Small-caliber polyurethane and polytetrafluoroethylene grafts: A comparative study in a canine aortoiliac model Thomas E. Brothers,* James C. Stanley, William E. Burkel,* and Linda M. Grahamt Section of Vascular Surgery, Department of Surgery; and *Department of Anatomy and Cell Biology, University of Michigan Medical Center, A n n Arbor, Michigan and 'Department of Surgery, Case Western Reserve University, Cleveland, Ohio
In uivo stability of a new small-caliber polyurethane graft (n8) was assessed in a canine aortoiliac model and compared to that of a conventional expanded polytetrafluoroethylene (eM'FE) graft (n8). Six months following implantation, marked aneurysmal dilatation to 230 t 80% ( x 2 SD) of the original diameter occurred in polyurethane grafts, while dilatation to 110 ? 8% of the original diameter occurred in ePTFE grafts ( p < 0.005). Interval patency was 75% for each graft type. Luminal thrombus
affected 59% of polyurethane graft surfaces compared to 22% of ePTFE graft surfaces ( p < 0.01). Qualitative examination of representative sections of polyurethane conduits demonstrated thick inner capsules with numerous small islands of graft material surrounded by macrophages and bands of mature fibrous tissue, in contrast to the thinner neointima and limited anastomotic pannus ingrowth observed in ePTFE grafts.
INTRODUCTION
The ideal small-caliber prosthetic vascular graft has not yet been developed. A variety of conduits have been used for vascular reconstructions in the absence of usable autogenous vein, but all demonstrate inferior longterm patency compared to autogenous vein, primarily because of progressive anastomotic intimal hyperplasia.' Factors linked to intimal hyperplasia include surface thrombogenicity of the graft and mechanical stresses due to anastomotic compliance mismatch.* Experimental studies involving 4.0mm-internal-diameter ePTFE and Dacron grafts have documented improved patencies by altering surface thrombogenicity through use of antiplatelet therapy or creation of antithrombotic luminal linings of endothelium following endothelial cell seeding, although the applicability of these interventions has not been substantiated in controlled clinical studElastomeric materials such as latex, Lycra, and various polyurethanes have been introduced in attempts to minimize discrepancies between vessel q o whom all correspondence should be addressed to at: University of Michigan Medical Center-2210 THCC, 1500 East Medical Center Drive, Ann Arbor, MI 48109. Journal of Biomedical Materials Research, Vol. 24, 761-771 (1990) 0 1990 John Wiley & Sons, Inc. CCC 0021-9304/90/060761-11$04.00
762
BROTHERS ET AL.
and conduit c~mpliance.~” The potential for these alternative conduits to improve patency has never been established clinically. In the current investigation, a new small-caliber polyurethane graft was compared to a conventional ePTFE graft in regard to long-term physical stability. MATERIALS AND METHODS
Eight female mongrel dogs weighing 20-35 kg underwent placement of bilateral aortoiliac bypass grafts using conduits of polyurethane on one side and ePTFE on the opposite side. All animals studied had normal leukocyte counts, hematocrits, platelet counts, prothrombin times, partial thromboplastin times, thrombin clotting times, and fibrinogen levels. In addition, all dogs had normal preoperative platelet aggregation to ADP, thrombin, and collagen. Acetylsalicylic acid (5 gr PO qd) was given 1 day preoperatively and continued postoperatively for 14 days. Surgical procedures and animal care complied with standards in ”The Guide for the Care and Use of Laboratory Animals” (DHEW Publication No. [NIH] 85-23, revised 1985). The two grafts evaluated in this experiment were a polyether polyurethane vascular graft (Medtronic, Inc., Minneapolis, MN) with a preimplantation internal diameter of 4.8-mm and a 5.0-mm-internal-diameter ePTFE graft (W. L. Gore, Flagstaff, AZ). The polyurethane grafts were prepared by phase inversion techniques from a solution of 8.F. Goodrich Estane 5714 (a polyether polyurethane) containing added salts to produce a porous graft following salt extraction (Fig. 1). These polyurethane grafts had a void volume of approximately 70%, with a porosity of approximately 30 mL/cm2/ min at 120 mm Hg. Prior to implantation, the polyurethane grafts were soaked for 20 min in normal saline, followed by preclotting with unheparinized autologous blood. The ePTFE grafts were conventional microporous conduits used in contemporary clinical practice, exhibiting zero porosity by the method of Weslowski.” Animals were anesthetized with thiamylal sodium induction and endotracheal halothane/N,O/O, maintenance. Operative hydration consisted of intravenous Ringer’s lactate, 10 mL/kg/h. Antibiotic prophylaxis was accomplished by intramuscular administration of penicillin benzathine (450,000 U) and penicillin procaine hydrochloride (450,000 U). The infrarenal aorta and external iliac arteries were exposed through a midline abdominal approach. Prior to graft implantation, animals were anticoagulated with intravenous heparin sodium (100 U/kg). Grafts were anastomosed end-to-side to both the infrarenal aorta and distal external iliac arteries using 6-0 polypropylene suture. Following simultaneous establishment of flow through both grafts, the heparin anticoagulation was reversed with intravenous protamine sulfate (1.0 mg/kg). Ligation of the middle sacral and external iliac arteries proximal to distal anastomoses was performed to exclude major lower extremity collateral blood flow. Intraoperative arteriogra-
POLYURETHANE GRAFTS
763
Figure 1. Scanning electron microscopic appearance of polyurethane graft (X80).
