- Email: [email protected]
Small-diameter vascular prostheses: Two designs of PTFE and endothelial cell—seeded and nonseeded Dacron
Small-diameter vascular prostheses: Two designs of PTFE and endothelial cell-seeded and nonseeded Dacron Steven P. Schmidt, Ph.D., Timothy J. Hunter, B.S., Mark Hirko, B.S., Terry A. Belden, M.S., M. Michelle Evancho, B.S., William V. Sharp, M.D., and Duane L. Donovan, M.D., Akron, Ohio Despite numerous advances in biomaterials design and utilization, the perfect artificial small-vessel substitute has yet to be developed. Dacron and expanded polytetrafluoroethylene (PTFE) are two materials potentially appropriate for use as small-vessel prostheses. We report the patencies of endothelial cell-seeded and nonseeded 4 mm I.D. D a c r o n grafts and two designs of nonseeded 4 mm I.D. PTFE (Gore-Tex and Impra) in the carotid position in dogs. All graft lengths exceeded the calculated maximum critical length for the material being tested. Dacron grafts, both endothelial cell-seeded and nonseeded, achieved higher patencies than both designs of PTFE. Endothelial cell-seeded Dacron grafts achieved the highest patencies. Endothelium was present to a significant extent only on endothelial cell-seeded Dacron grafts. There was little pannus ingrowth o r midgraft pseudointima on nonseeded Dacron o r o n patent PTFE grafts although thrombus-free surface areas of patent PTFE grafts were high. These comparative data support the utility of endothelial cell seeding in achieving high patencies of small-diameter vasoalar grafts. (J VAse SugG 1985; 2:292-7.)
The inherent thrombogenicity of nonbiologic surfaces has limited the successful development and implementation of most small-diameter prostheses in vascular surgery. Several approaches have been utilized in the design and processing of artificial materials suitable as small-vessel prostheses to mimic the interface that naturally occurs between blood and blood vessel walls. Although the mechanisms of reduction of thrombogenicity remain obscure, several groups of investigators have achieved encouraging long-term graft patencies with albumin-coated surfaces L2 and carbon-coated surfaces. 3 It has also been suggested that various designs of expanded polytetrafluoroethylene (PTFE) allow the development of neoendothelium lining the graft lumen that morphologically and functionally imitates the native arterial lining. 4 In addition, the successful seeding of From the Departments of Vascular and SurgicalResearch,Akron City Hospital. Presented at the Ninth Annual Meeting, The ClevelandVascular Society, Cleveland, Ohio, May 15, 1984. Supported in part by grants from the Akron District Chapter of the American Heart Association and the Akron City Hospital Foundation. Reprint requests: Steven P. Schmidt, Ph.D., Vascular Research Laboratory,Akron City Hospital, 525 East Market St., Akron, OH 44309. 292
autologous endothelial cells onto small-diameter graft surfaces has recently been reported in the canine model, s,6 This article summarizes data collected from our laboratory on the short-term healing characteristics of two designs of 4 mm I.D. expanded PTFE (GoreTex and Impra) and endothelial cell-seeded and nonseeded 4 mm double-velour Microvel Dacron grafts. These data are from a larger, ongoing effort to define the requirements and protocols for the successful utilization of small-vessel artificial grafts in vascular surgery. MATERIAL AND METHODS Fifty-five adult mongrel dogs of mixed sex and age averaging 20 to 25 kg in weight were utilized in these experiments. All procedures of animal care complied with the "Principles of Laboratory Animal Care" and the "Guide for the Care and Use of Laboratory Animals" (NIH Publication No. 80-23, revised 1978). Six-centimeter segments of 4 mm I.D. Microvel, double-velour knitted Dacron grafts (Meadox Medicals, Inc., Oakland, N.J.) were microsurgically interpolated with end-to-end anastomoses sutured with 7-0 Prolene into the paired arteries of 37 dogs utilizing the following surgical procedure. Each
Voh,mc 2 Number 2 March 1985
Small-diameter vascular prostheses
293
Table I. Patencies of 4 mm Gore-Tex and Impra PTFE and endothelial cell-seeded and nonseeded 4 mm Dacron vascular grafts No medications 3 wk
4 wk
Aspirin and dipyridamole for 4 wk; 1 wk nonmedicated
0 (0/3) 66% (2/3)
0 (0/3) 33% (1/3)
44.4% (4/9) 44.4% (4/9)
80% (7/9) 44% (4/9)
80% (7/9) 56% (5/9)
100.0% (6/6) 83.0% (5/6)
2 wk Gore-Tex Impra Dacron Endothelial cell-seeded Nonseeded
66% (2/3) 100% (3/3) 76% (10/13) 87% (11/13)
Parentheses contain ratios of patent grafts to total number of grafts per group.
