Expanded Microporous Polytetrafluoroethylene as Canine Arterial Bypass or Replacement Graft Bruce D. Walley, MD; Paul M. James, Jr, MD; Jesse H. Meredith, MD; Stuart K. Todd, MD; Nicholas Ogburn
microporous polytetrafluoroethylene (PTFE) has and clinically for small-vessel bypass the results vary widely. This study or assessed short (4-cm) femoral and carotid artery replacement PTFE grafts and longer (12-cm) femoral artery bypass PTFE grafts that crossed a flexion crease. Pore size at the blood-graft interface ranged from 10 to 30 \g=m\. At the end of three months, overall patency rate was 22%. No long bypass grafts remained patent. The pore size and type of anastomosis did not affect patency. The occlusion was always thrombotic, associated with fibrin at the suture line. Nine of 20 short carotid grafts (45%) and five of 22 short femoral grafts (22%) remained patent. These poor results indicate that further experimental studies are needed before PTFE is used clinically when an alternative exists. (Arch Surg 113:863-866, 1978) \s=b\ Expanded
been used
experimentally replacement grafting;
of
acceptance synthetic Thereplacement began 1952,
material for arterial when Voorhees and asso¬ ciates1 reported using Vinyon-"N" for successful bridging of a large-artery defect. Comparative experimental stud¬ ies2 and clinical experience have led to the successful current use of Dacron and Teflon for this purpose, both substances being strong and durable. The autogenous saphenous vein is still the graft of choice for replacement of medium-size vessels when available. In a prosthesis to replace recent years, the search for medium-size arteries (5- to 8-mm internal diameter) has spurred the development of numerous composite grafts of in
Accepted
for publication Feb 10, 1978. From the Department of Surgery, Bowman Gray School of Medicine, Wake Forest University, Winston-Salem, NC. Reprint requests to Department of Surgery, Bowman Gray School of Medicine, Winston-Salem, NC 27103 (Dr Walley).
synthetic material'·4 when venous autografts Purely biological grafts of heterologous collagen and homologous umbilical vein reinforced with a polyester mesh are also being used for medium-size artery replacements. Early results with many of these new grafts have been promising, but none of them have gained widespread acceptance. Replacement of small arteries (5-mm or less internal diameter) has met with even more limited success. Harri¬ collagen are
and
not available. '
in 1958 and Jacobson and associates" in 1963 concluded that only autogenous tissue (artery or vein) should be used for replacement of such small arteries, since none of the synthetic prostheses studied were satisfactory. Over the past five years, however, numerous investiga¬ tors have reported experimental success using expanded microporous polytetrafluoroethylene (PTFE) for smallvessel (3 to 4 mm) replacement in dogs.71" Expanded microporous PTFE (Teflon) is composed of nodes of PTFE interconnected by fine fibrils of PTFE forming a lattice. Discrepancy in patency rates (53% to 100%) has been ascribed to variations in the length of the fibrils between the nodes—or the "pore size"—which affect tissue ingrowth. Investigators disagree on what is the optimum "pore size," Campbell and associates8 having a patency rate of 96% with small-pore (22-µ) and 53% with larger-pore grafts, Florian and associates" and Matsumoto and asso¬ ciates'" having a patency rate of 100% with the more porous, longer-fibril (100-µ) grafts. Manufacture of the material is such now that pore size at the blood-graft interface can be smaller than the pore size on the outer surface of the graft. This prospective double-blind study was undertaken to compare patency rates of PTFE grafts of different lengths son2
Downloaded From: http://archsurg.jamanetwork.com/ by a University of California - San Diego User on 06/04/2015
and with different pore sizes. Both arterial and bypass models were studied.
replacement
MATERIALS AND METHODS
Thirty-nine healthy mongrel dogs of both sexes, averaging 22 kg in weight, were used in the study. Animals were anesthetized with 30 mg/kg of sodium pentobarbital and ventilated with a Harvard respirator through an endotracheal tube. Penicillin potas¬ sium, 500,000 units, was given just before and again just after the operative procedure. Normal saline, 50 mg/kg, was given intrave¬ nously during the operation. The abdomen and both femoral areas or the anterior cervical region was shaved as indicated for the operation to be done. The operative field was prepared with povidone-iodine and procedures were done under aseptic conditions. After adequate exposure of the arteries to be operated on, heparin sodium, 2 mg/kg, was administered intravenously. Ten minutes later vascular clamps were applied. The arterial anastomoses were performed with a continuous suture of 6-0 polyethylene. All anastomoses were done by experienced surgeons using 2 magnification loops and micro¬ surgery instruments. The prosthetic grafts utilized were expanded PTFE (Gore-Tex) grafts with an inside diameter of 4 mm and a wall thickness of 0.3 to 0.7 mm. Half of the grafts had an average fibril length of 10 µ at the blood-graft interface; the other half had a fibril length averaging 30 µ at the blood-graft interface. A graft from one group was randomly selected and placed on the left or right side of each animal, with a graft from the other group being placed in the contralateral side. The specifications of the graft were not known to the surgeons until the study had been completed.