phy at a standard distance of 1.5 m confirmed the technical adequacy of all reconstructions and allowed vernier caliper determination of the postimplantation in vivo internal diameter of the grafts with a precision of 0.1 mm. The graft status was assessed over a 6-month follow-up period. Patency was assessed by tiweekly femoral pulse palpation and at 3 and 6 months after graft implantation by arteriography performed in a manner similar to that undertaken intraoperatively. The latter was performed on anesthetized animals with a 30-cm-long, 16-gauge catheter placed through the left common carotid artery into the proximal abdominal aorta. One animal was sacrificed at 6 weeks because of absent femoral pulses and angiographic evidence of bilateral graft occlusion. At the completion of the 6-month study period, the dogs were anesthetized, and the abdominal aorta and femoral arteries were isolated. Animals were anticoagulated with intravenous heparin sodium (150 U/kg). An infusion catheter was placed in the left renal artery and outflow cannulas were inserted in both femoral arteries of each animal. The aorta was crossclamped, and the grafts were flushed with phosphate buffered saline (PBS) until the effluent was clear. The grafts were then perfusion-fixed in situ with 4% paraformaldehyde for 30 min. Subsequently, the grafts and adjacent arteries were removed, bisected longitudinally, and photographed while submerged in PBS. The 35-mm color transparencies of these grafts were
BROTHERS ET AL.
764
projected, and the total and thrombus-free luminal surfaces were outlined using a computer-linked-digitizer in order to quantitate the extent of surface thrombus. Sections from proximal and distal anastomoses and central grafts were prepared by standard means for light microscopy. Data was subjected to statistical analysis using Student's paired t test. RESULTS
Four grafts occluded during the investigation, two of each type, including two in the same animal, yielding a 6-month patency of 75% for both graft types. This patency is consistent with previous reports for unseeded smallcaliber vascular prostheses and illustrates the rigorous nature of the canine aortoiliac model for evaluation of graft p a t e n ~ yThe . ~ ~occlusions occurred between 24 and 98 days following insertion and were not associated with recognizable technical problems at either proximal or distal anastomoses. Results of serial measurement of the internal diameter of the polyurethane and ePTFE grafts revealed important differences in the two types of conduits (Table I). Polyurethane grafts became slightly larger in vivo due to their greater elasticity. Arteriograms as early as 3 months after graft implantation detected aneurysmal changes in polyurethane grafts, with continued aneurysmal dilatation observed at 6 months, corresponding to a mean 230 +. 80% increase from the original internal diameter (Fig. 2). Graft disruption was not evident. No aneurysmal changes were detected in the ePTFE conduits at 3 or 6 months. Measurements of initial and final graft lengths did not reveal differences between the two conduits studied. Graft surface thrombus (Fig. 3 ) was significantly greater ( p < 0.01) at 6 months on polyurethane grafts (59%)compared to ePTFE grafts (22%).It was unclear whether thrombus resulted from inherent thrombogenic properties of the polyurethane surface or was secondary to flow separation abnormalities associated with aneurysmal graft changes. Qualitative exami-
TABLE I Internal Diameter of Polyurethane and ePTFE Canine Aortoiliac Grafts Internal Diameter (mmp Postimplantation Preimplantation Polyurethane ePTFE
4.8 5 0.0 (n8) 5.0 i: 0.0 (n8)
Intraoperative
*
3 months
5.7 0.3 (r18)~ 8.6 4.9 f 0.2 (n8) 4.8
? f
2.3 (n6)bs' 0.9 (n7)
"Internal diameter expressed as mean k standard deviation. ' p < 0.001 polyurethane vs. ePTFE. ' p < 0.05 vs. intraoperative postimplantation diameter. dp < 0.005 vs. intraoperative postimplantation diameter.
6 months 12.8 5.4
* 4.1 (n6)b!d
* 0.2 (1161
POLYURETHANE GRAFTS
765
Figure 2. Angiographic appearance of polyurethane (subject’s right) and expanded polytetrafluoroethylene (subject’s left) aortoiliac grafts 6 months postimplantation.
nation of representative histological sections of the two graft types revealed marked differences (Figs. 4,5). Polyurethane grafts exhibited a thick neoin-
766
BROTHERS ET AL.
Figure 3. Gross appearance of polyurethane and expanded polytetrafluoroethylene prosthetic grafts excised following 6 months implantation. Anterior portion (left) and posterior portion (right) of grafts and arteries sectioned en bloc longitudinally.
tima with interposed foci of residual graft surrounded by macrophages, giant cells, and bands of mature fibrous tissue. This was in distinct contrast to the relatively thin neointima and pannus observed in ePTFE grafts.
767
POLYURETHANE GRAFTS
Figure 4. Histologic appearance of polyurethane graft 6 months postimplantation ( X 100). Relatively cellular thickened inner capsule is lined by endothelium in some areas. A paucity of residual graft material is readily apparent.