dog was anesthetized with sodium pentothal (20 mg/ kg IV) and maintained by inhalation with a mixture of 02 (4 L/min) and halothane (1% to 2%). All animals were hydrated with an IV solution of 5% dextrose and 0.9% saline solution during surgery. A 12 cm lateral incision was made just ventral to one external jugular vein and an 8 cm section of external jugular vein was excised following isolation and branch ligations. The vein was quickly cverted over a 6 mm diameter glass rod, stretched slightly up the rod, and tied firmly in place. The luminal venous endothelial cells thus exposed by the cversion technique were subsequently harvested by a modification of previously reported techniques. 7 The average number of endothelial cells derived by this techniquc was 6.75 x 105 cells as estimated by hemocytomcter counting. Both carotid arteries were isolated via a 15 cm ventral midline incision over the trachea while the endothelial cells were being harvested from the excised jugular vein. The 6 cm long nonsecded Dacron graft was preclotted with autologous blood by standard methodologies8 and was interpolated microsurgically into one transected carotid artery from which an equivalent length of vessel had been excised. The endothelial cell-seeded graft was interpolated into the contralateral carotid artery in a manner identical to that described for the nonsecdcd graft, except that the inoculum of harvested venous endothelial cells was incorporated into the preclot. Six of the dogs receiving the bilateral Dacron grafts were medicated with a daily dipyridamole regimen (50 mg twice a day) beginning 4 days preoperatively and with aspirin (325 mg daily) beginning 1 day preoperatively. This antiplatelet medication was continued for 4 weeks postoperatively and was subsequently discontinued for 1 week prior to the sacrifice of the dogs at 5 weeks. The remaining animals implanted with Dacron grafts did not receive any antiplatclet medications and were killed at 2
weeks (n = 13), 3 weeks (n = 9), and 4 weeks (n = 9) postoperatively. Eighteen dogs received a 6 cm length of Impra (mean pore size = 30 ~m) graft sutured into one carotid artery and an equal length of Gore-Tex (mean pore size = 22 ~m) sutured into the contralateral artery. The surgical procedure for the PTFE grafts was essentially the same as that described for Dacron except there was no jugular excision since these grafts were not endothelial cell seeded. Nine of the dogs receiving bilateral PTFE carotid implants were medicated with dipyridamole (50 mg twice a day) beginning 4 days preoperatively and aspirin (5 grains daily) beginning 1 day preoperatively. These medications were continued 4 weeks postoperatively and then discontinued for 1 week prior to sacrifice. In addition, nine dogs from this series were nonmedicated. From the latter group dogs were killed at 2 weeks (n = 3), 3 weeks (n = 3), and 4 weeks (n = 3) postoperatively. Prior to sacrifice each dog in this study was systemically heparinized (5000 units per dog) and the grafts were isolated and extirpated. Each graft was immediately opened longitudinally and carefully rinsed and flushed in Hanks' balanced salt solution to remove nonadherent blood cells. The grafts were immediately photographed with a 35 mm camera for comparisons of gross thrombus formation. Subsequently each graft was trisected longitudinally. One longitudinal section was stained for visualization of surface cndothelium by sequentially immersing the section for 5 minutes in a 10% glucose solution followed by 5 minutes of incubation in a 0.3% silver nitrate solution. The silver nitrate-stained grafts were stored in 10% formaldehyde and allowed to develop in light. These graft sections were examined with a Wild M5A stereomicroscope for estimation of surface area endothclialized. The second longitudinal section of each graft was reserved for hematoxylin-eosin staining for calcula-
294
Schmidt et al.
Fig. 1. Gross appearance of endothelial cell-seeded and nonseeded Dacron grafts (A) and patent PTFE GoreoTex and Impra grafts (B) at 5 weeks postoperatively.
tion of inner capsule thicknesses and for staining with peroxidase-antiperoxidase (PAP) for identification of the factor VIII antigen. The latter stain positively identified endothelial cells. The third longitudinal section of each graft was processed for scanning electron microscopy (SEM). The processing required glutaraldehyde fixation, dehydration, and critical point drying of the grafts. Graft sections were scanned with a JOEL-SMU 3 SEM. RESULTS The patencies of the 4 mm I.D. PTFE grafts and endothelial cell-seeded and nonseeded Dacron grafts at the time of postmortem examination are presented in Table I. Sixty-six percent of the 4 mm I.D. GoreTex grafts were patent 2 weeks postoperatively. However, all Gore-Tex grafts retrieved at 3 and 4 weeks postoperatively were occluded by gross thrombus formation lining the entire length of the lumen. Forty-four percent of the Gore-Tex grafts were pa-
Journal of VASCULAR SURGERY
tent in dogs medicated according to the described regimen. In contrast to the Gore-Tex grafts 100% of the Impra grafts harvested 2 weeks postoperatively were patent. However, with increasing postopcrativc. time in these nonmedicated dogs the patency of these PTFE grafts decreased to 66% at 3 weeks postoperatively and 33% at 4 weeks postoperatively. Fortyfour percent of the Impra grafts were patent at postmortem examination when the dogs were medicated with the antiplatelet regimen as described. Patencies of endothelial cell-seeded 4 mm Dacron grafts were high when nonmedicated dogs were evaluated at 2, 3, and 4 weeks postoperatively. One htmdred percent of the endothelial cell-seeded grafts were patent in dogs participating in the antiplatelet medication regimen. There was a high percentage (87%) of patent nonseeded Dacron grafts 2 weeks postoperatively. At' 3 and 4 weeks postoperatively, however, patencies of thcse grafts had fallen to 44% and 56%, respectively. Eighty-three percent of nonseedcd Dacron grafts were patent 5 weeks postoperatively when the dogs had been medicated with the antiplatelet agcnts for the first 4 postoperative weeks. Fig. 1 illustratcs the gross appearance of patent PTFE and endothelial cell-seeded and nonseeded Dacron grafts at 5 wecks postoperatively. These grafts wcrc extirpated from dogs medicated for 4 weeks with antiplatelet agcnts. Thrombus-free surface areas of the 4 mm I.D. PTFE grafts and endothelial cell-seeded and nonseeded Dacron grafts patent at the time of sacrifice. are listed in Table II. Thrombus-free surface arcas of the PTFE grafts that were patent when harvested were high. The thrombus-free surface areas of patent Impra grafts at 2, 3, and 4 wecks postoperatively were 79%, 89%, and 94%, respectively. There wcrc no differences in thrombus-frec surfaces of patcnt ~ endothelial cell-seeded grafts in nonmedicated dogs regardless of postoperative time of sacrifice. Ninetythree percent of endothelial cell-seedcd graft luminal surfaces were thrombus-free at sacrifice when the dogs were medicated with antiplatelet agents for the first 4 postoperative weeks. In contrast, even though patencies of nonseedcd Dacron grafts were high in medicated animals, the mean thrombus-free surfhcc area was only 27%. Thrombus-frcc surfaces of grafts from nonmedicated dogs wcrc also low. These values were 45%, 33%, and 22%, respectively, at 2, 3, and 4 weeks postoperatively. Endothelial cells were present to a significant extent only on endothelial cell-seeded grafts. Endothelium covered 22%, 53%, and 48% of endothelial cell-seedcd Dacron graft luminal surfaces at 2, 3, and
Volume 2 Number 2 March 1985
Small-diameter vascular prostheses
295
Table II. Thrombus-free surface areas of 4 mm Gore-Tex and Impra PTFE and endothelial cellseeded and nonseedcd 4 mm Dacron vascular grafts patent at time of sacrifice No medications
Gore-Tex Impra Dacron Endothelial cell-seeded Nonseeded
2 wk
3 wk
4 wk
Aspirin and dipyridamole for 4 wk; 1 wk nonmedicated
55% (2) 79% (3)
0 89% (2)
0 94% (1)
72% (3) 78% (3)
67% (10) 45% (11)
67% (7) 33% (4)
67% (7) 22% (5)
93% (6) 27% (6)
Numbers in parentheses indicate number of grafts evaluated.
