Takafumi Okoshi

Microporous small diameter PVDF-TrFE vascular grafts fabricated by a spray phase inversion technique

Microporous prostheses of 1.5 mm internal diameter were fabricated with a polyvinylidene fluoride-trifluoroethylene (PVDF-TrFE)n co-polymer by the spray phase inversion technique. Some of the grafts were made piezoelectric by poling under a high electrical field. Overall, 24 poled grafts (P) and 24 unpoled grafts (UP) (15-22 mm in length) were implanted in the infrarenal aorta of 48 adult rats. Patency rates in P were 100% (8/8) at 2 days, 100% (8/8) at 2 weeks, 75% (6/8) at 6 months, and 92% total (22 of 24). Patency rates in UP were 100% (8/8) at 2 days, 63% (5/8) at 2 weeks, 100% (8/8) at 6 months, and 88% total (21 of 24). Thus there was no significant difference in patency between the two types of grafts. Both showed similar macroscopic and microscopic findings. At 2 days, fibrin deposition was somewhat heavier on the poled grafts, but no difference in surface platelet deposition could be detected. Endothelialization was observed from both anastomoses at 2 weeks and was almost complete at 6 months. The excellent biocompatibility of PVDF-TrFE and the microporous structure of the grafts were probably the dominant factors in success with these grafts. Although piezoelectric activity in excised cleaned poled prostheses remained significantly higher than that in the control UP, the charges developed may have been too small to exert a biologic effect, either because of insufficient dipole orientation or inadequate mechanical deformation.

Penetrating micropores increase patency and achieve extensive endothelialization in small diameter polymer skin coated vascular grafts.

This article points to the importance of penetrating micropores through the graft wall to minimize thrombosis and to enhance endothelialization in small diameter polymer skin coated vascular grafts. Four types of spongy polyurethane-polydimethylsiloxane vascular grafts (PUG) fabricated by a spray, phase-inversion technique, 1.5 mm inner diameter, 1.5-1.9 cm in length, were implanted end-to-end in the infrarenal aorta of 26 adult rats. Some had a continuous inner skin and a hydraulic permeability (HP) of 0 ml/min/cm2/ 120 mmHg (PUG-S-O). Some had an inner skin with varying amounts of isolated penetrating micropores and a mean hydraulic permeability of 11 (PUG-S-11), 37 (PUG-S-37), or 58 ml/min/cm2/120 mmHg (PUG-S-58). Twelve PUG-S-O, 6 PUG-S-11, 4 PUG-S-11, and 4 PUG-S-58 were evaluated between 2 hr and 3 months after implantation. All PUG-S-O occluded soon after implantation. The PUG that had a HP of more than 11 ml/min/cm2 showed acceptable patency. However, endothelialization was limited to anastomoses in patent PUG-S-11. In contrast, the patent PUG-S-37 and PUG-S-58 were largely endothelialized. In all patent grafts at 3 months, numerous host cells had migrated, and newly formed capillaries were seen in the voids of the graft wall, which appeared moderately to highly cellular. In conclusion, it appears that penetrating micropores through the graft wall increase patency and that a highly porous structure is needed to achieve extensive endothelialization in small diameter polymer skin coated vascular grafts.

Fabrication of small-diameter polyurethane vascular grafts with microporous structure

Abstract Microporous polyurethane vascular grafts with a diameter of 3 mm were fabricated by a spray phase inversion technique (SPIT). The microporous structure and hydraulic permeability of the grafts were regulated by changing the fabrication conditions. The maximum hydraulic permeability of 26 ml/min/cm2, which was obtained in this series of grafts, satisfied the target value of about 10–40 ml/min/cm2, which our previous studies suggested would lead to satisfactory graft patency. Further investigation is, however, needed to optimize the microporous structure and hydraulic permeability of the grafts.

Endogenous basic fibroblast growth factor for endothelialization due to angiogenesis in fabric vascular prostheses

Angiogenesis by basic fibroblast growth factor (bFGF) is important for endothelialization in vascular prostheses and is significant from shortly after implantation. Canine adipose tissue was resected, minced into fragments, and suspended. This mixture was sieved through the wall of a fabric vascular prosthesis. Tissue-fragmented grafts (TF-grafts, 6 mm internal diameter, 6 cm long) were implanted into the abdominal aortae of four dogs, and four preclotted grafts were used as control subjects. Grafts were removed at 1-7 days after implantation. Removed grafts were evaluated microscopically, immunohistologically, and by scanning electron microscope. At day 1, a thrombus layer was on the TF-graft lumen. At day 3, cell proliferation and migration were observed. At day 5, endothelial-like cells were extending onto the luminal thrombus. Cell proliferation around the fragments was active, and those cells were bFGF positive. In the control subjects, at day 7, the perigraft tissue was bFGF positive, whereas no endothelialization on the lumen or no capillary infiltration into the graft wall was observed. Furthermore, bFGF was negative in the sites of thrombus and infection. These results demonstrate that endogenous bFGF is important for endothelialization due to angiogenesis in fabric vascular prostheses, whereas thrombus and infection might have a negative effect.