DISCUSSION
Research continues toward development of improved small-caliber prosthetic vascular grafts for use as alternatives to autogenous saphenous vein in coronary and peripheral vascular arterial reconstructions. Anastomotic intimal hyperplasia is a major cause of late failure of most currently available prosthetic grafts. Among those factors contributing to development of intimal hyperplasia are activation of platelets and other blood elements by prosthetic surfaces and compliance mismatches between grafts and host arteries. Polyurethanes are currently used clinically in artificial hearts, ventricular assist devices, and intravascular catheters because of their relatively favorable biocompatibility. Earlier laboratory investigations involving certain experimental polyurethane grafts suggested their potential usefulness in peripheral vascular reconstructions, with studies up to 1 year demonstrating endothelialization of luminal surfaces and acceptable patency rates without evidence of aneurysmal degeneration.’,17, l9 However, not all previous polyurethane conduits have fared so well. For example, Martz reported complete occlusion at 6 months of all 5-mm polyurethane grafts placed in canine carotid arteries.” Such descrepancies in outcome confirm the fact
768
BROTHERS ET AL.
Figure 5. Histologic appearance of polyurethane graft 6 months postimplantation (X450). Giant cells and macrophages surrounding polyurethane can be identified. Scattered bands of fibrous tissue are present.
that investigators should not extrapolate performance of one chemical and physical type of polyurethane to another.
POLYURETHANE GRAFTS
769
In vivo stability of polyurethane grafts remains a problem. Although early in vitro and in vivo testing suggested excellent stability, these polymers are susceptible to urethane group hydrolysis.21*22 However, simple chemical hydrolysis or soft-segment oxidation does not completely explain the deterior a t i ~ n . ’Although ~ some polyether urethanes degrade in water alone, addition of enzyme solutions enhances degradation.’* In regard to the latter, the presence of numerous macrophages and giant cells surrounding residual nests of graft material in the present study raises the question as to whether leukocyte enzyme release contributes to graft destruction. Enzymatic degradation of certain polyether polyurethanes has been reported to occur with exposure to hydrogen peroxide or enzymes of the type released from inflammatory cells during foreign body reactions.= In support of this phenomenon is the observation that papain, closely related to cathepsin B released by inflammatory cells, causes chain cleavage of some polyurethanes, possibly by urethane and urea linkage hydrolysis.’6 Effects of enzymatic activity on polyurethane materials may be magnified further under conditions of cyclic loading in the cardiovascular Increasing the porosity of polyurethanes will also increase the degradation rate, and external surfaces are most susceptible to enzymatic breakd~wn.’~ Development of progressive aneurysmal dilatation in the experimental polyurethane grafts used in the present study is of significant concern and renders this particular prosthesis unsuitable for chronic implantation. The fibrous tissue reaction coupled with the graft material degradation was insufficient to resist the repetitive circumferential distending forces that promoted eventual graft dilatation. Other investigators have observed similar tissue reactions to implanted polyurethane grafts, including absence of significant fibroblast invasion accompanying degeneration of the graft external surfaces 6 months following implantation. 12~17~19,20 Interestingly, in vitro studies have demonstrated that conditioned medium from human macrophages cultured in the presence of one polyurethane do not stimulate fibroblast proliferation despite normal interleukin-1 production, in contrast to stimulation of fibroblast proliferation provided by biomaterials such as Dacron, ePTFE, and polyethylene.” If these observations can be extrapolated to in vivo graft performance, then inhibition of fibroblast proliferation by polyurethane may account for the failure of autologous tissue ingrowth to provide adequate structural integrity as polyurethane degenerates. Support for this investigation was received from Medtronic, Inc., Minneapolis, MN, and W. L. Gore & Associates, Inc., Flagstaff, AZ.
References 1. A. D. Callow, ”Historical overview of experimental and clinical development of vascular grafts,” in Biological and Synthetic Vascular Prostheses, J. C . Stanley, et al. (eds.), Grune and Stratton, New York, 1982, pp. 11-36.
BROTHERS ET AL.
770 2.
3. 4.
5. 6.
7. 8. 9.
10. 11. 12.
13.
14.
15. 16.
17.
18.
19.
