHEAL PROSTHESIS
Ann Thorac Surg 1997;64:965–9
Fig 1. Tracheal prosthesis. (Upper left) The mesh cylinder reinforced with a continuous polypropylene spiral is the bulk structure of this prosthesis. (Lower left) Outer view of the prosthesis, after it has been sealed with collagen sponge. Views of cross-section (upper right) and longitudinal section (lower right) of the prosthesis are shown. The collagen sponge seals both the inner and outer surfaces of the mesh cylinder.
sponge was heated at 105°C under vacuum for 12 hours to generate moderate crosslinkage in the collagen molecules.
bital. The prosthesis was resected en bloc with both the distal and proximal sites of the native trachea, and after macroscopic examination, the specimens were prepared
Animal Experiment Ten mongrel dogs, weighing 12 to 19 kg, were anesthetized with an intramuscular injection of ketamine hydrochloride (10 mg/kg) and xylazine hydrochloride (4 mg/kg), and anesthesia was maintained with intravenous injections of sodium pentobarbital (7 mg/kg). Under spontaneous respiration, the cervical trachea was exposed through a midline incision in the neck. A 45-mm-long segment of the trachea including six to nine tracheal cartilage rings was resected circumferentially in the lower cervical trachea. A tracheal prosthesis was well soaked with canine blood, and then both cut ends of the native trachea were invaginated into the prosthesis and anastomosed using interrupted suturing with 3-0 Vicryl (Ethicon Inc). After the distal anastomosis was completed, a silicone tube (1 mm thick, outer diameter 4 mm less than the inner diameter of the prosthesis) was inserted into the prosthesis lumen and fixed with a 4-0 PDS suture (Ethicon Inc) (Fig 2). This silicone tube was designed as a stent for promoting transport of secretory products and protecting the prosthesis from local infection until it was covered with host tissue. The proximal stump of the trachea was then invaginated into the other end of the prosthesis and anastomosed in the same manner. After both anastomoses were completed, the cervical incision was closed by layers. Ampicillin was given intravenously at a dose of 1 g on the day of the operation and daily in an oral dose of 250 mg for 8 weeks. Bronchoscopic examinations were performed periodically under general anesthesia. The silicone tube was removed 8 weeks after the reconstruction using the forceps of a bronchofiberscope. For histologic examinations, the animals were killed at intervals of 6 to 24 months with an overdose injection of sodium pentobar-
Fig 2. (Top) Photograph after the tracheal was replaced with a tracheal prosthesis. (Bottom) Schema of the tracheal replacement with a prosthesis: a tracheal prosthesis is invaginated outside of the cut ends of the dog’s trachea. A silicone tube is inserted into the lumen of the prosthesis.
Ann Thorac Surg 1997;64:965–9
967
TERAMACHI ET AL COLLAGEN SPONGE-SEALED TRACHEAL PROSTHESIS
Table 1. Tracheal Reconstructions Removed Rings
Stenosis
Ulceration
Prognosis
Survival (mos)
1
9
2
Small
2
8
2
2
83%
Killed
6
44%
Killed
6
3
8
2
4
6
2
2
45%
Killed
24
Small
10%
Dieda
5
8
2
2
100%
Killed
12
6
9
7
8
1
2
,10%
Killed
12
2
2
42%
Killed
8
6
9
2
Large
,10%
Killed
12
9
8
1
2
,10%
Killed
12
10
8
1
2
0%
Diedb
Dog No.
a
Died of an anesthetic accident.
b
Epithelialization
1.5
2.5
Died of suffocation.
for light microscopic and scanning electron microscopic examinations. All animals received humane care in compliance with the “Guide for the Care and Use of Laboratory Animals” published by the National Institutes of Health (NIH publication 85-23, revised 1985).