4 weeks postoperatively in nonmedicated dogs as visualized by silver nitrate staining and SEM estimation. In contrast, endothelium derived from pannus ingrowth across the anastomoses covered only 4%, 8%, and 6% of control, nonseeded graft lumina in nonmedicated dogs at 2, 3, and 4 weeks postoperatively. There was no midgraft endothelium visualized by silver nitrate staining in these nonseeded grafts. In dogs medicated with aspirin and dipyridamole, endothelium lined 85% of the graft luminal surfaces. Pannus ingrowth constituted the extent of endothelium in medicated nonseeded grafts and covered only 18% of the luminal surfaces. There was virtually no pannus ingrowth or midgraft occurrence of endothelium in PTFE grafts regardless of the type of PTFE graft analyzed, the medication or nonmedication protocol, or the time of sacrifice. Inner capsule thicknesses of endothelial cellseeded grafts ranged from 10 to 50 ~m in the thinnest regions. Nonseeded control inner capsule thicknesses were typically greater, ranging between 50 to 200 Ixm in the thinnest regions and consisted of fibrin. Histologic investigations of PTFE grafts revealed the total absence of an inner capsule at the luminal interface between blood and graft. DISCUSSION
Reports from animal experiments have drawn attention to PTFE as a material potentially suitable for conduits in small-vessel replacement. Early reports noted 100% patency of 3 to 5 cm lengths of 3 mm I.D. IrFFE (Gore-Tex) grafts interpolated into the dissected femoral arteries of dogs. 9 No statistically significant differences were reported in patency rates between 4 cm lengths of 4 mm I.D. Gore-Tex, Impra, and Surgikos PTFE grafts. 1° The overall patency rate of these grafts in the femoral artery position was 62.5% when the grafts were examined 12 weeks postoperatively. A neoendothelial lining has been described on long-fibril Gore-Tex grafts 3 mm I.D. and
5 cm long when the grafts were tested in the dog's arterial circulation. H This neoendothelium has been further characterized from Gore-Tex grafts used for infrarenal aorta insertion. ~2 Brook et al. interpreted light and electron microscopic sections and described neoendothelium extending over the graft surfaces beginning 12 days postoperatively. In this latter report there was no mention of the internal diameter of PTFE graft utilized nor the positive confirmation of the neointimal cells as endothelial cells by immunofluorescence staining or PAP staining of the factor VIII antigen. In contrast to this study is the recent report of the absence of morphologic evidence of endothelium at the blood-prosthesis interface when 4 nun Gore-Tex prostheses were interpolated into the carotid arteries of dogs and examined up to 3 months after implantation. ~3 Nevertheless, conclusions from the majority of studies seem to indicate that the apparent success of small-diameter PTFE resides in its ability to promote rapid ncocndothclialization and consequently maintain patency more effectively. Tissue ingrowth is facilitated bv PTFE when internodal distances range between 20 to 30 ixm. 3 Our early experience with PTFE has been somewhat more discouraging than that reported by other authors using the canine artery model. In an effort to understand the patency performances we achieved with 4 mm I.D. PTFE, our attention was drawn to the concept of maximum critical lengths (Lc) proposed by Herring et al.~4,~s On the basis of the experience of these authors, the Lc for 4 mm I.D. PTFE grafts is 4.1 cm. Grafts with lengths less than the L~ will in all likelihood remain patent; this observation emphasizes that the lengths of the small-diameter PTFE grafts used for reconstruction are critical to the final determination of graft patency. Therefore the best experimental test of small-diameter PTFE graft patency is one that utilizes graft lengths greater than the L~. The 6 cm lengths of PTFE utilized in
296
Journal of VASCULAR SURGERY
Schmidt et al.