Group
1
In the five control dogs, a portion of splenic vein approximately 10 to 12 cm in length and approximately 4 mm in inside diameter was harvested and sewn end-to-side to the intraabdominal portion of the right femoral artery. A tunnel was made under the inguinal ligament, and the graft was brought through it and anastomosed end-to-side to the femoral artery in the upper leg. The femoral artery was then ligated 1 cm below the upper anastomosis and 1 cm above the lower anastomosis.
Group
2
In ten dogs, 12-cm-long PTFE grafts were beveled at 45° angles on both ends and sewn end-to-side to the femoral artery in the manner described for group 1. Ten grafts were sewn end-to-end to
the femoral artery, again being brought through an inguinal tunnel from the intraabdominal portion to the upper leg portion of the artery.
Group
3
Both femoral (12 dogs) or both carotid (12 dogs) arteries were isolated, a 3- to 4-cm portion was excised, and 4-cm-long PTFE grafts were sewn end-to-end into the defect.
Completion
of
Operation
and
Follow-up
When the anastomoses were accomplished and the vascular clamps removed, the effect of heparin was reversed with 2 mg/kg of protamine sulfate. The wounds were irrigated with normal saline and closed in layers with polyglycolic acid (Dexon). Postoperatively, the dogs were given chloramphenicol, 7 mg/kg, for three days. With the animals under sedation, patency of the grafts was assessed with the Doppler ultrasound flowmeter 48 hours postoperatively, weekly thereafter for three weeks, and again two months postoperatively. If no signal was audible, the
animal was anesthetized with pentobarbital sodium and the occlusion verified by transection of the vessel distal to the anastomosis. Animals with audible Doppler signals were followed up for three months, with patency or occlusion being verified at death.
Pathology All grafts were harvested and preserved for microscopic evalu¬ ation. Grafts were stained with hematoxylin-eosin and examined by a pathologist unaware of the specifications of the grafts used.
RESULTS The results in three dogs are excluded: two of the dogs had infected groin wounds and the third died of intussus¬ ception before the end of the study. Of the remaining dogs, the patency rate overall for the PTFE grafts was 22%.
Group
1
The five autogenous venous grafts remained patent (100%) for the duration of the three-month study. At the animals' deaths, the anastomoses were widely patent with¬ out evidence of stenosis. There was moderate dilation of the vein graft in all cases. Microscopically, there was minimal intimai thickening and fibrosis.
Group
2
None of the 20 synthetic expanded PTFE grafts placed in the iliofemoral area remained patent for three months. Two occluded 48 hours after implantation; two grafts maintained flow for two months but were subsequently found to be occluded at death (three months); most occluded between two days and one month postimplanta¬ tion. Neither the fibril length at the flow surface nor the manner in which the graft was sewn (end-to-side vs end-toend) appeared to affect the patency rate. Thrombosis was associated with an occluding, circumferential deposit of fibrin at the anastomoses. Microscopically, the fibrous sheath was seen to be loosely adherent to the graft with minimal ingrowth of fibrous tissue.
Group
3
Nine of the 20 grafts placed in the carotid
area
remained
patent for three months (45%). Two of these grafts
were
approximately
50% occluded with thrombosis. Doppler evaluation of these short carotid grafts was technically difficult, so we cannot determine how long the remaining 11 grafts remained patent before thrombosis occurred. Five of the 22 short grafts placed in the femoral area remained patent for three months (22%). Of the grafts that occluded, eight did so between two and three months after
implantation.
Pathology Gross examination of the occluded grafts showed throm¬ bus that appeared to originate at the suture lines. The grafts that remained patent had a smooth, glistening flow surface with some stenosis at the suture lines. Microscopi¬ cally, the patent grafts had a fairly smooth and cellular
Downloaded From: http://archsurg.jamanetwork.com/ by a University of California - San Diego User on 06/04/2015
flow surface that appeared to originate from through the interstices of the grafts.
