- Email: [email protected]
Ten year experience with the negatively charged glutaraldehyde-tanned vascular graft in peripheral vascular surgery
Ten Year Experience With the Negatively Charged Glutaraldehyde-Tanned Vascular Graft in Peripheral Vascular Surgery Initial Multicenter Trail
Philip N. Sawyer, MD, Brooklyn, New York, Joseph Fitzgerald,MD, Brooklyn, New York, Martin J. Kaptllt,MD, Manhassett,New York, Rkhard J. Sanders,MD, Denver,Colorado,GeorgeY. Wllllams,MD, Baltimore, Maryland, Robert P. Leather, MD, AllstalrKarmody,MD, Albany, New York’, RogerW. Hallln, MD, Portland, Oregon, Robert Taylor,MD, Surrey, England, and CharlesC. Fries, MD, Johnson City, New York
The first negatively charged glutaraldehyde-tanned vascular graft (NCGP, Interface Biomedical Laboratories, Brooklyn, NY) was implanted in a human subject in 1975. It was the result of an extended evaluation of grafts of collagenous origin using modified calf carotid arteries. Research began between 1968 and 1970 [I]. The grafts were made by digesting muscle and elastin out of bovine arteries [2]. They were then treated in order to negatively charge the amino acid side chains on the collagen triple helix in the mature collagen. In contrast to the Rosenberg bovine heterograft (Johnson and Johnson, Brooklyn, NY), they were glutaraldehyde tanned rather than dialdehyde starch tanned, yielding a graft with a bursting strength of 3,000 to 5,000 g [3]. The grafts were initially tested under a number of conditions in experimental animals, principally the dog, and evaluated against other grafts [4] (Figure 1). The initial study of the grafts in human subjects was carried out as a multicenter trial in the hope of evaluating enough of the grafts to get a statistical look at their performance [5]. The grafts were evaluated in an environment in which they were implanted when the autogenous saphenous vein had been used and failed, was unavailable, or was initially unsatisfactory. From the Vascular Surglcai Services, Departments of Surgery and Surgical Research, Downstate Medical Center and the Methodist Hospital, Brooklyn, New York. Supported in part by grants-in-aid from The Sawyer Foundation and Interface Biomedical Laboratories Corporation, Brooklyn, New York. ‘Deceased. Requests for reprints should be addressed to Philip N. Sawyer, MD, 7600 Ridge Boulevard. Brooklyn, New York 11209. Resented at the 15th Annual Meeting of the Society for Clinical Vascular Surgery. Scottsdale, Arizona, March 25-29, 1967.
Volume 154, November 1987
Material and Methods As already mentioned, the grafts were made by digesting out the muscle and elastin from the bovine arteries (carotid, brachial, deep thoracic iliac and femoral, and other arteries). They were then cross-linked. The positively charged amino acid side chains of the triple helix of the collagen were carboxylated to negatively charge and then glutaraldehyde tanned to give them physical strength [I]. The original bovine heterografts were dialdehyde starch tanned for two reasons: dialdehyde starch tanning increased the tensile strength of the graft, and clean glutaraldehyde was not available to allow glutaraldehyde tanning of the grafts from 1955 to 1960 [S]. Formalin tanning had already been discarded as formalin was found to desorb rapidly when the graft was implanted in experimental animals, which yielded a weak aneurysmal graft within a matter of days. We now know that dialdehyde starch also desorbs over 12 to 36 months in human subjects, producing loss of collagen cross-linking with the passage of time. Ultimately, these grafts also become aneurysmal [7]. As with human homografts, with the passage of time, the number of aneurysms increases, approaching 100 percent 5 to 7 years after placement [8]. The Dardick-Meadoxa umbilical vein grafts present a different problem. Although glutaraldehyde tanned, they consist of juvenile tropocollagen. According to a number of investigators, the umbilical vein graft, surrounded by an outer Dacron net, disappears histologically by 24 months after placement [9,10]. The interstices of the Dacron net are filled with a thin layer of host fibroblasts. Aneurysm formation then tends to occur, particularly in hypertensive patients. The modern negatively charged ghrtaraidehyde-
533
Sawyer et al
Ffgufe 1. Resuffs of fmplanfafion of collagen prostheses in canine blood vessels hiulfipie modiflcafbns of fhe dfgesfed bovhe caroffd artery were carried auf to determine the elfecfs of the chemical frantdormafion on the tendency lo both fhrombose and defer/orate affer lmplanfafbn In the carofid arfery andaorta In the dog. These munlple mcdlfkzaflons are IdenfMledln column one. Also shown lo the rate of fhrombosls among the various lmpianfed grafftx
TABLE
I
Comparisonof the St. Jude Blopolymerlc Graft and the Dardlck-Meadox Blografl
Criteria Radial tensile strength (lb f force) Suture retention strength (lb f force) Wall thickness (mm) Anastomosed longitudinal tensile strength (lb f force) Zeta potential Collagen content (percentage of dry weight)
St. Jude Biopolymeric Graft 6.57 f
Dardick Meadox Umbilical Vein Bioaraft
1.65
2.31 f 0.76
2.64 f 0.57
0.76 f 0.17
1.08 f 0.007
1.33 f 0.25
16.7 f 4.4
7.2 f 2.1
9.89 f 1.33 56.59 f 6.60
... . .