W. M. Abbott and R. P. Cambria, ”Elasticity and compliance of vascular grafts,” in Biological and Synthetic Vascular Prostheses, J. C. Stanley et al. (eds.), Grune and Stratton, New York, 1982, pp. 189-220. T. A. Belden, S. P. Schmidt, L. J. Falkow, and W. V. Sharp, “Endothelial cell seeding of small-diameter vascular grafts,” Trans. Am. SOC.Artif. Intern. Organs, 28, 173-177 (1982). C. D. Campbell, D. Goldfarb, and R. Roe, “A small arterial substitute: Expanded microporous polytetrafluoroethylene: Patency versus porosity,” Ann. Surg., 182, 138-143 (1975). D.Goldfarb, J. A. Houk, J. L. Moore, Sr., and D. L. Gain, “Graphiteexpanded polytetrafluoroethylene: An improved small artery prosthesis,“ Trans. Am. SOC.Artif. Intern. Organs, 23, 268-276 (1977). L. M. Graham, J. C. Stanley, and W. E. Burkel, “Improved patency of endothelial-cell-seeded, long, knitted Dacron and PTFE vascular prostheses,’’ ASAIO J., 8, 65-73 (1985). B.T. Allen, J.A. Long, M. J. Welch, K.T. Hopkins, G. A. Sicard, and R. E. Clark, ”Effect of aspirin therapy and its withdrawal on control and endothelial cell seeded grafts,” Surg. Forum, 34, 470-472 (1983). D. Annis, A. Bornat, R. 0. Edwards, A. Higham, B. Loveday, and J. Wilson, “An elastomeric vascular prosthesis,” Trans. Am. SOC.Artif. Intern. Organs, 24, 209-214 (1978). B. Dreyer, T. Akutsu, and W. J. Kolff, “Aortic grafts of polyurethane in dogs,” J. Appl. Physiol., 15, 18-22 (1960). D. J. Lyman, D. Albo, Jr., R. Jackson, and K. Knutson, “Development of small diameter vascular prostheses,” Trans. Am. Soc. Artif. Intern. Organs, 23, 253-256 (1977). V. Marinescu, E. Pausescu, and S. Carnaru, ”Long-term biological fate of polyurethane aortic prostheses,” Thorax, 206, 108-111 (1971). H. Martz, R. Paynter, S.B. Slimane, G. Beaudoin, R. Guidoin, J. Borzone, H. B. Simhon, R. Satin, and N. Sheiner, “Hydrophilic microporous polyurethane versus expanded PTFE grafts as substitutes in the carotid arteries of dogs. A limited study,” 1. Biomed. Muter. Res., 22, 6369 (1988). K. S. Murabayashi, H. Kambic, H. Harasaki, T. Morimoto, R. Yozu, and Y. Nose. “Fabrication and long-term imdantation of semi-comuliant small vascular prosthesis,” Trins. Am. h c . Artif. Intern. Organ’s, 31, 50-54 f1985). W. V. Sharp’, A. F. Finelli, W. H. Falor, and J. W. Ferraro, ”Latex vascular prostheses: Patency rate and neointimization related to prosthesis lining and electrical conductivity,” Circulation, 29, (supplj, 165-170 (1964). M. Wagner, G. Red, J. Teresi, and K. L. Kayser, ”Experimental observations on a new and inherently elastic material for sutures and vascular prostheses: Lycra,” Am. 1. Surg., 3, 838-841 (1966). R. A. White, E. W. White, E. L. Hanson, R. E Rohner, and W. R. Webb, ”Preliminary report: Evaluation of tissue ingrowth into experimental replamineform vascular prostheses,” Surgery, 79, 229-232 (1976). G. J. Wilson, D. C. MacGregor, P. Klement, J. M. Lee, P. J. del Nido, E. W. C. Wong, and J. Leidner, “Anisotropic polyurethane nonwoven conduits: A new approach to the design of a vascular prosthesis,” Trans. Am. Soc. Artif. Intern. Organs, 24, 260-268 (1983). W. F. Abbott and R. P. Cambria, ”Control of physical characteristics (elasticity and compliance) of vascular grafts, in Biologic and Synthetic Vascular Prostheses, J. C. Stanley et al. (eds.), Grune and Stratton, New York, 1982, p. 212. F. Hess, B. Braun, C . Jerusalem, R. van Det, S. Steeghs, S. Skotnicki, and P. Grande, “Endothelialization of polyurethane vascular prosthe-
POLYURETHANE GRAFTS
20.
21. 22.
23. 24. 25.
26. 27. 28.
771
ses implanted in the dog carotid and femoral artery,” J. Curdiovusc. Surg., 29, 458 (1988). H.Martz, R. Paynter, S. B. Slimane, G. Beaudoin, and R. Guiloin, ”Hydrophylic microporous polyurethane versus expanded PTFE grafts as substitutes in the carotid arteries of dogs. A limited study,” I. Biorned. Muter. Res., 22, 63 (1988). A. S. Chawla, P. Blais, I. Hinberg, and D. L. Johnson, “Experience with retrieved cardiac pacing leads,” Symp. Retrieval & Analysis of Surgical lmplunts & Biomuterials, 52, (1988). D.L. Coleman, H. L. C. Meuzelaar, T. R. Kessler, W. H. McClennen, J. M. Richards, and D. E. Gregonis, ”Retrieval and analysis of a clinical total artificial heart,” J. Biomed. Muter. Res., 20, 417 (1986). R. J. Thoma and R. E. Phillips, ”Note: Studies of poly(ether)urethane pacemaker lead insulation oxidation,” I . Biomed. Muter. Res., 21, 525 (1987). R. E. Marchant, Q. Zhao, J. M. Anderson, and A. Hiltner, “Degradation of a poly(ether urethane urea) elastomer: infra-red and XPS studies,” Polymer, 28, 2032 (1987). B.D. Ratner, K. W. Gladhill, and T. A. Horbett, ”Analysis of in vitro enzymatic and oxidative degradation of polyurethanes,” J. Biomed. Muter. Res., 22, 509 (1988). S.K. Phua, E. Castillo, J. M. Anderson, and A. Hiltner, “Biodegradation of a polyurethane in vitro,” I. Biomed. Muter. Res., 21, 231 (1987). Q.Zhao, R. E. Marchant, J. M. Anderson, and A. Hiltner, “Long term biodegradation in vitro of poly(ether urethane urea): a mechanical property study,” Polymer, 28, 2040 (1987). K.M. Miller and J. M. Anderson, ”In vitro stimulation of fibroblast activity by factors generated from human monocytes activated by biomedical polymers,” J . Biomed. Muter. Res., 23, 911 (1989).