growth
COMMENT
Over the past 20 years, there has been extensive
screen¬
ing and development of materials for vascular reconstruc¬ tion. Pioneers such as Wesolowski, Sawyer, Sauvage, and colleagues11" have contributed greatly to our knowledge of how synthetic materials are incorporated into the arterial tree; DeBakey,"1' Sanger,1" Edwards,17 Voorhees and colleagues,1 to our knowledge of techniques and of the complications of graft placement in various parts of the body. Within 24 hours of implantation of a graft, fibrin and blood elements coat both surfaces of the graft. Over the next few weeks, the outer fibrin coating becomes replaced by an encapsulating sheath of granulation tissue; the inner fibrin layer must be converted to fibrous tissue by ingrowth of granulation tissue through the interstices of the graft.11 Experiments have shown the inner capsule to be complete at one month in swine and at four months in dogs." The inner capsule may never be complete in
humans." When this inner, fibrous-tissue "scar" contracts, its capillary blood supply in the wall of the graft may be disrupted, and this, in turn, may lead to sloughing of a portion of the inner lining, creating a more thrombogenic fibrin-flow surface. If the graft wall is porous, ingrowth of capillaries will allow rehealing in these areas." Thus, porosity is a primary determinant of the continued integrity of the flow surface of the prosthesis. However, the graft must not be so porous that it allows hemorrhage through the interstices. Other specifications for an ideal synthetic graft as outlined by Wesolowski et al1-' are no allergenic potential, durability for prolonged periods, and desirable handling properties. An ideal vascular substitute should also not kink across flexion sites, resist infection, and possess a permanent nonthrombogenic surface at the blood-graft interface.1 Phillips and co-workers18 have reported that synthetic grafts in small vessels probably fail because of thrombosis at the suture line. Our results support this finding. Physical stress at the suture line appears to be an important factor leading to that thrombosis.'9 Bypassing or replacing the femoral artery in dogs has proved a difficult test for prosthetic grafts. Although patency rates of 100%" and 93%" have been reported with autogenous veins used for this purpose, only 36% patency was achieved with knitted Dacron, 25% with woven Teflon, and 60% with knitted Teflon." Collagen tubes (bovine heterografts) had a patency rate of 18%-" to 50%21 when used to bypass the femoral artery of dogs. Since autoge¬ nous veins are not always available, the recent reports of high patency rates with grafts of expanded PTFE have excited interest in this material for small-vessel grafting. However, we found that we could not duplicate those high patency rates, having an overall patency rate after carotid artery replacement and femoral artery bypass or replace¬ ment of only 22%. Similarly poor results have been mentioned by others.--"1 Our grafts had either a highdensity flow surface of 10-µ fibril length or a flow surface
of 30-µ fibril length, both considered small-pore surfaces when considered against grafts with 100-µ fibril lengths. Our results were equally poor with both pore sizes, and suture line thrombosis occurred regardless of the extent of graft healing. Our poor results led us to compare our methods with those of investigators reporting patency rates as high as 96%8 and 100%."·"' Unlike the previous models, 20 of our grafts were implanted so that they passed beneath the inguinal ligament, thus crossing a joint (0% patency with PTFE graft; 100% patency with autogenous vein graft). Crossing a flexion crease has been shown to adversely affect other synthetic grafts of small arteries.'1 The grafts we used were also longer than those in previous studies. Our follow-up of all 62 grafts was for 90 days or until occlusion. A previous study by Campbell et al7 had an average follow-up of 18.5 days; animals were killed at regular intervals and only eight grafts were left to be followed up for longer than four months. Palpation was used to determine patency in other studies; in our study, Doppler flow measurements were used because palpation was found to be unreliable. We do not believe our suture technique was at fault, since all of the grafts were placed by surgeons who had participated in the experimental anastomosis of more than 100 autologous arteries and veins with only one confirmed technical error. Failure of a prosthetic graft for arterial replacement in humans may be followed by devastating consequences. Experiences with Sparks' mandril graft21 and some of the ultralightweight Dacron grafts2" are tragic examples of this fact. Expanded PTFE is presently being used clinically to replace and bypass medium-sized arteries, in some instances to cross a flexion crease. Aneurysms of such grafts have already been seen."·28 We are concerned about this clinical use when the experimental results are so varied and the factors causing graft failure are still unknown. This
investigation
Surgery.
Grafts
were
was
supported
provided by
in part
by
the Bradshaw
W. L. Gore and Associates,
Fellowship in
Flagstaff,
Ariz.
References 1. Voorhees AB Jr, Jaretzki A III, Blakemore AH: The use of tubes constructed from Vinyon-"N" cloth in bridging arterial defects: A preliminary report. Ann Surg 135:332-336, 1952. 2. Harrison JH: Synthetic materials as vascular prostheses: I. A comparative study in small vessels of nylon, Dacron, Orlon, Ivalon sponge, and Teflon. Am J Surg 95:3-15, 1958. 3. Parsonnet V, Alpert J, Brief DK: Autogenous polypropylene-supported collagen tubes for long-term arterial replacement. Surgery 70:935-939, 1971. 4. Sparks CH: Silicone mandril method of femoropopliteal artery bypass: Clinical experience and surgical technics. Am J Surg 124:244-249, 1972. 5. Dardik H, Ibrahim IM, Sprayregen S, et al: Clinical experience with modified human umbilical cord vein for arterial bypass. Surgery 79:618-624, 1976. 6. Jacobson JH II, Rosario ES, Katsumura T: Influence of prosthesis diameter in small artery replacement. Circulation 28:742-743, 1963. 7. Campbell CD, Goldfarb D, Detton DD, et al: Expanded polytetrafluoroethylene as a small artery substitute. Trans Am Soc Artif Intern Organs 20:86-90, 1974. 8. Campbell CD, Goldfarb D, Roe R: A small arterial substitute:
Expanded microporous polytetrafluoroethylene: Patency Surg 182:138-143, 1975.