tanned vascular graft, now called the St. Jude Medical Biopolymerica graft, is made of mature glutaraldehydetanned bovine collagen. By 1970, clean glutaraldehyde became available because of a need in both electron microscopy and other sophisticated histologic applications. The initial grafts were manufactured by Interface Bio-
534
medical Laboratories from 1971 to 1984. They were first implanted in the canine aorta, carotid arteries, and iliac arteries from 1968 to 1971 and followed up to 10 years. The grafts were compared with grafts in which other means of digesting and tanning were employed (Table I) [1,2,41. After 1 month to 5 years of follow-up in experimental animals, the grafts were first implanted in human subjects in 1975 under an Investigational Device Exemption with the Food and Drug Administration. These initial negatively charged glutaraldehyde-tanned vascular grafts, made in lengths of from 10 to 50 cm, were carefully evaluated. When long lengths were needed, two grafts would occasionally be sutured together in the operating room in order to obtain sufficient length. A multicenter trial was eventually carried out [5,11]. With experience came improvements in graft length, consistency, and quality. Each of the early batches of grafts manufactured from 1965 to 1978 was initially tested in dogs. In several early experimental preparations, the grafts were discarded as the processing chemicals did not produce the desired result. However, the experience was extremely informative and made possible the initial human trial in 1975. It should be noted that many of the grafts discussed in the present study were implanted to replace failed autogenous saphenous veins (Table II) [4].
The American Journal of Surgery
Negatively Charged Glutaraldehyde-Tanned Vascular Graft
TABLE II
Succemful Revmcularlzatlon blng Nogatlvely Charged Giutarakkhyde-Tannod Copolymerlc Collagen Grafl (St. Jude Blopolymwlc Graft) After Occlusion of Orlglnal Vascular Reconstruction
Date of Second Operation 1012175
Original Reconstruction
Duration Patency
10/23/75 1212175
Right limb aortofemoral would not stay open; 3 thrombectomies Femoropopliteal endarterectomy Right femoropopliteal saphenous vein bypass
204 10 mo
12/11/75
Left femoropopliteal saphenous vein bypass
10 mo
213176
Aortoiliacfemoral thrombectomy and embolectomy with endarterectomy Aortofemoral bypass
22d
214176 2110176
7127176 7129176
l
Long femoropopliteal Rosenberg bovine heterograft with multiple reconstructions, thrombectomies, 8 aneurysm formation Bilateral femoropopliteal endarterectomy with venous patch graft Bilateral femoropopllteal endarterectomy with venous patch graft
4d
Duration of FlNlction (mo)’
Secondary Reconstruction Right limb of graft to popliteal bovine glutaraldehyde heterograft Negative bovine femoropopliteal heterograft Replaced with negative bovine glutaraldehyde heterograft Replaced with negative bovine glutaraldehyde heterograft Right femoropopliteal negatively charged bovine heterograft. Amputated left leg Femorofemoral negatively charged bovine glutaraldehyde heterograft Replaced with negative glutaraldehyde-tanned bovine heterograft
24 mo 24 mo
12 12 16 16 12 6 10
5mo
Long femoropopliteal bovine heterograft bypass
1
3mo
Long femoropopliteal bovine heterograft bypass
1
Patency confirmed by Doppler analysis or angiography.