Rapid neointima formation with elastic laminae similar to the natural arterial wall on an adipose tissue fragmented vascular prosthesis

A vascular prosthesis that can induce a neointima similar to a natural arterial wall is reported. The authors have developed a sealing method using autologous tissue fragments. The sealed graft showed many advantages, with characteristic neointima formation in an animal study. The grafts were implanted in the thoracic descending aortae of 40 dogs and were removed from 1 hour to 608 days after implantation. Another 40 dogs, used as controls, had a fabric graft implanted using the preclotting method. The luminal surface of the sealed graft was completely endothelialized and the original adipose tissue fragments were absorbed within 1 month. Smooth muscle cells infiltrated and proliferated at the same time as endothelialization took place. Most of the smooth muscle cells were arranged in parallel rows and oriented circumferentially within the graft. At 1 month, elastic fibers appeared around the smooth muscle cells near the anastomotic sites. In the long-term specimens, these elastic fibers constituted a fine lamina in the neointima. Intimal hyperplasia and degenerative changes in the neointima were not observed. These results indicated that the sealing method could induce a very stable neointima with a smooth muscle cell layer and elastic laminae similar to a natural arterial wall within a short period of time throughout the graft wall, with maintenance of the neointima for a long period of time after implantation.

A functional neointima with regularly arranged smooth muscle cells in a fabric vascular prosthesis transplanted with autologous venous tissue fragments

Regular arrangement of smooth muscle cells underneath an endothelial cell layer was observed in the neointima of a fabric vascular prosthesis treated with new technology to accelerate endothelialization, i.e., transplantation of autologous venous tissue fragments in the graft wall. This finding indicated that the neointima has a vital function as the intima of the blood vessel. A canine left jugular vein was minced and stirred into 20 ml of saline containing 1,000 IU heparin. It was injected with pressure into a fabric prosthesis (4 mm inner diameter [ID], 3.5 cm in length, Water porosity: 4,000 ml) to create the tissue fragmented, heparinized graft. The graft was implanted into the same animal from which the jugular vein was taken. Forty tissue fragmented heparinized (TFH) grafts were implanted in both carotid arteries of 20 dogs and explanted from 1 hr to 400 days after implantation. In this study, the neointimae of the grafts implanted for more than 1 month are analyzed, with a focus on the arrangement of smooth muscle cells in the neointima. A circumferential arrangement of smooth muscle cells with a thin layer of longitudinally arranged cells underneath was seen in the neointimae, which resemble the arrangement of smooth muscle cells in the natural arterial wall. Some areas had a thin smooth muscle cell layer in the longitudinal direction just under the endothelial cell layer. At anastomotic sites, they ran in parallel rows in the longitudinal direction. The authors previously clarified that the smooth muscle cells arrange in parallel rows in the direction of strain caused by tensile stress.(ABSTRACT TRUNCATED AT 250 WORDS)

A method to reduce thrombogenicity of a graft for small diameter arterial substitution seeded with autologous venous tissue fragments.

The authors successfully applied a method to accelerate endothelialization by tissue fragmentation to a small diameter fabric vascular prosthesis. Tissue fragment seeded grafts showed rapid healing of the neointima. The thrombogenicity of the collagen fibrils in the fragments, however, caused major problems when the method was applied to small diameter grafts: the positively charged collagen fibrils aggregated the negatively charged platelets. The authors masked the fibrils electrostatically with heparin molecules, which are negatively charged. A canine jugular vein was resected, minced into tissue fragments, and suspended in the heparin solution; it then was sieved through the wall of a fabric prosthesis. The grafts (4 mm internal diameter and 3.5 cm in length) were implanted into both carotid arteries of six dogs (12 grafts). Tissue fragment seeded grafts without heparin also were implanted into six dogs. As a control, preclotted fabric grafts were implanted into six dogs (12 grafts). These grafts occluded within 1 week, whereas all the masked grafts were patent without thrombi. In vitro examination of heparin release revealed that approximately 92% of heparin in the graft was released during the first 5 hr, but approximately 6% remained after 25 hr. These results indicate that the method is applicable to small diameter arterial grafts.