Ann
Downloaded From: http://archsurg.jamanetwork.com/ by a University of California - San Diego User on 06/04/2015
versus
porosity.
9. Florian A, Cohn LH, Dammin GJ, et al: Small vessel replacement with Gore-Tex (expanded polytetrafluoroethylene). Arch Surg 111:267-270, 1976. 10. Matsumoto H, Hasegawa T, Fuse K, et al: A new vascular prosthesis for a small caliber artery. Surgery 74:519-523, 1973. 11. Berger K, Sauvage LR, Rao AM, et al: Healing of arterial prostheses in man: Its incompleteness. Am Surg 175:118-127, 1972. 12. Wesolowski SA, Fries CC, Domingo RT, et al: The compound prosthetic vascular graft: A pathologic survey. Surgery 53:19-44, 1963. 13. Wesolowski SA, Fries CC, Karlson KE, et al: Primary determinant of ultimate fate of synthetic vascular grafts. Surgery 50:91-96, 1961. 14. Wesolowski SA, Fries CC, Martinez A, et al: Arterial prosthetic materials. Ann NY Acad Sci 146:325-344, 1968. 15. Shirkey AL, Beall AC Jr, DeBakey ME: The problem of small vessel grafting and the flexion crease: A comparison betweeen autogenous vein and Dacron grafts. Am J Surg 106:558-565, 1963. 16. Sanger PW, Taylor FH, McCall RE, et al: Seamless synthetic arterial grafts. JAMA 160:1403-1404, 1956. 17. Edwards WS, Tapp JS: Chemically treated nylon tubes as arterial grafts. Surgery 38:61-70, 1955. 18. Phillips CE Jr, DeWeese JA, Campeti FL: Comparison of peripheral artery grafts. Arch Surg 82:38-48, 1961. 19. Clark RE, Apostolou S, Kardos JL; Mismatch of mechanical properties as a cause of arterial prostheses thrombosis. Surg Forum 27:208-210, 1976. 20. Dale WA, Lewis MR: Modified bovine heterografts for arterial
replacement. Ann Surg 169:927-946, 1969. 21. Weymon AK, Plume SK, DeWeese JA: Bovine heterografts and autogenous veins as canine arterial bypass grafts. Arch Surg 110:746-750, 1975. 22. Clark
RE, in discussion, Campbell CD, Goldfarb D, Detton DD, et al: Expanded polytetrafluoro-ethylene as a small artery substitute. Trans Am Soc Artif Intern Organs 20:86-90, 1974. 23. Hopeman AR, in discussion, Gazzaniga AB, Lamberti JJ, Siewers RD, et al: Arterial prosthesis of microporous expanded polytetrafluoroethylene for construction of aorta-pulmonary shunts. J Thorac Cardiovasc Surg 72:357-362, 1976. 24. Smeloff EA, in discussion, Gazzaniga AB, Lamberti JJ, Siewers RD,
et al: Arterial prosthesis of microporous expanded polytetrafluoroethylene for construction of aorta-pulmonary shunts. J Thorac Cardiovasc Surg
72:357-362, 1976. 25. Hallin RW, Sweetman WR: The Sparks' mandril graft: A
seven
year
follow-up of mandril grafts placed by Charles H. Sparks and his associates. Am J Surg 132:221-223, 1976. 26. Ottinger LW, Darling C, Wirthlin LS, et al: Failure of ultralight\x=req-\ weight knitted Dacron grafts in arterial reconstruction. Arch Surg 111:146\x=req-\ 149, 1976.
27. Campbell CD, Brooks DH, Webster MW, et al: The use of expanded microporous polytetrafluoroethylene for limb salvage: A preliminary report. Surgery 79:485-493, 1976. 28. Roberts AK, Johnson N: Aneurysm formation in an expanded microporous polytetrafluoroethylene graft. Arch Surg 113:211-213, 1978.