An analysis of results from many years of effort to apply gas endarterectomy, autogenous saphenous vein, and the Rosenberg bovine heterograft, yielded a comparative trial of the first 31 implants of the negatively charged glutaraldehyde-tanned vascular grafts, during a comparative, nonrandomized trial. The results from this early comparative study are presented in Figure 2 and were first published in 1976. Ultimately, the grafts were made in increasingly smaller sizes to determine the minimum diameter that would remain patent. These too were first implanted in the carotid and iliofemoral arteries of dogs. The small-diameter grafts were initially used in distal femoral, posterior tibial, and carotid axillary arteries in human subjects. These small grafts are now available in naturally tapering 5 to 4 mm, 6 to 4 mm, 5 mm, and 4 mm isodiametric diameters. By 1986, 146 grafts were implanted in 108 patients in this multicenter trial. Thirty-three percent were women and 67 were men. Patient ages ranged from 36 to 89 years, with an average age of 64 years. Graft follow-up ranged from 1 day to 9*/z years, for an average follow-up of 2 years. The combined length of follow-up was 3,504 months. Indications for reconstruction were foot salvage for chronic limb-threatening disease and limb-threatening vascular insufficiency in 80 percent of the patients. Claudication at rest or severe claudication on walking very short distances were the indications in the other 20 percent. Many of these patients were hypertensive, had a previous myocardial infarction, were diabetic or had a combination of these conditions. Operative runoff was limited in many of the patients, averaging from one to
Volume
154, November
1987
> y w
60
ii
60
2
4
6
6
IO
12
14
2yr
3yr
4yr
TIME (MO) Figure 2. Patency of fetnoropoPlttea/ reconetructfone ushg lffe table analysk Patency In femoropopllteat reconstructIons was stat/st/ca//y Identkal for three modes of reconstruction during a follow-up perfod of up to 4 years The modes of reconstruct/on were old and new bovlne heterografts, autogenous &#tenous vein bypass, and gas endarterectomy. At 4 years, the old bovine heterografts had already started to develop aneurysms and d/splayed decreasing patency.
one-and-a-half vessels below the trifurcation. Seven percent of the patients were classified as having good operative runoff, 32 percent fair runoff, 42 percent poor runoff, and 18 percent undetermined. The implantation sites are illustrated in Table III. Twenty-nine percent were implanted above or at the knee joint and 62 percent below the knee joint. Sixty-nine percent of the grafts were large (6 mm or greater at the proximal and distal ends) and 28 percent were small (less than 6 mm at the proximal and distal end). Secondary patency rates, demonstrating duration of function, were calculated by the life table method using
535
Sawyer et al
TABLE III
Graft Locatlons for the St. Jude Blopolymerlc Graft n
Graft Location
1 6 1 2 4 5 2 18 85 3 2 9 6 1 1
Aortorenal Axillofemoral Axillofemoral-tibia1 Axillopopliteal Carotid-axillary Femorofemoral Femorofemoral-popliteal Femoroperoneal Femoropopliteal Femoropopliteal-peroneal Femoropopliteal-tibia1 Femorotibial lliopopliteal Iliopopliteal-tibia1 Innominate-brachial
146
Total
40
-‘,3otl3(3) C_+__+3ef13w
t -Good/Fair *--*Poor
’
I
I
I
I
I
I
I
I
Runoff Runoff
I
123456769 Years Postimpfant
Figure 4. Secondary patency by operat/ve runoff for the negatively charged glutaraldehyde-tanned vascular graft. The patency rate for all grafts was less at all t/me intervals If the runoff was poor (less than one-vessel runoff below the knee).