Results Neither operative death nor dehiscence of the tracheal prosthesis occurred in any animal. Two dogs died within 3 months of tracheal reconstruction. One died of an anesthetic accident during bronchoscopic examination at 6 weeks, and the other of suffocation by sputum at 10 weeks. Eight dogs were killed between 6 and 24 months; 3 dogs at 6 months, 4 at 12 months, and 1 at 24 months (Table 1).
Bronchoscopic Examination Before the removal of the silicone stent, a circumferential stenosis was recognized in 3 dogs just proximal to the stent 4 weeks after reconstruction, and 1 of these dogs died of suffocation at 10 weeks. This animal was the only one in this study that died of complications directly related to the prosthesis. When the silicone tube was removed 8 weeks after reconstruction, the luminal surface of the tracheal prosthesis was almost completely covered with host tissue. The luminal surface was lustrous and was still reddish at this time. Significant stenosis was not observed except in 3 dogs in which circumferential stenosis occurred at the proximal site before the removal of the silicone stent. A large area of the mesh was exposed to the tracheal lumen (we call this “ulceration”) in another dog even at 12 months.
to which the prosthesis had adhered. In 6 dogs, the inner diameters of the prostheses remained at more than twothirds of the diameter of the adjacent native trachea. The luminal surface was almost entirely covered with the host tissue. In 3 dogs, however, ulceration was observed (Fig 3). Light microscopic examination showed that connective tissue, including small vessels, had invaded into the mesh pores and that a layer of connective tissue had covered the prosthesis. The collagen sponge of the prosthesis was totally absorbed even at 6 weeks. Neoepithelium extended over the connective tissue layer from both tracheal stumps (Fig 4). Varying degrees of epithelial lining were seen to have formed on the prosthesis surface (see Table 1). Although ciliated columnar epithelium was observed near the anastomoses (Fig 5A), only squamous epithelium was recognized at the center of the prosthesis (Fig 5B). Scanning electron microscopic examination showed that the regenerated epithelial cells near the anastomoses had cilia similar to those of normal tracheal epithelium.
Comment Reconstruction of circumferential tracheal defects is performed after tissue extirpation in cancer or resection of
Histologic Evaluation In 8 dogs that survived more than 3 months after reconstruction, the postoperative courses were uneventful. The prostheses were incorporated into the native trachea and adhered to the surrounding organs, especially to the esophagus. However, no erosion was observed on the vessels
Fig 3. Macroscopic finding of the inner surface of the extirpated prosthesis of dog 8, which had the largest ulceration of all the dogs at 12 months. The inner surface showed a lustrous appearance with a large ulceration (arrowheads).
968
TERAMACHI ET AL COLLAGEN SPONGE-SEALED TRACHEAL PROSTHESIS
Fig 4. Microscopic finding of the mucosa regenerated on the prosthesis at 12 months: Neomucosa with epithelial lining was regenerated on the mesh of the prosthesis. (P 5 polypropylene mesh.) (Hematoxylin and eosin; original magnification, 310.)