Editorial Comment This report on the experimental results of a particular vascular substitute has a distinctively different tone than reports by other investigators. There are a large number of variables in experi¬ ments that could be responsible for this difference, including the characteristics of the prosthesis (pore size and fiber length), the anatomical location and length of the substitution, and the method of follow-up and evaluation. In order to come to firm conclusions as to the importance of any one of these variables, it is necessary to control the others. It is time for standard laboratory preparations and techniques to be established so that comparisons can be made between vascular substitutes and the important determinants of success determined. The vascular surgical groups would do well to
experimental evaluation of prostheses. A substitute could then be measured against the standard and there would be more meaningful evaluation. Until such time as standards are developed allowing for clear identification that the efficacy of a particular substitute is equal to that of autogenous tissue, clinical use of small-vessel substitutes should be reserved to limb- or life-threatening circumstances where no suitable autogenous material is available. This is partic¬ ularly true in the use of small-vessel substitutes for coronary replacement where failure of a graft can be a serious event. set standards for the
Downloaded From: http://archsurg.jamanetwork.com/ by a University of California - San Diego User on 06/04/2015
V. L. WlLLMAN, MD St Louis
microporous polytetrafluoroethylene (PTFE) has and clinically for small-vessel bypass the results vary widely. This study or assessed short (4-cm) femoral and carotid artery replacement PTFE grafts and longer (12-cm) femoral artery bypass PTFE grafts that crossed a flexion crease. Pore size at the blood-graft interface ranged from 10 to 30 \g=m\. At the end of three months, overall patency rate was 22%. No long bypass grafts remained patent. The pore size and type of anastomosis did not affect patency. The occlusion was always thrombotic, associated with fibrin at the suture line. Nine of 20 short carotid grafts (45%) and five of 22 short femoral grafts (22%) remained patent. These poor results indicate that further experimental studies are needed before PTFE is used clinically when an alternative exists. (Arch Surg 113:863-866, 1978) \s=b\ Expanded
been used
experimentally replacement grafting;
of
acceptance synthetic Thereplacement began 1952,
material for arterial when Voorhees and asso¬ ciates1 reported using Vinyon-"N" for successful bridging of a large-artery defect. Comparative experimental stud¬ ies2 and clinical experience have led to the successful current use of Dacron and Teflon for this purpose, both substances being strong and durable. The autogenous saphenous vein is still the graft of choice for replacement of medium-size vessels when available. In a prosthesis to replace recent years, the search for medium-size arteries (5- to 8-mm internal diameter) has spurred the development of numerous composite grafts of in
Accepted
for publication Feb 10, 1978. From the Department of Surgery, Bowman Gray School of Medicine, Wake Forest University, Winston-Salem, NC. Reprint requests to Department of Surgery, Bowman Gray School of Medicine, Winston-Salem, NC 27103 (Dr Walley).
synthetic material'·4 when venous autografts Purely biological grafts of heterologous collagen and homologous umbilical vein reinforced with a polyester mesh are also being used for medium-size artery replacements. Early results with many of these new grafts have been promising, but none of them have gained widespread acceptance. Replacement of small arteries (5-mm or less internal diameter) has met with even more limited success. Harri¬ collagen are
and
not available. '
in 1958 and Jacobson and associates" in 1963 concluded that only autogenous tissue (artery or vein) should be used for replacement of such small arteries, since none of the synthetic prostheses studied were satisfactory. Over the past five years, however, numerous investiga¬ tors have reported experimental success using expanded microporous polytetrafluoroethylene (PTFE) for smallvessel (3 to 4 mm) replacement in dogs.71" Expanded microporous PTFE (Teflon) is composed of nodes of PTFE interconnected by fine fibrils of PTFE forming a lattice. Discrepancy in patency rates (53% to 100%) has been ascribed to variations in the length of the fibrils between the nodes—or the "pore size"—which affect tissue ingrowth. Investigators disagree on what is the optimum "pore size," Campbell and associates8 having a patency rate of 96% with small-pore (22-µ) and 53% with larger-pore grafts, Florian and associates" and Matsumoto and asso¬ ciates'" having a patency rate of 100% with the more porous, longer-fibril (100-µ) grafts. Manufacture of the material is such now that pore size at the blood-graft interface can be smaller than the pore size on the outer surface of the graft. This prospective double-blind study was undertaken to compare patency rates of PTFE grafts of different lengths son2
Downloaded From: http://archsurg.jamanetwork.com/ by a University of California - San Diego User on 06/04/2015
and with different pore sizes. Both arterial and bypass models were studied.
replacement
MATERIALS AND METHODS
Thirty-nine healthy mongrel dogs of both sexes, averaging 22 kg in weight, were used in the study. Animals were anesthetized with 30 mg/kg of sodium pentobarbital and ventilated with a Harvard respirator through an endotracheal tube. Penicillin potas¬ sium, 500,000 units, was given just before and again just after the operative procedure. Normal saline, 50 mg/kg, was given intrave¬ nously during the operation. The abdomen and both femoral areas or the anterior cervical region was shaved as indicated for the operation to be done. The operative field was prepared with povidone-iodine and procedures were done under aseptic conditions. After adequate exposure of the arteries to be operated on, heparin sodium, 2 mg/kg, was administered intravenously. Ten minutes later vascular clamps were applied. The arterial anastomoses were performed with a continuous suture of 6-0 polyethylene. All anastomoses were done by experienced surgeons using 2 magnification loops and micro¬ surgery instruments. The prosthetic grafts utilized were expanded PTFE (Gore-Tex) grafts with an inside diameter of 4 mm and a wall thickness of 0.3 to 0.7 mm. Half of the grafts had an average fibril length of 10 µ at the blood-graft interface; the other half had a fibril length averaging 30 µ at the blood-graft interface. A graft from one group was randomly selected and placed on the left or right side of each animal, with a graft from the other group being placed in the contralateral side. The specifications of the graft were not known to the surgeons until the study had been completed.