the Statistical Package for the Social Sciences, a computerized statistical package. Where appropriate, statistical comparisons were made using statistic D, calculated from actuarial scores using the algorithm of Lee and Desu [Z]. Grafts were followed from implant to the point at which function ceased due to death, amputation, explant, development of an occlusion that could not be thrombectomized, or to the point of most recent follow-up. Graft closure for which no extrinsic factors could be identified was considered a terminal event. Results Overall patency at 5 years was 67 percent for all grafts (Figure 3). Above-knee and below-knee grafts functioned equally well at 5 years (above-knee grafts 60 f 9 and below-knee grafts 172 f 7). In addition, there were no significant dilferences in function at 5 years between large-diameter and small-diameter grafts (large-diameter grafts 69 f 5 536
6054fod)
’
1
I I I 12345678
I
I
I
I
I
II 9 9.6
Years Post-Implant Figure 3. Secondary patency for the negatively charged glutaraldehyde-tanned vascular graft. Long-term patency for both large- and small-diameter grafts combined was approximately 67 percent at 6 years, w/th a gradual decrease to 54 percent at 9.6 years. The resolts Indicate 5 year patency among survivors of approximately 69 percent for large-diameter grafts and 59 percent for small-dlameter grafts Values In parentheses indicate number of vessels at risk.
and small-diameter grafts 59 f 10). Operative runoff produced the only significant difference (Figure 4). As expected, grafts with poor operative runoff did not function as well as those with good or fair runoff. Complications observed in this study included 59 grafts (40 percent) that had at least one graft-related occlusion, 12 incidents of infection (9 percent), 1 instance of calcification (0.7 percent), and 6 aneurysms (4 percent). Two of the aneurysms occurred in one extremely hypertensive patient. After 4 years of adequate function, the graft in the left leg required complete replacement, and the graft in the right leg required segmental replacement. Another two aneurysms occurred in two patients who previously had received massive chemotherapy. As terminus approached, an anastomotic infection with a pseudoaneurysm developed. On the other hand, two other patients in whom the negatively charged glutaraldehyde-tanned vascular graft had been implanted underwent massive streptokinese therapy without bleeding and without subsequent aneurysm formation 1 and 3 years after implantation. Each of these thrombolytic efforts resulted in long-term patency. All available patients with patent negatively charged glutaraldehyde-tanned vascular grafts have recently undergone duplex scanning. These nine patients with both large- and small-diameter grafts in place for periods of 2 to 7 years were found to be free of aneurysms*. In three of these patients, arteriograms confirmed these findings. ‘Specific data concerning these patients are available from the senior author (PNS).
The American Journal of Surgery
NegativelyChargedGlutaraldehyde-Tanned Vascular Graft
comments The principal objection to the use of biologic grafts appears to be aneurysm formation. Actually, careful scrutiny of the literature and our personal experience suggest that biologic grafts perform more satisfactorily over the long term, with respect to patency, than any of the polymeric grafts [13,14]. It is of interest that even the autogenous saphenous vein itself is capable of aneurysm formation under certain unsupported conditions, such as aortorenal bypass in the hypertensive patient [14,15]. All homologous arterial grafts, such as the Rosenberg bovine heterograft and the Meadox-Dardick umbilical vein, are reported to have functioned better than their synthetic equivalents, such as the Dacron cloth, polytetrafluoroethylene, and all of the small-diameter prostheses, until aneurysms occurred at 36 months and thereafter. The negatively charged glutaraldehyde-tanned vascular graft (now called the St. Jude Biopolymeric graft) seems to have significantly better resistance to aneurysm formation and long-term patency, even though it has been implanted in a limited number of patients. This graft is currently available as an arterial substitute, and its use for dialysis access, as well as in other applications, has been applied for. Its longterm utility in the hands of the vascular surgical community will only be determined when we know the results in large numbers of patients in a clinical trial. Summary The negatively charged glutaraldehyde-tanned vascular graft was developed to determine the utility of grafts of biologic origin to provide satisfactory revascularization for ischemic limbs with advanced arteriosclerotic peripheral vascular disease. The bovine carotid artery was modified under a number of conditions using a large number of variables and evaluated in the carotid artery and aorta of dogs. Eventually ficin-digested carboxylated glutaraldehyde-tanned grafts were selected as being most antithrombotic and most resistant to aneurysm formation. Under an Investigational Device Exemption with the Food and Drug Administration, 146 grafts were
Volume 154, November 1997
evaluated in 108 patients with patency approaching 67 percent for all grafts at 5 years and 59 percent at 9 years plus. Total aneurysm formation has been 4 percent (six grafts) occurring only in hypertensive patients. The graft is now undergoing clinical use and evaluation worldwide. References 1. Sawyer PN, O’Shaughnessy AM, Sophie Z. Development and 2.