Long-term results of a new antithrombogenic cardiac wall substitute

We evaluated the long-term results of a new antithrombogenic cardiac wall substitute which is composed of collagen-coated ultrafine polyester mesh cross-linked by a hydrophilic polyepoxy compound (CUFP), and compared it with glutaraldehyde-treated equine pericardium (GA graft). In an animal study, 20 CUFPs and 18 GA control grafts were implanted as patches in the right ventricular outflow tract in 38 dogs. In the CUFP at 28 days after implantation, a thin neointima, which was almost endothelialized, had been formed. Fibroblasts and vasa vasorum were seen inside both the neointima and the graft wall. At 168 days, transmission electron microscopy revealed smooth muscle-like cells in the neointima, with endothelialization. The CUFP at 486 days had maintained a white, shiny, smooth, thin, and uniform neointima with endothelialization. Neither ulcer nor thrombus formation was seen on the surface, while in the GA graft at 353 days, no endothelialization was noticed in the center and almost no fibroblasts had infiltrated into the graft wall. From these results, the CUFP should be an ideal candidate for cardiac wall substitution.

Microporous Polyurethane Inhibits Critical Mural Thrombosis and Enhances Endothelialization at Blood-Contacting Surface

This study addresses a long-lasting controversy, ‘Which is a better blood-contacing surface morphology, smooth or rough?’ and indicates the importance of micropores penetrating through the graft wall in minimizing thrombosis. Four types of spongy polyurethanepolydimethylsiloxane (Cardiothane 51; Kontron Instruments, Everett, MA, USA) vascular grafts (PUG), 1.5 mm ID and 1.5–2cm in length, fabricated by a spray, phase-inversion technique, were implanted end-to-end in the infrarenal aorta of 58 adult rats. Some grafts had a continuous inner skin and, consequently, a hydraulic permeability of 0 ml/min per cm2 (PUG-S-0). Some had an inner skin with isolated pores and a mean hydraulic permeability of 11 ml/min per cm2 (PUG-S-11). Some had a microporous luminal surface with a mean hydraulic permeability of either 2.7 ml/min per cm2 (PUG-2.7) or 39 ml/min per cm2 (PUG-39). Twelve PUG-S-0, 6 PUG-S-11, 23 PUG-2.7, and 17 PUG-39 were evaluated between 2 hours and 3 months post-implantation. Almost all PUS-S-0 occluded with thrombus soon after implantation. PUG-2.7 had poor patency. Both PUG-S-11 and PUG-39 showed acceptable patency. Endothelialization, however, was limited to 1–2 mm from proximal and distal anastomoses in any patent PUG-S-11. In contrast, all but one of the patent PUG-39 were completely endothelialized. The extent of mural thrombosis decreased in the order from PUG-39 to PUG-S-11, PUG-2.7, and PUG-S-0. In conclusion, micropores penetrating through the graft wall, as reflected by hydraulic permeability values, appear to inhibit critical mural thrombosis and to promote a high degree of endothelialization.

Influence of Microporous Structures on Mural Thrombosis and Endothelialization at Blood-Contacting Surfaces

The influence of microporous structures in the walls of small-diameter arterial prostheses was investigated with the aim of minimizing thrombosis and enhancing endothelialization of blood-contacting surfaces. Six types of spongy polyurethane-polydimethylsiloxane grafts (PUG), 1.5-mm in an internal diameter and 1.5–2 cm in length, were implanted end-to-end in the infrarenal aorta of 66 adult rats. Some had a continuous inner skin and a hydraulic permeability (HP) of Oml/min/cm2at the standard transmural pressure of 120mmHg (PUG-S-0). Some had a discontinuous inner skin with some isolated windows connecting penetrating micropores though the graft wall and a mean HP ranging from 11 (PUG-S-11) to 37 (PUG-S-37) or 58 (PUG-S-58) ml/ min/cm2. The rest had a microporous inner surface with penetrating micropores through the graft wall and a mean HP of 2.7 (PUG-2.7) or 39 (PUG-39) ml/min/cm2. PUG which had a HP of less than 2.7 ml/min/cm2showed poor patency. PUG with a HP of more than 11 ml/min/cm2had acceptable patency, but endothelialization was limited to their anastomoses. In contrast, the patent PUG-S-37 and PUG-S-58 were largely endothelialized and all but one of the patent PUG-39 implants were completely endothelialized. In conclusion, penetrating micropores through the graft wall appear to inhibit critical mural thrombosis. A microporous inner surface seems to be superior to a skinned inner surface in achieving a high degree of endothelialization.