Group
1
In the five control dogs, a portion of splenic vein approximately 10 to 12 cm in length and approximately 4 mm in inside diameter was harvested and sewn end-to-side to the intraabdominal portion of the right femoral artery. A tunnel was made under the inguinal ligament, and the graft was brought through it and anastomosed end-to-side to the femoral artery in the upper leg. The femoral artery was then ligated 1 cm below the upper anastomosis and 1 cm above the lower anastomosis.
Group
2
In ten dogs, 12-cm-long PTFE grafts were beveled at 45° angles on both ends and sewn end-to-side to the femoral artery in the manner described for group 1. Ten grafts were sewn end-to-end to
the femoral artery, again being brought through an inguinal tunnel from the intraabdominal portion to the upper leg portion of the artery.
Group
3
Both femoral (12 dogs) or both carotid (12 dogs) arteries were isolated, a 3- to 4-cm portion was excised, and 4-cm-long PTFE grafts were sewn end-to-end into the defect.
Completion
of
Operation
and
Follow-up
When the anastomoses were accomplished and the vascular clamps removed, the effect of heparin was reversed with 2 mg/kg of protamine sulfate. The wounds were irrigated with normal saline and closed in layers with polyglycolic acid (Dexon). Postoperatively, the dogs were given chloramphenicol, 7 mg/kg, for three days. With the animals under sedation, patency of the grafts was assessed with the Doppler ultrasound flowmeter 48 hours postoperatively, weekly thereafter for three weeks, and again two months postoperatively. If no signal was audible, the
animal was anesthetized with pentobarbital sodium and the occlusion verified by transection of the vessel distal to the anastomosis. Animals with audible Doppler signals were followed up for three months, with patency or occlusion being verified at death.
Pathology All grafts were harvested and preserved for microscopic evalu¬ ation. Grafts were stained with hematoxylin-eosin and examined by a pathologist unaware of the specifications of the grafts used.
RESULTS The results in three dogs are excluded: two of the dogs had infected groin wounds and the third died of intussus¬ ception before the end of the study. Of the remaining dogs, the patency rate overall for the PTFE grafts was 22%.
Group
1
The five autogenous venous grafts remained patent (100%) for the duration of the three-month study. At the animals' deaths, the anastomoses were widely patent with¬ out evidence of stenosis. There was moderate dilation of the vein graft in all cases. Microscopically, there was minimal intimai thickening and fibrosis.
Group
2
None of the 20 synthetic expanded PTFE grafts placed in the iliofemoral area remained patent for three months. Two occluded 48 hours after implantation; two grafts maintained flow for two months but were subsequently found to be occluded at death (three months); most occluded between two days and one month postimplanta¬ tion. Neither the fibril length at the flow surface nor the manner in which the graft was sewn (end-to-side vs end-toend) appeared to affect the patency rate. Thrombosis was associated with an occluding, circumferential deposit of fibrin at the anastomoses. Microscopically, the fibrous sheath was seen to be loosely adherent to the graft with minimal ingrowth of fibrous tissue.
Group
3
Nine of the 20 grafts placed in the carotid
area
remained
patent for three months (45%). Two of these grafts
were
approximately
50% occluded with thrombosis. Doppler evaluation of these short carotid grafts was technically difficult, so we cannot determine how long the remaining 11 grafts remained patent before thrombosis occurred. Five of the 22 short grafts placed in the femoral area remained patent for three months (22%). Of the grafts that occluded, eight did so between two and three months after
implantation.
Pathology Gross examination of the occluded grafts showed throm¬ bus that appeared to originate at the suture lines. The grafts that remained patent had a smooth, glistening flow surface with some stenosis at the suture lines. Microscopi¬ cally, the patent grafts had a fairly smooth and cellular
Downloaded From: http://archsurg.jamanetwork.com/ by a University of California - San Diego User on 06/04/2015
flow surface that appeared to originate from through the interstices of the grafts.