3.
4.
5. 8.
7. 8. 9.
10.
11.
12.
13.
14.
15.
performance characteristics of a new vascular graft: J Biomed Mater Res 1985; 19: 991-1010. Sawyer PN, Stanczewski B, Sivakoff SJ, Lucas TR, Kirchenbaum D. Search for the ideal collagen vascular prosthesis. Trans Am Sot Artif Intern Organs 1977; 23: 288-92. Rosenberg N, Henderson J, Lord GH, Bothwell JW. An arterial prosthesis of heterologous vascular origin. JAMA 1984; 187: 74-85. Sawyer PN, Stanczewski B, Lucas TR, et al. In: Sawyer PN, Kapplitt MI, ed. Experimental and clinical evaluation of a new negatively charged bovine heterograft for use in peripheral and coronary revascularization vascular grafts. New York: Appleton-Century-Crofts, 1977: 282-397. Reddy K, Haque SN, Cohen L, et al. A clinical experience with the NCGT graft. J Biomed Mater Res 1981; 15: 335-41. Rosenberg N. Dialdehyde starch-tanned bovine heterograft. In: Stanley JC, ed. Biologic and synthetic vascular prostheses. New York: Grune and Stratton, 1982; 423-32. Cutler BS, Thompson JE, Patman DE. The modified bovine arterial graft: a clinical study. Surgery 1974; 36: 431-42. Szilagyi DE, McDonald RT, Smith RF. Biologic fate of human arterial homografts. Arch Surg 1957; 75:506-14. Baier RE. Phvsical chemistrv of blood. In: Stanlev JC. ed. Biologic and synthetic vascular prostheses. New York: Grune and Stratton, 1982: 83-99. Hasson E, Newton WD, Waltman AC, et al. Mural degeneration in the glutaraldehyde-tanned umbilical vein graft: incidence and implications. J Vast Surg 1986; 4: 243-50. Sawyer PN, Adamson R, Butt K, et al. Long-term function of NCGT vascular conduits in a multicenter trial: evaluation of physical chemical parameters. Biomater Med Devices Artlf Organs 1980; 8: 345-67. Lee E, Desu M. A computer program for comparing K samples with right-censored data. Comput Programs Blamed 1972; 2: 315-21. Sottiurdi VS, Batson RC. Autogenous vein graft: experimental studies. In: Stanley JC, ed. Biologic and synthetic vascular prostheses. New York: tine and Stratton, 1982. Stewart SFC, Lyman DJ. Essential physical characteristics of vascular grafts. In: Sawyer PN, ed. Modern vascular grafts. New York: McGraw Hill, 1986: 115-32. Stanley JC, Graham LM, Whitehall WM, Linderman SM. Zerolock GB, Cronenwett JL. Autogenous saphenous vein as an arterial graft: clinical status. In: Stanley JC, ed. Biologic and synthetic vascular prostheses. New York: Grune and Stratton, 1982: 333-49.
537
Philip N. Sawyer, MD, Brooklyn, New York, Joseph Fitzgerald,MD, Brooklyn, New York, Martin J. Kaptllt,MD, Manhassett,New York, Rkhard J. Sanders,MD, Denver,Colorado,GeorgeY. Wllllams,MD, Baltimore, Maryland, Robert P. Leather, MD, AllstalrKarmody,MD, Albany, New York’, RogerW. Hallln, MD, Portland, Oregon, Robert Taylor,MD, Surrey, England, and CharlesC. Fries, MD, Johnson City, New York
The first negatively charged glutaraldehyde-tanned vascular graft (NCGP, Interface Biomedical Laboratories, Brooklyn, NY) was implanted in a human subject in 1975. It was the result of an extended evaluation of grafts of collagenous origin using modified calf carotid arteries. Research began between 1968 and 1970 [I]. The grafts were made by digesting muscle and elastin out of bovine arteries [2]. They were then treated in order to negatively charge the amino acid side chains on the collagen triple helix in the mature collagen. In contrast to the Rosenberg bovine heterograft (Johnson and Johnson, Brooklyn, NY), they were glutaraldehyde tanned rather than dialdehyde starch tanned, yielding a graft with a bursting strength of 3,000 to 5,000 g [3]. The grafts were initially tested under a number of conditions in experimental animals, principally the dog, and evaluated against other grafts [4] (Figure 1). The initial study of the grafts in human subjects was carried out as a multicenter trial in the hope of evaluating enough of the grafts to get a statistical look at their performance [5]. The grafts were evaluated in an environment in which they were implanted when the autogenous saphenous vein had been used and failed, was unavailable, or was initially unsatisfactory. From the Vascular Surglcai Services, Departments of Surgery and Surgical Research, Downstate Medical Center and the Methodist Hospital, Brooklyn, New York. Supported in part by grants-in-aid from The Sawyer Foundation and Interface Biomedical Laboratories Corporation, Brooklyn, New York. ‘Deceased. Requests for reprints should be addressed to Philip N. Sawyer, MD, 7600 Ridge Boulevard. Brooklyn, New York 11209. Resented at the 15th Annual Meeting of the Society for Clinical Vascular Surgery. Scottsdale, Arizona, March 25-29, 1967.