Received September 22, 1989 Accepted December 14, 1989
In uivo stability of a new small-caliber polyurethane graft (n8) was assessed in a canine aortoiliac model and compared to that of a conventional expanded polytetrafluoroethylene (eM'FE) graft (n8). Six months following implantation, marked aneurysmal dilatation to 230 t 80% ( x 2 SD) of the original diameter occurred in polyurethane grafts, while dilatation to 110 ? 8% of the original diameter occurred in ePTFE grafts ( p < 0.005). Interval patency was 75% for each graft type. Luminal thrombus
affected 59% of polyurethane graft surfaces compared to 22% of ePTFE graft surfaces ( p < 0.01). Qualitative examination of representative sections of polyurethane conduits demonstrated thick inner capsules with numerous small islands of graft material surrounded by macrophages and bands of mature fibrous tissue, in contrast to the thinner neointima and limited anastomotic pannus ingrowth observed in ePTFE grafts.
INTRODUCTION
The ideal small-caliber prosthetic vascular graft has not yet been developed. A variety of conduits have been used for vascular reconstructions in the absence of usable autogenous vein, but all demonstrate inferior longterm patency compared to autogenous vein, primarily because of progressive anastomotic intimal hyperplasia.' Factors linked to intimal hyperplasia include surface thrombogenicity of the graft and mechanical stresses due to anastomotic compliance mismatch.* Experimental studies involving 4.0mm-internal-diameter ePTFE and Dacron grafts have documented improved patencies by altering surface thrombogenicity through use of antiplatelet therapy or creation of antithrombotic luminal linings of endothelium following endothelial cell seeding, although the applicability of these interventions has not been substantiated in controlled clinical studElastomeric materials such as latex, Lycra, and various polyurethanes have been introduced in attempts to minimize discrepancies between vessel q o whom all correspondence should be addressed to at: University of Michigan Medical Center-2210 THCC, 1500 East Medical Center Drive, Ann Arbor, MI 48109. Journal of Biomedical Materials Research, Vol. 24, 761-771 (1990) 0 1990 John Wiley & Sons, Inc. CCC 0021-9304/90/060761-11$04.00
762
BROTHERS ET AL.
and conduit c~mpliance.~” The potential for these alternative conduits to improve patency has never been established clinically. In the current investigation, a new small-caliber polyurethane graft was compared to a conventional ePTFE graft in regard to long-term physical stability. MATERIALS AND METHODS
Eight female mongrel dogs weighing 20-35 kg underwent placement of bilateral aortoiliac bypass grafts using conduits of polyurethane on one side and ePTFE on the opposite side. All animals studied had normal leukocyte counts, hematocrits, platelet counts, prothrombin times, partial thromboplastin times, thrombin clotting times, and fibrinogen levels. In addition, all dogs had normal preoperative platelet aggregation to ADP, thrombin, and collagen. Acetylsalicylic acid (5 gr PO qd) was given 1 day preoperatively and continued postoperatively for 14 days. Surgical procedures and animal care complied with standards in ”The Guide for the Care and Use of Laboratory Animals” (DHEW Publication No. [NIH] 85-23, revised 1985). The two grafts evaluated in this experiment were a polyether polyurethane vascular graft (Medtronic, Inc., Minneapolis, MN) with a preimplantation internal diameter of 4.8-mm and a 5.0-mm-internal-diameter ePTFE graft (W. L. Gore, Flagstaff, AZ). The polyurethane grafts were prepared by phase inversion techniques from a solution of 8.F. Goodrich Estane 5714 (a polyether polyurethane) containing added salts to produce a porous graft following salt extraction (Fig. 1). These polyurethane grafts had a void volume of approximately 70%, with a porosity of approximately 30 mL/cm2/ min at 120 mm Hg. Prior to implantation, the polyurethane grafts were soaked for 20 min in normal saline, followed by preclotting with unheparinized autologous blood. The ePTFE grafts were conventional microporous conduits used in contemporary clinical practice, exhibiting zero porosity by the method of Weslowski.” Animals were anesthetized with thiamylal sodium induction and endotracheal halothane/N,O/O, maintenance. Operative hydration consisted of intravenous Ringer’s lactate, 10 mL/kg/h. Antibiotic prophylaxis was accomplished by intramuscular administration of penicillin benzathine (450,000 U) and penicillin procaine hydrochloride (450,000 U). The infrarenal aorta and external iliac arteries were exposed through a midline abdominal approach. Prior to graft implantation, animals were anticoagulated with intravenous heparin sodium (100 U/kg). Grafts were anastomosed end-to-side to both the infrarenal aorta and distal external iliac arteries using 6-0 polypropylene suture. Following simultaneous establishment of flow through both grafts, the heparin anticoagulation was reversed with intravenous protamine sulfate (1.0 mg/kg). Ligation of the middle sacral and external iliac arteries proximal to distal anastomoses was performed to exclude major lower extremity collateral blood flow. Intraoperative arteriogra-
POLYURETHANE GRAFTS
763
Figure 1. Scanning electron microscopic appearance of polyurethane graft (X80).