our studies provided a very rigorous test of the suitability of this material for small-vessel replacement. Our data indicated that when 6 cm lengths of 4 mm I.D. PTFE grafts are tested in nonmedicated dogs in the carotid position, patency is very poor. All Gore-Tex PTFE grafts occluded by 3 weeks postoperatively in our study. Impra grafts performed somewhat more effectively than Gore-Tex. Thirtythree percent of Impra grafts were still patent 4 weeks postoperatively. The Impra grafts utilized in our study did have a greater pore size than the Gore-Tex grafts. Although there was some histologic indication of a greater depth of ingrowth of cells from the adventitial side of the Impra grafts compared with Gore-Tex, even at 5 weeks postoperatively no pseudointima was observed in the patent Impra grafts. Therefore pore size probably did not contribute to the differences in patency we observed between these types of PTFE. Antiplatelet medication of the dogs for 4 weeks followed by 1 week of nonmedication improved graft performances of both types of PTFE slightly. Patency rates of 44% were observed in both Impra and Gore-Tex grafts following the medication protocol we employed. It is interesting to note that those PTFE grafts remaining patent had high thrombus-free surface areas, indicating that when smalldiameter PTFE grafting is successful, it is very nonthrombogenic. Our data would seem to confirm the concept of the maximum critical length for 4 mm PTFE. The implications of this concept are important if one attempts to utilize experimental data from dogs as one predictive measure of the success of human reconstruction with small-diameter PTFE. Many potential applications of small-vessel prostheses in human arteries might require lengths of synthetic 4 mm I.D. prosthesis far longer than the Lc. Our studies suggest that longer lengths of PTFE will certainly fail. The maximum critical length of knitted Dacron grafts of 4 mm I.D. calculated by Herring et al. Is is 3.25 cm. In contrast to the PTFE studies, most investigators testing 4 mm Dacron have utilized lengths of graft in excess of this maximum critical length. Patency appears to be limited for these smallvessel Dacron conduits unless antiplatelet medications are utilized. 16Recent investigations have sought to circumvent the inherent thrombogenicity of the small-vessel Dacron prostheses by seeding autologous endothelial cells onto the Dacron graft surfaces in the preclot. We initially reported success with endothelial cell seeding of 6 cm lengths of 4 mm I.D.
Microvel knitted Dacron. s More recently, wc reported the efficacy of endothelial cell seeding of 4 mm I.D. double-velour Dacron grafts under conditions of controlled low blood flows, y'17The comparative data we have accumulated indicate that patency in small-vessel Dacron grafts is better than in PTFE grafts when tested under identical experimental conditions. In addition, there was a more consistent pattern of patency in endothelial cell-seeded Dacron grafts compared with nonseeded Dacron grafts in nonmedicated dogs when the grafts were exanained 2 to 4 weeks postoperatively. When antiplatelct agents were administered daily for 4 weeks postoperatively, endothelial cell-seeded grafts were 100% patent. REFERENCES
1. Lyman DJ, Metcalf LC, Albo Jr D, Richards KF, Lamb J. The effect of chemical structure and surface properties of synthetic polymers on the coagulation of blood. III. In vivo adsorpt!on of proteins on polymer surfaces. Trans Am Soc Artif Intern Organs 1974; 20:474-8. 2. Sharp WV, Teague PC, Richenbacher WE. Thrombogenic potential of Dacron grafts after prior exposure to whole blood, plasma, or albumin. Trans Am Soc Artif Intern Organs 1979; 25:275-8. 3. Goldfarb D, Houk JA, Moore Sr L, Gain DL. Graphiteexpanded polytetrafluoroethylene: An improved small artery prosthesis. Trans Am Soc Artif Intern Organs 1977; 23:26874. 4. Campbell CD, Goldfarb D, Roe R. A small arterial substitute: Expanded microporous polytetrafluoroethylene: Patcncv versus porosity. Ann Surg 1975; 182:138-43. 5. Bclden TA, Schmidt SP, Falkow LJ, Sharp WV. Endothelial cell seeding of small-diameter vascular grafts. Trans ?un Soc Artif Intern Organs 1982; 28:173-7. 6. Stanley JC, Burkel WE, Ford JW, Vintcr DW, Kahn RH, Whitehouse WM, Graham LM. Enhanced patcncy of smalldiameter externally supported Dacron iliofemoral grafts seeded with endothelial cells. Surge~ 1982; 92:994-1005. 7. Schmidt SP, Hunter TJ, Sharp WV, Malindzak GS, Evancho MM. Endothelial cell-seeded 4 mm Dacron vascular grafts: Effects of blood flow manipulation through the grafts. J VASe SURG 1984; 1:434-41. 8. Yates SG, Barros D'Sa AAB, Berger KE, Fernandez LG, Woop SJ, Rittenhouse EA, Davis CC, Mansfield PB, Sauvage LR. The preclotting of porous arterial prostheses. Ann Surg 1978; 188:611-2. 9. Matsumoto H, Hasegawa T, Fuse K, Yamamoto M, Saigusa M. A new vascular prosthesis for a small caliber artery. Surgery 1973; 74:519-23. 10. Hastings OM, Jain KM, Hobson II RW, Swan KG. A prospective randomized study of three expanded polytetrafluoroethylene (FFFE) grafts as small arterial substitutes. Ann Surg 1978; 188:743-7. 11. Florian A, Cohn LH, Dammin GJ, Collins Jr J. Small vessel replacement with Gore-tex® (expanded polytetrafluoroethylcnc). Arch Surg 1976; 111:267-70.
Volume 2 Number 2 March 1985
12. Brook WH, Phornphibulaya P, Scott DF. Light and electron microscopy of polytetrafluoroethylene arterial grafts in dogs. Aust NZ J Surg 1980; 50:634-9. 13. Chignier E, Guidollet J, Serres M, Clendinnen G, Louisot P, Eloy R. Macromolecular, histological, ultrastructural and immunocytochemical characteristics of the neointima developed within PTFE vascular grafts. Experimental study in dogs. J Biomed Mater Res 1983; 17:623-36. 14. Herring MB, Dilley R, Peterson G, Wiggans J, Gardener A, Glover J. Graft material, length, and diameter determine the patency of small arterial prostheses in dogs. J Surg Res 1982; 32:138-42.