Volume 154, November 1987
Material and Methods As already mentioned, the grafts were made by digesting out the muscle and elastin from the bovine arteries (carotid, brachial, deep thoracic iliac and femoral, and other arteries). They were then cross-linked. The positively charged amino acid side chains of the triple helix of the collagen were carboxylated to negatively charge and then glutaraldehyde tanned to give them physical strength [I]. The original bovine heterografts were dialdehyde starch tanned for two reasons: dialdehyde starch tanning increased the tensile strength of the graft, and clean glutaraldehyde was not available to allow glutaraldehyde tanning of the grafts from 1955 to 1960 [S]. Formalin tanning had already been discarded as formalin was found to desorb rapidly when the graft was implanted in experimental animals, which yielded a weak aneurysmal graft within a matter of days. We now know that dialdehyde starch also desorbs over 12 to 36 months in human subjects, producing loss of collagen cross-linking with the passage of time. Ultimately, these grafts also become aneurysmal [7]. As with human homografts, with the passage of time, the number of aneurysms increases, approaching 100 percent 5 to 7 years after placement [8]. The Dardick-Meadoxa umbilical vein grafts present a different problem. Although glutaraldehyde tanned, they consist of juvenile tropocollagen. According to a number of investigators, the umbilical vein graft, surrounded by an outer Dacron net, disappears histologically by 24 months after placement [9,10]. The interstices of the Dacron net are filled with a thin layer of host fibroblasts. Aneurysm formation then tends to occur, particularly in hypertensive patients. The modern negatively charged ghrtaraidehyde-
533
Sawyer et al
Ffgufe 1. Resuffs of fmplanfafion of collagen prostheses in canine blood vessels hiulfipie modiflcafbns of fhe dfgesfed bovhe caroffd artery were carried auf to determine the elfecfs of the chemical frantdormafion on the tendency lo both fhrombose and defer/orate affer lmplanfafbn In the carofid arfery andaorta In the dog. These munlple mcdlfkzaflons are IdenfMledln column one. Also shown lo the rate of fhrombosls among the various lmpianfed grafftx
TABLE
I
Comparisonof the St. Jude Blopolymerlc Graft and the Dardlck-Meadox Blografl
Criteria Radial tensile strength (lb f force) Suture retention strength (lb f force) Wall thickness (mm) Anastomosed longitudinal tensile strength (lb f force) Zeta potential Collagen content (percentage of dry weight)
St. Jude Biopolymeric Graft 6.57 f
Dardick Meadox Umbilical Vein Bioaraft
1.65
2.31 f 0.76
2.64 f 0.57
0.76 f 0.17
1.08 f 0.007
1.33 f 0.25
16.7 f 4.4
7.2 f 2.1
9.89 f 1.33 56.59 f 6.60
... . .