growth
COMMENT
Over the past 20 years, there has been extensive
screen¬
ing and development of materials for vascular reconstruc¬ tion. Pioneers such as Wesolowski, Sawyer, Sauvage, and colleagues11" have contributed greatly to our knowledge of how synthetic materials are incorporated into the arterial tree; DeBakey,"1' Sanger,1" Edwards,17 Voorhees and colleagues,1 to our knowledge of techniques and of the complications of graft placement in various parts of the body. Within 24 hours of implantation of a graft, fibrin and blood elements coat both surfaces of the graft. Over the next few weeks, the outer fibrin coating becomes replaced by an encapsulating sheath of granulation tissue; the inner fibrin layer must be converted to fibrous tissue by ingrowth of granulation tissue through the interstices of the graft.11 Experiments have shown the inner capsule to be complete at one month in swine and at four months in dogs." The inner capsule may never be complete in
humans." When this inner, fibrous-tissue "scar" contracts, its capillary blood supply in the wall of the graft may be disrupted, and this, in turn, may lead to sloughing of a portion of the inner lining, creating a more thrombogenic fibrin-flow surface. If the graft wall is porous, ingrowth of capillaries will allow rehealing in these areas." Thus, porosity is a primary determinant of the continued integrity of the flow surface of the prosthesis. However, the graft must not be so porous that it allows hemorrhage through the interstices. Other specifications for an ideal synthetic graft as outlined by Wesolowski et al1-' are no allergenic potential, durability for prolonged periods, and desirable handling properties. An ideal vascular substitute should also not kink across flexion sites, resist infection, and possess a permanent nonthrombogenic surface at the blood-graft interface.1 Phillips and co-workers18 have reported that synthetic grafts in small vessels probably fail because of thrombosis at the suture line. Our results support this finding. Physical stress at the suture line appears to be an important factor leading to that thrombosis.'9 Bypassing or replacing the femoral artery in dogs has proved a difficult test for prosthetic grafts. Although patency rates of 100%" and 93%" have been reported with autogenous veins used for this purpose, only 36% patency was achieved with knitted Dacron, 25% with woven Teflon, and 60% with knitted Teflon." Collagen tubes (bovine heterografts) had a patency rate of 18%-" to 50%21 when used to bypass the femoral artery of dogs. Since autoge¬ nous veins are not always available, the recent reports of high patency rates with grafts of expanded PTFE have excited interest in this material for small-vessel grafting. However, we found that we could not duplicate those high patency rates, having an overall patency rate after carotid artery replacement and femoral artery bypass or replace¬ ment of only 22%. Similarly poor results have been mentioned by others.--"1 Our grafts had either a highdensity flow surface of 10-µ fibril length or a flow surface
of 30-µ fibril length, both considered small-pore surfaces when considered against grafts with 100-µ fibril lengths. Our results were equally poor with both pore sizes, and suture line thrombosis occurred regardless of the extent of graft healing. Our poor results led us to compare our methods with those of investigators reporting patency rates as high as 96%8 and 100%."·"' Unlike the previous models, 20 of our grafts were implanted so that they passed beneath the inguinal ligament, thus crossing a joint (0% patency with PTFE graft; 100% patency with autogenous vein graft). Crossing a flexion crease has been shown to adversely affect other synthetic grafts of small arteries.'1 The grafts we used were also longer than those in previous studies. Our follow-up of all 62 grafts was for 90 days or until occlusion. A previous study by Campbell et al7 had an average follow-up of 18.5 days; animals were killed at regular intervals and only eight grafts were left to be followed up for longer than four months. Palpation was used to determine patency in other studies; in our study, Doppler flow measurements were used because palpation was found to be unreliable. We do not believe our suture technique was at fault, since all of the grafts were placed by surgeons who had participated in the experimental anastomosis of more than 100 autologous arteries and veins with only one confirmed technical error. Failure of a prosthetic graft for arterial replacement in humans may be followed by devastating consequences. Experiences with Sparks' mandril graft21 and some of the ultralightweight Dacron grafts2" are tragic examples of this fact. Expanded PTFE is presently being used clinically to replace and bypass medium-sized arteries, in some instances to cross a flexion crease. Aneurysms of such grafts have already been seen."·28 We are concerned about this clinical use when the experimental results are so varied and the factors causing graft failure are still unknown. This
investigation
Surgery.
Grafts
were
was
supported
provided by
in part
by
the Bradshaw
W. L. Gore and Associates,
Fellowship in
Flagstaff,
Ariz.
References 1. Voorhees AB Jr, Jaretzki A III, Blakemore AH: The use of tubes constructed from Vinyon-"N" cloth in bridging arterial defects: A preliminary report. Ann Surg 135:332-336, 1952. 2. Harrison JH: Synthetic materials as vascular prostheses: I. A comparative study in small vessels of nylon, Dacron, Orlon, Ivalon sponge, and Teflon. Am J Surg 95:3-15, 1958. 3. Parsonnet V, Alpert J, Brief DK: Autogenous polypropylene-supported collagen tubes for long-term arterial replacement. Surgery 70:935-939, 1971. 4. Sparks CH: Silicone mandril method of femoropopliteal artery bypass: Clinical experience and surgical technics. Am J Surg 124:244-249, 1972. 5. Dardik H, Ibrahim IM, Sprayregen S, et al: Clinical experience with modified human umbilical cord vein for arterial bypass. Surgery 79:618-624, 1976. 6. Jacobson JH II, Rosario ES, Katsumura T: Influence of prosthesis diameter in small artery replacement. Circulation 28:742-743, 1963. 7. Campbell CD, Goldfarb D, Detton DD, et al: Expanded polytetrafluoroethylene as a small artery substitute. Trans Am Soc Artif Intern Organs 20:86-90, 1974. 8. Campbell CD, Goldfarb D, Roe R: A small arterial substitute:
Expanded microporous polytetrafluoroethylene: Patency Surg 182:138-143, 1975.