Small-diameter vascular prostheses 297
15. Herring M, Hubbard A, Baughman S, Smith D, Miller B, Dilley R, Gardner A, Glover J. Endothelium-lined small artery prostheses. A preliminary report. ASAIO J 1983; 6:93-102. 16. Sauvage LR, Walker MW, Bergcr K, Robel SB, Lischko MM, Yates SG, Logan GA. Current arterial prostheses. Experimental evaluation by implantation in the carotid and circumflex coronary arterics of the dog. Arch Surg 1979; 114:68791. 17. Hunter TJ, Schmidt SP, Sharp WV, Malindzak GS. Controlled flow studies in 4 mm endothelialized Dacron grafts. Trans Am Soc Artif Intern Organs 1983; 29:177-82.
The inherent thrombogenicity of nonbiologic surfaces has limited the successful development and implementation of most small-diameter prostheses in vascular surgery. Several approaches have been utilized in the design and processing of artificial materials suitable as small-vessel prostheses to mimic the interface that naturally occurs between blood and blood vessel walls. Although the mechanisms of reduction of thrombogenicity remain obscure, several groups of investigators have achieved encouraging long-term graft patencies with albumin-coated surfaces L2 and carbon-coated surfaces. 3 It has also been suggested that various designs of expanded polytetrafluoroethylene (PTFE) allow the development of neoendothelium lining the graft lumen that morphologically and functionally imitates the native arterial lining. 4 In addition, the successful seeding of From the Departments of Vascular and SurgicalResearch,Akron City Hospital. Presented at the Ninth Annual Meeting, The ClevelandVascular Society, Cleveland, Ohio, May 15, 1984. Supported in part by grants from the Akron District Chapter of the American Heart Association and the Akron City Hospital Foundation. Reprint requests: Steven P. Schmidt, Ph.D., Vascular Research Laboratory,Akron City Hospital, 525 East Market St., Akron, OH 44309. 292
autologous endothelial cells onto small-diameter graft surfaces has recently been reported in the canine model, s,6 This article summarizes data collected from our laboratory on the short-term healing characteristics of two designs of 4 mm I.D. expanded PTFE (GoreTex and Impra) and endothelial cell-seeded and nonseeded 4 mm double-velour Microvel Dacron grafts. These data are from a larger, ongoing effort to define the requirements and protocols for the successful utilization of small-vessel artificial grafts in vascular surgery. MATERIAL AND METHODS Fifty-five adult mongrel dogs of mixed sex and age averaging 20 to 25 kg in weight were utilized in these experiments. All procedures of animal care complied with the "Principles of Laboratory Animal Care" and the "Guide for the Care and Use of Laboratory Animals" (NIH Publication No. 80-23, revised 1978). Six-centimeter segments of 4 mm I.D. Microvel, double-velour knitted Dacron grafts (Meadox Medicals, Inc., Oakland, N.J.) were microsurgically interpolated with end-to-end anastomoses sutured with 7-0 Prolene into the paired arteries of 37 dogs utilizing the following surgical procedure. Each
Voh,mc 2 Number 2 March 1985
Small-diameter vascular prostheses
293
Table I. Patencies of 4 mm Gore-Tex and Impra PTFE and endothelial cell-seeded and nonseeded 4 mm Dacron vascular grafts No medications 3 wk
4 wk
Aspirin and dipyridamole for 4 wk; 1 wk nonmedicated
0 (0/3) 66% (2/3)
0 (0/3) 33% (1/3)
44.4% (4/9) 44.4% (4/9)
80% (7/9) 44% (4/9)
80% (7/9) 56% (5/9)
100.0% (6/6) 83.0% (5/6)
2 wk Gore-Tex Impra Dacron Endothelial cell-seeded Nonseeded
66% (2/3) 100% (3/3) 76% (10/13) 87% (11/13)
Parentheses contain ratios of patent grafts to total number of grafts per group.
dog was anesthetized with sodium pentothal (20 mg/ kg IV) and maintained by inhalation with a mixture of 02 (4 L/min) and halothane (1% to 2%). All animals were hydrated with an IV solution of 5% dextrose and 0.9% saline solution during surgery. A 12 cm lateral incision was made just ventral to one external jugular vein and an 8 cm section of external jugular vein was excised following isolation and branch ligations. The vein was quickly cverted over a 6 mm diameter glass rod, stretched slightly up the rod, and tied firmly in place. The luminal venous endothelial cells thus exposed by the cversion technique were subsequently harvested by a modification of previously reported techniques. 7 The average number of endothelial cells derived by this techniquc was 6.75 x 105 cells as estimated by hemocytomcter counting. Both carotid arteries were isolated via a 15 cm ventral midline incision over the trachea while the endothelial cells were being harvested from the excised jugular vein. The 6 cm long nonsecded Dacron graft was preclotted with autologous blood by standard methodologies8 and was interpolated microsurgically into one transected carotid artery from which an equivalent length of vessel had been excised. The endothelial cell-seeded graft was interpolated into the contralateral carotid artery in a manner identical to that described for the nonsecdcd graft, except that the inoculum of harvested venous endothelial cells was incorporated into the preclot. Six of the dogs receiving the bilateral Dacron grafts were medicated with a daily dipyridamole regimen (50 mg twice a day) beginning 4 days preoperatively and with aspirin (325 mg daily) beginning 1 day preoperatively. This antiplatelet medication was continued for 4 weeks postoperatively and was subsequently discontinued for 1 week prior to the sacrifice of the dogs at 5 weeks. The remaining animals implanted with Dacron grafts did not receive any antiplatclet medications and were killed at 2