phy at a standard distance of 1.5 m confirmed the technical adequacy of all reconstructions and allowed vernier caliper determination of the postimplantation in vivo internal diameter of the grafts with a precision of 0.1 mm. The graft status was assessed over a 6-month follow-up period. Patency was assessed by tiweekly femoral pulse palpation and at 3 and 6 months after graft implantation by arteriography performed in a manner similar to that undertaken intraoperatively. The latter was performed on anesthetized animals with a 30-cm-long, 16-gauge catheter placed through the left common carotid artery into the proximal abdominal aorta. One animal was sacrificed at 6 weeks because of absent femoral pulses and angiographic evidence of bilateral graft occlusion. At the completion of the 6-month study period, the dogs were anesthetized, and the abdominal aorta and femoral arteries were isolated. Animals were anticoagulated with intravenous heparin sodium (150 U/kg). An infusion catheter was placed in the left renal artery and outflow cannulas were inserted in both femoral arteries of each animal. The aorta was crossclamped, and the grafts were flushed with phosphate buffered saline (PBS) until the effluent was clear. The grafts were then perfusion-fixed in situ with 4% paraformaldehyde for 30 min. Subsequently, the grafts and adjacent arteries were removed, bisected longitudinally, and photographed while submerged in PBS. The 35-mm color transparencies of these grafts were
BROTHERS ET AL.
764
projected, and the total and thrombus-free luminal surfaces were outlined using a computer-linked-digitizer in order to quantitate the extent of surface thrombus. Sections from proximal and distal anastomoses and central grafts were prepared by standard means for light microscopy. Data was subjected to statistical analysis using Student's paired t test. RESULTS
Four grafts occluded during the investigation, two of each type, including two in the same animal, yielding a 6-month patency of 75% for both graft types. This patency is consistent with previous reports for unseeded smallcaliber vascular prostheses and illustrates the rigorous nature of the canine aortoiliac model for evaluation of graft p a t e n ~ yThe . ~ ~occlusions occurred between 24 and 98 days following insertion and were not associated with recognizable technical problems at either proximal or distal anastomoses. Results of serial measurement of the internal diameter of the polyurethane and ePTFE grafts revealed important differences in the two types of conduits (Table I). Polyurethane grafts became slightly larger in vivo due to their greater elasticity. Arteriograms as early as 3 months after graft implantation detected aneurysmal changes in polyurethane grafts, with continued aneurysmal dilatation observed at 6 months, corresponding to a mean 230 +. 80% increase from the original internal diameter (Fig. 2). Graft disruption was not evident. No aneurysmal changes were detected in the ePTFE conduits at 3 or 6 months. Measurements of initial and final graft lengths did not reveal differences between the two conduits studied. Graft surface thrombus (Fig. 3 ) was significantly greater ( p < 0.01) at 6 months on polyurethane grafts (59%)compared to ePTFE grafts (22%).It was unclear whether thrombus resulted from inherent thrombogenic properties of the polyurethane surface or was secondary to flow separation abnormalities associated with aneurysmal graft changes. Qualitative exami-
TABLE I Internal Diameter of Polyurethane and ePTFE Canine Aortoiliac Grafts Internal Diameter (mmp Postimplantation Preimplantation Polyurethane ePTFE
4.8 5 0.0 (n8) 5.0 i: 0.0 (n8)
Intraoperative
*
3 months
5.7 0.3 (r18)~ 8.6 4.9 f 0.2 (n8) 4.8
? f
2.3 (n6)bs' 0.9 (n7)
"Internal diameter expressed as mean k standard deviation. ' p < 0.001 polyurethane vs. ePTFE. ' p < 0.05 vs. intraoperative postimplantation diameter. dp < 0.005 vs. intraoperative postimplantation diameter.
6 months 12.8 5.4
* 4.1 (n6)b!d
* 0.2 (1161
POLYURETHANE GRAFTS
765
Figure 2. Angiographic appearance of polyurethane (subject’s right) and expanded polytetrafluoroethylene (subject’s left) aortoiliac grafts 6 months postimplantation.
nation of representative histological sections of the two graft types revealed marked differences (Figs. 4,5). Polyurethane grafts exhibited a thick neoin-
766
BROTHERS ET AL.
Figure 3. Gross appearance of polyurethane and expanded polytetrafluoroethylene prosthetic grafts excised following 6 months implantation. Anterior portion (left) and posterior portion (right) of grafts and arteries sectioned en bloc longitudinally.
tima with interposed foci of residual graft surrounded by macrophages, giant cells, and bands of mature fibrous tissue. This was in distinct contrast to the relatively thin neointima and pannus observed in ePTFE grafts.
767
POLYURETHANE GRAFTS
Figure 4. Histologic appearance of polyurethane graft 6 months postimplantation ( X 100). Relatively cellular thickened inner capsule is lined by endothelium in some areas. A paucity of residual graft material is readily apparent.