stenotic regions, and an end-to-end anastomosis is now considered to be the method of choice for defects of up to approximately 6 cm in adults [7–9]. However, in the case of larger defects, another method of reconstruction is required. Researchers encouraged by the success of artificial vascular grafts have developed a variety of tracheal prostheses using synthetic biomaterials. However, prosthetic tracheal reconstructions have met with only limited success because of problems like local infection, anastomotic leakage, dislocation, granuloma formation, and stenosis or rupture of adjacent vessels [1–3]. Such potentially fatal complications and the associated miserable results preclude the clinical use of such reconstructions at present [1–3, 10 –13]. Recently, we produced a tracheal prosthesis made from collagen-grafted polypropylene mesh and designed to be completely covered with host tissue. Animal experiments revealed that it provided high biocompatibility and seemed to overcome the major serious complications such as anastomotic leakage or local infection [5, 6]. However, the problems of luminal stenosis and ulceration (ie, the exposure of mesh on the luminal surface) have not yet been solved. Because stenoses have often been seen in the middle part of the mesh-type trachea [6, 10 –12], we speculated that the stenosis was attributable to a lack of epithelial lining and consequently excessive granuloma formation. In addition, we speculated that ulceration might occur in places where the implanted collagen had been absorbed before the tissue fully invaded the pores of the mesh. In our laboratory, we have been investigating the application of collagen composite to the surface of synthetic polymers in an attempt to enhance their biocompatibility. Until recently, collagen was thought to serve mainly as a relatively inert scaffolding to stabilize the physical structure of tissue. However, progress in molecular biology has clarified that the collagen matrix plays a far more active role in regulating the behavior of the cells that contact it by influencing their development, migration, and proliferation [14]. Moreover, it was reported
Ann Thorac Surg 1997;64:965–9
that the three-dimensional structure of the matrix had advantages of not only providing a better template for cell migration, but also enhancing the synthesis of collagen and other proteins by migrated cells [15, 16]. We, therefore, used collagen sponge, which had a threedimensional porous structure, instead of the amorphous collagen used in our previous study. We have already applied collagen sponge as a bulk material for an artificial esophagus [17]. In animal experiments using dogs, we confirmed that collagen sponge on our artificial esophagus promoted tissue self-repair, namely induced fibroblast invasion within 2 weeks and complete epithelial lining with mature submucosal tissue including muscle cells within 4 weeks in 5-cm replacement [18]. Because collagen has a high affinity to platelets, collagen soaked with blood absorbs a large number of platelets. In particular, collagen sponge can contain much more blood than the amorphous collagen used in the previous prosthesis. Platelets release various factors that stimulate wound healing, the most well known of which is plateletderived growth factor. Platelet-derived growth factor
Fig 5. Microscopic finding with a higher magnification near the anastomoses. (A) The luminal surface was covered with ciliated columnar epithelium. (B) Squamous epithelium was only observed near the center. (Both, hematoxylin and eosin staining; original magnification, 350.)
Ann Thorac Surg 1997;64:965–9
liberated from blood clots in the body plays a major role in stimulating cell division during wound healing [19, 20]. Thus, it may be that collagen sponge not only serves as a scaffold for tissue regeneration but also stimulates cell proliferation through the surrounding clot formation. In the present study, we confirmed that collagen sponge, which was soaked with blood, could seal the mesh-type tracheal prosthesis and provide adequate tissue proliferation into the mesh as a template for the formation of connective tissue. No significant central stenosis has occurred at the replaced part in any dogs. Microscopically, confluent epithelization was observed in 1 dog at 12 months. These findings indicate that this prosthesis has high biocompatibility and may inhibit excessive formation of granulation tissue even where epithelial lining is lacking. Stenoses were observed in 3 dogs at the sites of the anastomosis in this study. Anastomotic stenoses were not observed in the previous study in which the silicone stents were removed at 4 weeks [5]. Because the silicone stents were left in place for 8 weeks in this study, the stimulation evoked by the stent, in particular by its edge, might have induced granulation in the tracheal mucosa. Moreover, collagen sponge was easily detached from the mesh cylinder in the prosthesis when handled carelessly, and extreme care was needed during the anastomosis procedure. Therefore, these stenoses might have occurred not as a result of the properties of the prosthesis, but owing to detachment of the sponge by the surgical procedures. With the use of this prosthesis, ulceration was still observed in some cases. Although the incidence and extent of ulceration were less (3 in 8 dogs) than those (9 in 17 dogs) observed in the former studies using amorphous collagen [6], a large ulceration appeared in 1 of 8 dogs. Judging from the fact that ulceration caused no significant complications even after long implantation in our previous studies, a small ulceration on the luminal surface may not be associated with any serious complications. Nevertheless, we consider that any exposure of the mesh should be avoided because it represents a potential source of infection. The present study has revealed that collagen sponge on the tracheal prosthesis provided air-tightness in the early stage, promoted tissue regeneration into the mesh, and overcame stenosis of the prosthesis lumen even in the longer 50-mm replacement. This prosthesis is very promising for the repair of tracheal defects, with further improvements in promoting the epithelial lining.