weeks (n = 13), 3 weeks (n = 9), and 4 weeks (n = 9) postoperatively. Eighteen dogs received a 6 cm length of Impra (mean pore size = 30 ~m) graft sutured into one carotid artery and an equal length of Gore-Tex (mean pore size = 22 ~m) sutured into the contralateral artery. The surgical procedure for the PTFE grafts was essentially the same as that described for Dacron except there was no jugular excision since these grafts were not endothelial cell seeded. Nine of the dogs receiving bilateral PTFE carotid implants were medicated with dipyridamole (50 mg twice a day) beginning 4 days preoperatively and aspirin (5 grains daily) beginning 1 day preoperatively. These medications were continued 4 weeks postoperatively and then discontinued for 1 week prior to sacrifice. In addition, nine dogs from this series were nonmedicated. From the latter group dogs were killed at 2 weeks (n = 3), 3 weeks (n = 3), and 4 weeks (n = 3) postoperatively. Prior to sacrifice each dog in this study was systemically heparinized (5000 units per dog) and the grafts were isolated and extirpated. Each graft was immediately opened longitudinally and carefully rinsed and flushed in Hanks' balanced salt solution to remove nonadherent blood cells. The grafts were immediately photographed with a 35 mm camera for comparisons of gross thrombus formation. Subsequently each graft was trisected longitudinally. One longitudinal section was stained for visualization of surface cndothelium by sequentially immersing the section for 5 minutes in a 10% glucose solution followed by 5 minutes of incubation in a 0.3% silver nitrate solution. The silver nitrate-stained grafts were stored in 10% formaldehyde and allowed to develop in light. These graft sections were examined with a Wild M5A stereomicroscope for estimation of surface area endothclialized. The second longitudinal section of each graft was reserved for hematoxylin-eosin staining for calcula-
294
Schmidt et al.
Fig. 1. Gross appearance of endothelial cell-seeded and nonseeded Dacron grafts (A) and patent PTFE GoreoTex and Impra grafts (B) at 5 weeks postoperatively.
tion of inner capsule thicknesses and for staining with peroxidase-antiperoxidase (PAP) for identification of the factor VIII antigen. The latter stain positively identified endothelial cells. The third longitudinal section of each graft was processed for scanning electron microscopy (SEM). The processing required glutaraldehyde fixation, dehydration, and critical point drying of the grafts. Graft sections were scanned with a JOEL-SMU 3 SEM. RESULTS The patencies of the 4 mm I.D. PTFE grafts and endothelial cell-seeded and nonseeded Dacron grafts at the time of postmortem examination are presented in Table I. Sixty-six percent of the 4 mm I.D. GoreTex grafts were patent 2 weeks postoperatively. However, all Gore-Tex grafts retrieved at 3 and 4 weeks postoperatively were occluded by gross thrombus formation lining the entire length of the lumen. Forty-four percent of the Gore-Tex grafts were pa-
Journal of VASCULAR SURGERY
tent in dogs medicated according to the described regimen. In contrast to the Gore-Tex grafts 100% of the Impra grafts harvested 2 weeks postoperatively were patent. However, with increasing postopcrativc. time in these nonmedicated dogs the patency of these PTFE grafts decreased to 66% at 3 weeks postoperatively and 33% at 4 weeks postoperatively. Fortyfour percent of the Impra grafts were patent at postmortem examination when the dogs were medicated with the antiplatelet regimen as described. Patencies of endothelial cell-seeded 4 mm Dacron grafts were high when nonmedicated dogs were evaluated at 2, 3, and 4 weeks postoperatively. One htmdred percent of the endothelial cell-seeded grafts were patent in dogs participating in the antiplatelet medication regimen. There was a high percentage (87%) of patent nonseeded Dacron grafts 2 weeks postoperatively. At' 3 and 4 weeks postoperatively, however, patencies of thcse grafts had fallen to 44% and 56%, respectively. Eighty-three percent of nonseedcd Dacron grafts were patent 5 weeks postoperatively when the dogs had been medicated with the antiplatelet agcnts for the first 4 postoperative weeks. Fig. 1 illustratcs the gross appearance of patent PTFE and endothelial cell-seeded and nonseeded Dacron grafts at 5 wecks postoperatively. These grafts wcrc extirpated from dogs medicated for 4 weeks with antiplatelet agcnts. Thrombus-free surface areas of the 4 mm I.D. PTFE grafts and endothelial cell-seeded and nonseeded Dacron grafts patent at the time of sacrifice. are listed in Table II. Thrombus-free surface arcas of the PTFE grafts that were patent when harvested were high. The thrombus-free surface areas of patent Impra grafts at 2, 3, and 4 wecks postoperatively were 79%, 89%, and 94%, respectively. There wcrc no differences in thrombus-frec surfaces of patcnt ~ endothelial cell-seeded grafts in nonmedicated dogs regardless of postoperative time of sacrifice. Ninetythree percent of endothelial cell-seedcd graft luminal surfaces were thrombus-free at sacrifice when the dogs were medicated with antiplatelet agents for the first 4 postoperative weeks. In contrast, even though patencies of nonseedcd Dacron grafts were high in medicated animals, the mean thrombus-free surfhcc area was only 27%. Thrombus-frcc surfaces of grafts from nonmedicated dogs wcrc also low. These values were 45%, 33%, and 22%, respectively, at 2, 3, and 4 weeks postoperatively. Endothelial cells were present to a significant extent only on endothelial cell-seeded grafts. Endothelium covered 22%, 53%, and 48% of endothelial cell-seedcd Dacron graft luminal surfaces at 2, 3, and
Volume 2 Number 2 March 1985
Small-diameter vascular prostheses
295
Table II. Thrombus-free surface areas of 4 mm Gore-Tex and Impra PTFE and endothelial cellseeded and nonseedcd 4 mm Dacron vascular grafts patent at time of sacrifice No medications
Gore-Tex Impra Dacron Endothelial cell-seeded Nonseeded
2 wk
3 wk
4 wk
Aspirin and dipyridamole for 4 wk; 1 wk nonmedicated
55% (2) 79% (3)
0 89% (2)
0 94% (1)
72% (3) 78% (3)
67% (10) 45% (11)
67% (7) 33% (4)
67% (7) 22% (5)
93% (6) 27% (6)
Numbers in parentheses indicate number of grafts evaluated.