tanned vascular graft, now called the St. Jude Medical Biopolymerica graft, is made of mature glutaraldehydetanned bovine collagen. By 1970, clean glutaraldehyde became available because of a need in both electron microscopy and other sophisticated histologic applications. The initial grafts were manufactured by Interface Bio-
534
medical Laboratories from 1971 to 1984. They were first implanted in the canine aorta, carotid arteries, and iliac arteries from 1968 to 1971 and followed up to 10 years. The grafts were compared with grafts in which other means of digesting and tanning were employed (Table I) [1,2,41. After 1 month to 5 years of follow-up in experimental animals, the grafts were first implanted in human subjects in 1975 under an Investigational Device Exemption with the Food and Drug Administration. These initial negatively charged glutaraldehyde-tanned vascular grafts, made in lengths of from 10 to 50 cm, were carefully evaluated. When long lengths were needed, two grafts would occasionally be sutured together in the operating room in order to obtain sufficient length. A multicenter trial was eventually carried out [5,11]. With experience came improvements in graft length, consistency, and quality. Each of the early batches of grafts manufactured from 1965 to 1978 was initially tested in dogs. In several early experimental preparations, the grafts were discarded as the processing chemicals did not produce the desired result. However, the experience was extremely informative and made possible the initial human trial in 1975. It should be noted that many of the grafts discussed in the present study were implanted to replace failed autogenous saphenous veins (Table II) [4].
The American Journal of Surgery
Negatively Charged Glutaraldehyde-Tanned Vascular Graft
TABLE II
Succemful Revmcularlzatlon blng Nogatlvely Charged Giutarakkhyde-Tannod Copolymerlc Collagen Grafl (St. Jude Blopolymwlc Graft) After Occlusion of Orlglnal Vascular Reconstruction
Date of Second Operation 1012175
Original Reconstruction
Duration Patency
10/23/75 1212175
Right limb aortofemoral would not stay open; 3 thrombectomies Femoropopliteal endarterectomy Right femoropopliteal saphenous vein bypass
204 10 mo
12/11/75
Left femoropopliteal saphenous vein bypass
10 mo
213176
Aortoiliacfemoral thrombectomy and embolectomy with endarterectomy Aortofemoral bypass
22d
214176 2110176
7127176 7129176
l
Long femoropopliteal Rosenberg bovine heterograft with multiple reconstructions, thrombectomies, 8 aneurysm formation Bilateral femoropopliteal endarterectomy with venous patch graft Bilateral femoropopllteal endarterectomy with venous patch graft
4d
Duration of FlNlction (mo)’
Secondary Reconstruction Right limb of graft to popliteal bovine glutaraldehyde heterograft Negative bovine femoropopliteal heterograft Replaced with negative bovine glutaraldehyde heterograft Replaced with negative bovine glutaraldehyde heterograft Right femoropopliteal negatively charged bovine heterograft. Amputated left leg Femorofemoral negatively charged bovine glutaraldehyde heterograft Replaced with negative glutaraldehyde-tanned bovine heterograft
24 mo 24 mo
12 12 16 16 12 6 10
5mo
Long femoropopliteal bovine heterograft bypass
1
3mo
Long femoropopliteal bovine heterograft bypass
1
Patency confirmed by Doppler analysis or angiography.
An analysis of results from many years of effort to apply gas endarterectomy, autogenous saphenous vein, and the Rosenberg bovine heterograft, yielded a comparative trial of the first 31 implants of the negatively charged glutaraldehyde-tanned vascular grafts, during a comparative, nonrandomized trial. The results from this early comparative study are presented in Figure 2 and were first published in 1976. Ultimately, the grafts were made in increasingly smaller sizes to determine the minimum diameter that would remain patent. These too were first implanted in the carotid and iliofemoral arteries of dogs. The small-diameter grafts were initially used in distal femoral, posterior tibial, and carotid axillary arteries in human subjects. These small grafts are now available in naturally tapering 5 to 4 mm, 6 to 4 mm, 5 mm, and 4 mm isodiametric diameters. By 1986, 146 grafts were implanted in 108 patients in this multicenter trial. Thirty-three percent were women and 67 were men. Patient ages ranged from 36 to 89 years, with an average age of 64 years. Graft follow-up ranged from 1 day to 9*/z years, for an average follow-up of 2 years. The combined length of follow-up was 3,504 months. Indications for reconstruction were foot salvage for chronic limb-threatening disease and limb-threatening vascular insufficiency in 80 percent of the patients. Claudication at rest or severe claudication on walking very short distances were the indications in the other 20 percent. Many of these patients were hypertensive, had a previous myocardial infarction, were diabetic or had a combination of these conditions. Operative runoff was limited in many of the patients, averaging from one to
Volume
154, November
1987
> y w
60
ii
60
2
4
6
6
IO
12
14
2yr
3yr
4yr
TIME (MO) Figure 2. Patency of fetnoropoPlttea/ reconetructfone ushg lffe table analysk Patency In femoropopllteat reconstructIons was stat/st/ca//y Identkal for three modes of reconstruction during a follow-up perfod of up to 4 years The modes of reconstruct/on were old and new bovlne heterografts, autogenous &#tenous vein bypass, and gas endarterectomy. At 4 years, the old bovine heterografts had already started to develop aneurysms and d/splayed decreasing patency.