DISCUSSION
Research continues toward development of improved small-caliber prosthetic vascular grafts for use as alternatives to autogenous saphenous vein in coronary and peripheral vascular arterial reconstructions. Anastomotic intimal hyperplasia is a major cause of late failure of most currently available prosthetic grafts. Among those factors contributing to development of intimal hyperplasia are activation of platelets and other blood elements by prosthetic surfaces and compliance mismatches between grafts and host arteries. Polyurethanes are currently used clinically in artificial hearts, ventricular assist devices, and intravascular catheters because of their relatively favorable biocompatibility. Earlier laboratory investigations involving certain experimental polyurethane grafts suggested their potential usefulness in peripheral vascular reconstructions, with studies up to 1 year demonstrating endothelialization of luminal surfaces and acceptable patency rates without evidence of aneurysmal degeneration.’,17, l9 However, not all previous polyurethane conduits have fared so well. For example, Martz reported complete occlusion at 6 months of all 5-mm polyurethane grafts placed in canine carotid arteries.” Such descrepancies in outcome confirm the fact
768
BROTHERS ET AL.
Figure 5. Histologic appearance of polyurethane graft 6 months postimplantation (X450). Giant cells and macrophages surrounding polyurethane can be identified. Scattered bands of fibrous tissue are present.
that investigators should not extrapolate performance of one chemical and physical type of polyurethane to another.
POLYURETHANE GRAFTS
769
In vivo stability of polyurethane grafts remains a problem. Although early in vitro and in vivo testing suggested excellent stability, these polymers are susceptible to urethane group hydrolysis.21*22 However, simple chemical hydrolysis or soft-segment oxidation does not completely explain the deterior a t i ~ n . ’Although ~ some polyether urethanes degrade in water alone, addition of enzyme solutions enhances degradation.’* In regard to the latter, the presence of numerous macrophages and giant cells surrounding residual nests of graft material in the present study raises the question as to whether leukocyte enzyme release contributes to graft destruction. Enzymatic degradation of certain polyether polyurethanes has been reported to occur with exposure to hydrogen peroxide or enzymes of the type released from inflammatory cells during foreign body reactions.= In support of this phenomenon is the observation that papain, closely related to cathepsin B released by inflammatory cells, causes chain cleavage of some polyurethanes, possibly by urethane and urea linkage hydrolysis.’6 Effects of enzymatic activity on polyurethane materials may be magnified further under conditions of cyclic loading in the cardiovascular Increasing the porosity of polyurethanes will also increase the degradation rate, and external surfaces are most susceptible to enzymatic breakd~wn.’~ Development of progressive aneurysmal dilatation in the experimental polyurethane grafts used in the present study is of significant concern and renders this particular prosthesis unsuitable for chronic implantation. The fibrous tissue reaction coupled with the graft material degradation was insufficient to resist the repetitive circumferential distending forces that promoted eventual graft dilatation. Other investigators have observed similar tissue reactions to implanted polyurethane grafts, including absence of significant fibroblast invasion accompanying degeneration of the graft external surfaces 6 months following implantation. 12~17~19,20 Interestingly, in vitro studies have demonstrated that conditioned medium from human macrophages cultured in the presence of one polyurethane do not stimulate fibroblast proliferation despite normal interleukin-1 production, in contrast to stimulation of fibroblast proliferation provided by biomaterials such as Dacron, ePTFE, and polyethylene.” If these observations can be extrapolated to in vivo graft performance, then inhibition of fibroblast proliferation by polyurethane may account for the failure of autologous tissue ingrowth to provide adequate structural integrity as polyurethane degenerates. Support for this investigation was received from Medtronic, Inc., Minneapolis, MN, and W. L. Gore & Associates, Inc., Flagstaff, AZ.
References 1. A. D. Callow, ”Historical overview of experimental and clinical development of vascular grafts,” in Biological and Synthetic Vascular Prostheses, J. C . Stanley, et al. (eds.), Grune and Stratton, New York, 1982, pp. 11-36.
BROTHERS ET AL.
770 2.
3. 4.
5. 6.
7. 8. 9.
10. 11. 12.
13.
14.
15. 16.
17.
18.
19.