4 weeks postoperatively in nonmedicated dogs as visualized by silver nitrate staining and SEM estimation. In contrast, endothelium derived from pannus ingrowth across the anastomoses covered only 4%, 8%, and 6% of control, nonseeded graft lumina in nonmedicated dogs at 2, 3, and 4 weeks postoperatively. There was no midgraft endothelium visualized by silver nitrate staining in these nonseeded grafts. In dogs medicated with aspirin and dipyridamole, endothelium lined 85% of the graft luminal surfaces. Pannus ingrowth constituted the extent of endothelium in medicated nonseeded grafts and covered only 18% of the luminal surfaces. There was virtually no pannus ingrowth or midgraft occurrence of endothelium in PTFE grafts regardless of the type of PTFE graft analyzed, the medication or nonmedication protocol, or the time of sacrifice. Inner capsule thicknesses of endothelial cellseeded grafts ranged from 10 to 50 ~m in the thinnest regions. Nonseeded control inner capsule thicknesses were typically greater, ranging between 50 to 200 Ixm in the thinnest regions and consisted of fibrin. Histologic investigations of PTFE grafts revealed the total absence of an inner capsule at the luminal interface between blood and graft. DISCUSSION
Reports from animal experiments have drawn attention to PTFE as a material potentially suitable for conduits in small-vessel replacement. Early reports noted 100% patency of 3 to 5 cm lengths of 3 mm I.D. IrFFE (Gore-Tex) grafts interpolated into the dissected femoral arteries of dogs. 9 No statistically significant differences were reported in patency rates between 4 cm lengths of 4 mm I.D. Gore-Tex, Impra, and Surgikos PTFE grafts. 1° The overall patency rate of these grafts in the femoral artery position was 62.5% when the grafts were examined 12 weeks postoperatively. A neoendothelial lining has been described on long-fibril Gore-Tex grafts 3 mm I.D. and
5 cm long when the grafts were tested in the dog's arterial circulation. H This neoendothelium has been further characterized from Gore-Tex grafts used for infrarenal aorta insertion. ~2 Brook et al. interpreted light and electron microscopic sections and described neoendothelium extending over the graft surfaces beginning 12 days postoperatively. In this latter report there was no mention of the internal diameter of PTFE graft utilized nor the positive confirmation of the neointimal cells as endothelial cells by immunofluorescence staining or PAP staining of the factor VIII antigen. In contrast to this study is the recent report of the absence of morphologic evidence of endothelium at the blood-prosthesis interface when 4 nun Gore-Tex prostheses were interpolated into the carotid arteries of dogs and examined up to 3 months after implantation. ~3 Nevertheless, conclusions from the majority of studies seem to indicate that the apparent success of small-diameter PTFE resides in its ability to promote rapid ncocndothclialization and consequently maintain patency more effectively. Tissue ingrowth is facilitated bv PTFE when internodal distances range between 20 to 30 ixm. 3 Our early experience with PTFE has been somewhat more discouraging than that reported by other authors using the canine artery model. In an effort to understand the patency performances we achieved with 4 mm I.D. PTFE, our attention was drawn to the concept of maximum critical lengths (Lc) proposed by Herring et al.~4,~s On the basis of the experience of these authors, the Lc for 4 mm I.D. PTFE grafts is 4.1 cm. Grafts with lengths less than the L~ will in all likelihood remain patent; this observation emphasizes that the lengths of the small-diameter PTFE grafts used for reconstruction are critical to the final determination of graft patency. Therefore the best experimental test of small-diameter PTFE graft patency is one that utilizes graft lengths greater than the L~. The 6 cm lengths of PTFE utilized in
296
Journal of VASCULAR SURGERY
Schmidt et al.
our studies provided a very rigorous test of the suitability of this material for small-vessel replacement. Our data indicated that when 6 cm lengths of 4 mm I.D. PTFE grafts are tested in nonmedicated dogs in the carotid position, patency is very poor. All Gore-Tex PTFE grafts occluded by 3 weeks postoperatively in our study. Impra grafts performed somewhat more effectively than Gore-Tex. Thirtythree percent of Impra grafts were still patent 4 weeks postoperatively. The Impra grafts utilized in our study did have a greater pore size than the Gore-Tex grafts. Although there was some histologic indication of a greater depth of ingrowth of cells from the adventitial side of the Impra grafts compared with Gore-Tex, even at 5 weeks postoperatively no pseudointima was observed in the patent Impra grafts. Therefore pore size probably did not contribute to the differences in patency we observed between these types of PTFE. Antiplatelet medication of the dogs for 4 weeks followed by 1 week of nonmedication improved graft performances of both types of PTFE slightly. Patency rates of 44% were observed in both Impra and Gore-Tex grafts following the medication protocol we employed. It is interesting to note that those PTFE grafts remaining patent had high thrombus-free surface areas, indicating that when smalldiameter PTFE grafting is successful, it is very nonthrombogenic. Our data would seem to confirm the concept of the maximum critical length for 4 mm PTFE. The implications of this concept are important if one attempts to utilize experimental data from dogs as one predictive measure of the success of human reconstruction with small-diameter PTFE. Many potential applications of small-vessel prostheses in human arteries might require lengths of synthetic 4 mm I.D. prosthesis far longer than the Lc. Our studies suggest that longer lengths of PTFE will certainly fail. The maximum critical length of knitted Dacron grafts of 4 mm I.D. calculated by Herring et al. Is is 3.25 cm. In contrast to the PTFE studies, most investigators testing 4 mm Dacron have utilized lengths of graft in excess of this maximum critical length. Patency appears to be limited for these smallvessel Dacron conduits unless antiplatelet medications are utilized. 16Recent investigations have sought to circumvent the inherent thrombogenicity of the small-vessel Dacron prostheses by seeding autologous endothelial cells onto the Dacron graft surfaces in the preclot. We initially reported success with endothelial cell seeding of 6 cm lengths of 4 mm I.D.