TERAMACHI ET AL COLLAGEN SPONGE-SEALED TRACHEAL PROSTHESIS
969
References 1. Neville WE, Bolanowski PJ, Kotia GG. Clinical experience with the silicone tracheal prosthesis. J Thorac Cardiovasc Surg 1990;99:604–13. 2. Toomes H, Mickisch G, Vogt-Moykopf I. Experiences with prosthetic reconstruction of the trachea and bifurcation. Thorax 1985;40:32–7. 3. Deslauriers J, Ginsberg RJ, Nelems JM, Pearson FG. Innominate artery rupture—a major complication of tracheal surgery. Ann Thorac Surg 1975;20:671–7. 4. Richardt LF, Tomaselli KJ. Extracellular matrix molecules and their receptors: function in neural development. Annu Rev Neurosci 1990;14:531–70. 5. Okumura N, Nakamura T, Natsume T, Tomihata K, Ikada Y, Shimizu Y. Experimental study on a new tracheal prosthesis made from collagen-conjugated mesh. J Thorac Cardiovasc Surg 1994;108:337– 45. 6. Okumura N, Nakamura T, Takimoto Y, et al. A new tracheal prosthesis made from collagen grafted mesh. ASAIO J 1993; 39:M475–9. 7. Grillo HC. Circumferential resection and reconstruction of the mediastinal and cervical trachea. Ann Surg 1965;162: 374– 88. 8. Mulliken JB, Grillo HC. The limits of tracheal resection with primary anastomosis. J Thorac Cardiovasc Surg 1968;55:418–21. 9. Maassen W, Greschuchna D, Vogt-Moykopf I, Toomes H, Lu¨llig H. Tracheal resection—state of the art. Thorac Cardiovasc Surg 1985;33:2–7. 10. Pearson FG, Henderson RD, Gross AE, Ginsberg RJ, Stone RM. The reconstruction of circumferential tracheal defects with a porous prosthesis. An experimental and clinical study using heavy Marlex mesh. J Thorac Cardiovasc Surg 1968;55:605–16. 11. Jorge RG, Ramos AS-P, de Guevara ACL, Huedo FM, de Vega MG, Romeu FP. Experimental study of a new porous tracheal prosthesis. Ann Thorac Surg 1990;50:281–7. 12. Jacobs JR. Investigations into tracheal prosthetic reconstruction. Laryngoscope 1988;98:1239– 45. 13. Cull DL, Lally KP, Mair EA, Daidone M, Parsons DS. Tracheal reconstruction with polytetrafluoroethylene graft in dogs. Ann Thorac Surg 1990;50:899 –901. 14. Birk DE, Silver FH, Trelstad RL. Matrix assembly. In: Hay ED, ed. Cell biology of extracellular matrix. 2nd ed. New York: Plenum, 1991:221–54. 15. Berthod F, Hayek D, Damour O, Collombel C. Collagen synthesis by fibroblasts cultured within a collagen sponge. Biomaterials 1993;14:749–50. 16. Bellamkonda R, Ranieri JP, Bouche N, Aebischer P. Hydrogelbased three dimensional matrix for neural cells. J Biomed Mater Res 1995;29:663–71. 17. Natsume T, Ike O, Okada T, Takimoto Y, Shimizu Y, Ikada Y. Porous collagen sponge for esophageal replacement. J Biomed Mater Res 1993;27:867–75. 18. Takimoto Y, Okumura N, Nakamura T, Natsume T, Shimizu Y. Long-term follow-up of the experimental replacement of the esophagus with a collagen-silicone composite tube. ASAIO J 1993;39:M736 –9. 19. Cross M, Dexter TM. Growth factors in development, transformation, and tumorigenesis. Cell 1991;64:271– 80. 20. Heldin C-H. Structural and functional studies on plateletderived growth factor. EMBO J 1992;11:4251–9.