W. M. Abbott and R. P. Cambria, ”Elasticity and compliance of vascular grafts,” in Biological and Synthetic Vascular Prostheses, J. C. Stanley et al. (eds.), Grune and Stratton, New York, 1982, pp. 189-220. T. A. Belden, S. P. Schmidt, L. J. Falkow, and W. V. Sharp, “Endothelial cell seeding of small-diameter vascular grafts,” Trans. Am. SOC.Artif. Intern. Organs, 28, 173-177 (1982). C. D. Campbell, D. Goldfarb, and R. Roe, “A small arterial substitute: Expanded microporous polytetrafluoroethylene: Patency versus porosity,” Ann. Surg., 182, 138-143 (1975). D.Goldfarb, J. A. Houk, J. L. Moore, Sr., and D. L. Gain, “Graphiteexpanded polytetrafluoroethylene: An improved small artery prosthesis,“ Trans. Am. SOC.Artif. Intern. Organs, 23, 268-276 (1977). L. M. Graham, J. C. Stanley, and W. E. Burkel, “Improved patency of endothelial-cell-seeded, long, knitted Dacron and PTFE vascular prostheses,’’ ASAIO J., 8, 65-73 (1985). B.T. Allen, J.A. Long, M. J. Welch, K.T. Hopkins, G. A. Sicard, and R. E. Clark, ”Effect of aspirin therapy and its withdrawal on control and endothelial cell seeded grafts,” Surg. Forum, 34, 470-472 (1983). D. Annis, A. Bornat, R. 0. Edwards, A. Higham, B. Loveday, and J. Wilson, “An elastomeric vascular prosthesis,” Trans. Am. SOC.Artif. Intern. Organs, 24, 209-214 (1978). B. Dreyer, T. Akutsu, and W. J. Kolff, “Aortic grafts of polyurethane in dogs,” J. Appl. Physiol., 15, 18-22 (1960). D. J. Lyman, D. Albo, Jr., R. Jackson, and K. Knutson, “Development of small diameter vascular prostheses,” Trans. Am. Soc. Artif. Intern. Organs, 23, 253-256 (1977). V. Marinescu, E. Pausescu, and S. Carnaru, ”Long-term biological fate of polyurethane aortic prostheses,” Thorax, 206, 108-111 (1971). H. Martz, R. Paynter, S.B. Slimane, G. Beaudoin, R. Guidoin, J. Borzone, H. B. Simhon, R. Satin, and N. Sheiner, “Hydrophilic microporous polyurethane versus expanded PTFE grafts as substitutes in the carotid arteries of dogs. A limited study,” 1. Biomed. Muter. Res., 22, 6369 (1988). K. S. Murabayashi, H. Kambic, H. Harasaki, T. Morimoto, R. Yozu, and Y. Nose. “Fabrication and long-term imdantation of semi-comuliant small vascular prosthesis,” Trins. Am. h c . Artif. Intern. Organ’s, 31, 50-54 f1985). W. V. Sharp’, A. F. Finelli, W. H. Falor, and J. W. Ferraro, ”Latex vascular prostheses: Patency rate and neointimization related to prosthesis lining and electrical conductivity,” Circulation, 29, (supplj, 165-170 (1964). M. Wagner, G. Red, J. Teresi, and K. L. Kayser, ”Experimental observations on a new and inherently elastic material for sutures and vascular prostheses: Lycra,” Am. 1. Surg., 3, 838-841 (1966). R. A. White, E. W. White, E. L. Hanson, R. E Rohner, and W. R. Webb, ”Preliminary report: Evaluation of tissue ingrowth into experimental replamineform vascular prostheses,” Surgery, 79, 229-232 (1976). G. J. Wilson, D. C. MacGregor, P. Klement, J. M. Lee, P. J. del Nido, E. W. C. Wong, and J. Leidner, “Anisotropic polyurethane nonwoven conduits: A new approach to the design of a vascular prosthesis,” Trans. Am. Soc. Artif. Intern. Organs, 24, 260-268 (1983). W. F. Abbott and R. P. Cambria, ”Control of physical characteristics (elasticity and compliance) of vascular grafts, in Biologic and Synthetic Vascular Prostheses, J. C. Stanley et al. (eds.), Grune and Stratton, New York, 1982, p. 212. F. Hess, B. Braun, C . Jerusalem, R. van Det, S. Steeghs, S. Skotnicki, and P. Grande, “Endothelialization of polyurethane vascular prosthe-
POLYURETHANE GRAFTS
20.
21. 22.
23. 24. 25.
26. 27. 28.
771
ses implanted in the dog carotid and femoral artery,” J. Curdiovusc. Surg., 29, 458 (1988). H.Martz, R. Paynter, S. B. Slimane, G. Beaudoin, and R. Guiloin, ”Hydrophylic microporous polyurethane versus expanded PTFE grafts as substitutes in the carotid arteries of dogs. A limited study,” I. Biorned. Muter. Res., 22, 63 (1988). A. S. Chawla, P. Blais, I. Hinberg, and D. L. Johnson, “Experience with retrieved cardiac pacing leads,” Symp. Retrieval & Analysis of Surgical lmplunts & Biomuterials, 52, (1988). D.L. Coleman, H. L. C. Meuzelaar, T. R. Kessler, W. H. McClennen, J. M. Richards, and D. E. Gregonis, ”Retrieval and analysis of a clinical total artificial heart,” J. Biomed. Muter. Res., 20, 417 (1986). R. J. Thoma and R. E. Phillips, ”Note: Studies of poly(ether)urethane pacemaker lead insulation oxidation,” I . Biomed. Muter. Res., 21, 525 (1987). R. E. Marchant, Q. Zhao, J. M. Anderson, and A. Hiltner, “Degradation of a poly(ether urethane urea) elastomer: infra-red and XPS studies,” Polymer, 28, 2032 (1987). B.D. Ratner, K. W. Gladhill, and T. A. Horbett, ”Analysis of in vitro enzymatic and oxidative degradation of polyurethanes,” J. Biomed. Muter. Res., 22, 509 (1988). S.K. Phua, E. Castillo, J. M. Anderson, and A. Hiltner, “Biodegradation of a polyurethane in vitro,” I. Biomed. Muter. Res., 21, 231 (1987). Q.Zhao, R. E. Marchant, J. M. Anderson, and A. Hiltner, “Long term biodegradation in vitro of poly(ether urethane urea): a mechanical property study,” Polymer, 28, 2040 (1987). K.M. Miller and J. M. Anderson, ”In vitro stimulation of fibroblast activity by factors generated from human monocytes activated by biomedical polymers,” J . Biomed. Muter. Res., 23, 911 (1989).
Received September 22, 1989 Accepted December 14, 1989