Microvel knitted Dacron. s More recently, wc reported the efficacy of endothelial cell seeding of 4 mm I.D. double-velour Dacron grafts under conditions of controlled low blood flows, y'17The comparative data we have accumulated indicate that patency in small-vessel Dacron grafts is better than in PTFE grafts when tested under identical experimental conditions. In addition, there was a more consistent pattern of patency in endothelial cell-seeded Dacron grafts compared with nonseeded Dacron grafts in nonmedicated dogs when the grafts were exanained 2 to 4 weeks postoperatively. When antiplatelct agents were administered daily for 4 weeks postoperatively, endothelial cell-seeded grafts were 100% patent. REFERENCES
1. Lyman DJ, Metcalf LC, Albo Jr D, Richards KF, Lamb J. The effect of chemical structure and surface properties of synthetic polymers on the coagulation of blood. III. In vivo adsorpt!on of proteins on polymer surfaces. Trans Am Soc Artif Intern Organs 1974; 20:474-8. 2. Sharp WV, Teague PC, Richenbacher WE. Thrombogenic potential of Dacron grafts after prior exposure to whole blood, plasma, or albumin. Trans Am Soc Artif Intern Organs 1979; 25:275-8. 3. Goldfarb D, Houk JA, Moore Sr L, Gain DL. Graphiteexpanded polytetrafluoroethylene: An improved small artery prosthesis. Trans Am Soc Artif Intern Organs 1977; 23:26874. 4. Campbell CD, Goldfarb D, Roe R. A small arterial substitute: Expanded microporous polytetrafluoroethylene: Patcncv versus porosity. Ann Surg 1975; 182:138-43. 5. Bclden TA, Schmidt SP, Falkow LJ, Sharp WV. Endothelial cell seeding of small-diameter vascular grafts. Trans ?un Soc Artif Intern Organs 1982; 28:173-7. 6. Stanley JC, Burkel WE, Ford JW, Vintcr DW, Kahn RH, Whitehouse WM, Graham LM. Enhanced patcncy of smalldiameter externally supported Dacron iliofemoral grafts seeded with endothelial cells. Surge~ 1982; 92:994-1005. 7. Schmidt SP, Hunter TJ, Sharp WV, Malindzak GS, Evancho MM. Endothelial cell-seeded 4 mm Dacron vascular grafts: Effects of blood flow manipulation through the grafts. J VASe SURG 1984; 1:434-41. 8. Yates SG, Barros D'Sa AAB, Berger KE, Fernandez LG, Woop SJ, Rittenhouse EA, Davis CC, Mansfield PB, Sauvage LR. The preclotting of porous arterial prostheses. Ann Surg 1978; 188:611-2. 9. Matsumoto H, Hasegawa T, Fuse K, Yamamoto M, Saigusa M. A new vascular prosthesis for a small caliber artery. Surgery 1973; 74:519-23. 10. Hastings OM, Jain KM, Hobson II RW, Swan KG. A prospective randomized study of three expanded polytetrafluoroethylene (FFFE) grafts as small arterial substitutes. Ann Surg 1978; 188:743-7. 11. Florian A, Cohn LH, Dammin GJ, Collins Jr J. Small vessel replacement with Gore-tex® (expanded polytetrafluoroethylcnc). Arch Surg 1976; 111:267-70.
Volume 2 Number 2 March 1985
12. Brook WH, Phornphibulaya P, Scott DF. Light and electron microscopy of polytetrafluoroethylene arterial grafts in dogs. Aust NZ J Surg 1980; 50:634-9. 13. Chignier E, Guidollet J, Serres M, Clendinnen G, Louisot P, Eloy R. Macromolecular, histological, ultrastructural and immunocytochemical characteristics of the neointima developed within PTFE vascular grafts. Experimental study in dogs. J Biomed Mater Res 1983; 17:623-36. 14. Herring MB, Dilley R, Peterson G, Wiggans J, Gardener A, Glover J. Graft material, length, and diameter determine the patency of small arterial prostheses in dogs. J Surg Res 1982; 32:138-42.
Small-diameter vascular prostheses 297
15. Herring M, Hubbard A, Baughman S, Smith D, Miller B, Dilley R, Gardner A, Glover J. Endothelium-lined small artery prostheses. A preliminary report. ASAIO J 1983; 6:93-102. 16. Sauvage LR, Walker MW, Bergcr K, Robel SB, Lischko MM, Yates SG, Logan GA. Current arterial prostheses. Experimental evaluation by implantation in the carotid and circumflex coronary arterics of the dog. Arch Surg 1979; 114:68791. 17. Hunter TJ, Schmidt SP, Sharp WV, Malindzak GS. Controlled flow studies in 4 mm endothelialized Dacron grafts. Trans Am Soc Artif Intern Organs 1983; 29:177-82.