Ann
Downloaded From: http://archsurg.jamanetwork.com/ by a University of California - San Diego User on 06/04/2015
versus
porosity.
9. Florian A, Cohn LH, Dammin GJ, et al: Small vessel replacement with Gore-Tex (expanded polytetrafluoroethylene). Arch Surg 111:267-270, 1976. 10. Matsumoto H, Hasegawa T, Fuse K, et al: A new vascular prosthesis for a small caliber artery. Surgery 74:519-523, 1973. 11. Berger K, Sauvage LR, Rao AM, et al: Healing of arterial prostheses in man: Its incompleteness. Am Surg 175:118-127, 1972. 12. Wesolowski SA, Fries CC, Domingo RT, et al: The compound prosthetic vascular graft: A pathologic survey. Surgery 53:19-44, 1963. 13. Wesolowski SA, Fries CC, Karlson KE, et al: Primary determinant of ultimate fate of synthetic vascular grafts. Surgery 50:91-96, 1961. 14. Wesolowski SA, Fries CC, Martinez A, et al: Arterial prosthetic materials. Ann NY Acad Sci 146:325-344, 1968. 15. Shirkey AL, Beall AC Jr, DeBakey ME: The problem of small vessel grafting and the flexion crease: A comparison betweeen autogenous vein and Dacron grafts. Am J Surg 106:558-565, 1963. 16. Sanger PW, Taylor FH, McCall RE, et al: Seamless synthetic arterial grafts. JAMA 160:1403-1404, 1956. 17. Edwards WS, Tapp JS: Chemically treated nylon tubes as arterial grafts. Surgery 38:61-70, 1955. 18. Phillips CE Jr, DeWeese JA, Campeti FL: Comparison of peripheral artery grafts. Arch Surg 82:38-48, 1961. 19. Clark RE, Apostolou S, Kardos JL; Mismatch of mechanical properties as a cause of arterial prostheses thrombosis. Surg Forum 27:208-210, 1976. 20. Dale WA, Lewis MR: Modified bovine heterografts for arterial
replacement. Ann Surg 169:927-946, 1969. 21. Weymon AK, Plume SK, DeWeese JA: Bovine heterografts and autogenous veins as canine arterial bypass grafts. Arch Surg 110:746-750, 1975. 22. Clark
RE, in discussion, Campbell CD, Goldfarb D, Detton DD, et al: Expanded polytetrafluoro-ethylene as a small artery substitute. Trans Am Soc Artif Intern Organs 20:86-90, 1974. 23. Hopeman AR, in discussion, Gazzaniga AB, Lamberti JJ, Siewers RD, et al: Arterial prosthesis of microporous expanded polytetrafluoroethylene for construction of aorta-pulmonary shunts. J Thorac Cardiovasc Surg 72:357-362, 1976. 24. Smeloff EA, in discussion, Gazzaniga AB, Lamberti JJ, Siewers RD,
et al: Arterial prosthesis of microporous expanded polytetrafluoroethylene for construction of aorta-pulmonary shunts. J Thorac Cardiovasc Surg
72:357-362, 1976. 25. Hallin RW, Sweetman WR: The Sparks' mandril graft: A
seven
year
follow-up of mandril grafts placed by Charles H. Sparks and his associates. Am J Surg 132:221-223, 1976. 26. Ottinger LW, Darling C, Wirthlin LS, et al: Failure of ultralight\x=req-\ weight knitted Dacron grafts in arterial reconstruction. Arch Surg 111:146\x=req-\ 149, 1976.
27. Campbell CD, Brooks DH, Webster MW, et al: The use of expanded microporous polytetrafluoroethylene for limb salvage: A preliminary report. Surgery 79:485-493, 1976. 28. Roberts AK, Johnson N: Aneurysm formation in an expanded microporous polytetrafluoroethylene graft. Arch Surg 113:211-213, 1978.
Editorial Comment This report on the experimental results of a particular vascular substitute has a distinctively different tone than reports by other investigators. There are a large number of variables in experi¬ ments that could be responsible for this difference, including the characteristics of the prosthesis (pore size and fiber length), the anatomical location and length of the substitution, and the method of follow-up and evaluation. In order to come to firm conclusions as to the importance of any one of these variables, it is necessary to control the others. It is time for standard laboratory preparations and techniques to be established so that comparisons can be made between vascular substitutes and the important determinants of success determined. The vascular surgical groups would do well to
experimental evaluation of prostheses. A substitute could then be measured against the standard and there would be more meaningful evaluation. Until such time as standards are developed allowing for clear identification that the efficacy of a particular substitute is equal to that of autogenous tissue, clinical use of small-vessel substitutes should be reserved to limb- or life-threatening circumstances where no suitable autogenous material is available. This is partic¬ ularly true in the use of small-vessel substitutes for coronary replacement where failure of a graft can be a serious event. set standards for the
Downloaded From: http://archsurg.jamanetwork.com/ by a University of California - San Diego User on 06/04/2015
V. L. WlLLMAN, MD St Louis