one-and-a-half vessels below the trifurcation. Seven percent of the patients were classified as having good operative runoff, 32 percent fair runoff, 42 percent poor runoff, and 18 percent undetermined. The implantation sites are illustrated in Table III. Twenty-nine percent were implanted above or at the knee joint and 62 percent below the knee joint. Sixty-nine percent of the grafts were large (6 mm or greater at the proximal and distal ends) and 28 percent were small (less than 6 mm at the proximal and distal end). Secondary patency rates, demonstrating duration of function, were calculated by the life table method using
535
Sawyer et al
TABLE III
Graft Locatlons for the St. Jude Blopolymerlc Graft n
Graft Location
1 6 1 2 4 5 2 18 85 3 2 9 6 1 1
Aortorenal Axillofemoral Axillofemoral-tibia1 Axillopopliteal Carotid-axillary Femorofemoral Femorofemoral-popliteal Femoroperoneal Femoropopliteal Femoropopliteal-peroneal Femoropopliteal-tibia1 Femorotibial lliopopliteal Iliopopliteal-tibia1 Innominate-brachial
146
Total
40
-‘,3otl3(3) C_+__+3ef13w
t -Good/Fair *--*Poor
’
I
I
I
I
I
I
I
I
Runoff Runoff
I
123456769 Years Postimpfant
Figure 4. Secondary patency by operat/ve runoff for the negatively charged glutaraldehyde-tanned vascular graft. The patency rate for all grafts was less at all t/me intervals If the runoff was poor (less than one-vessel runoff below the knee).
the Statistical Package for the Social Sciences, a computerized statistical package. Where appropriate, statistical comparisons were made using statistic D, calculated from actuarial scores using the algorithm of Lee and Desu [Z]. Grafts were followed from implant to the point at which function ceased due to death, amputation, explant, development of an occlusion that could not be thrombectomized, or to the point of most recent follow-up. Graft closure for which no extrinsic factors could be identified was considered a terminal event. Results Overall patency at 5 years was 67 percent for all grafts (Figure 3). Above-knee and below-knee grafts functioned equally well at 5 years (above-knee grafts 60 f 9 and below-knee grafts 172 f 7). In addition, there were no significant dilferences in function at 5 years between large-diameter and small-diameter grafts (large-diameter grafts 69 f 5 536
6054fod)
’
1
I I I 12345678
I
I
I
I
I
II 9 9.6
Years Post-Implant Figure 3. Secondary patency for the negatively charged glutaraldehyde-tanned vascular graft. Long-term patency for both large- and small-diameter grafts combined was approximately 67 percent at 6 years, w/th a gradual decrease to 54 percent at 9.6 years. The resolts Indicate 5 year patency among survivors of approximately 69 percent for large-diameter grafts and 59 percent for small-dlameter grafts Values In parentheses indicate number of vessels at risk.
and small-diameter grafts 59 f 10). Operative runoff produced the only significant difference (Figure 4). As expected, grafts with poor operative runoff did not function as well as those with good or fair runoff. Complications observed in this study included 59 grafts (40 percent) that had at least one graft-related occlusion, 12 incidents of infection (9 percent), 1 instance of calcification (0.7 percent), and 6 aneurysms (4 percent). Two of the aneurysms occurred in one extremely hypertensive patient. After 4 years of adequate function, the graft in the left leg required complete replacement, and the graft in the right leg required segmental replacement. Another two aneurysms occurred in two patients who previously had received massive chemotherapy. As terminus approached, an anastomotic infection with a pseudoaneurysm developed. On the other hand, two other patients in whom the negatively charged glutaraldehyde-tanned vascular graft had been implanted underwent massive streptokinese therapy without bleeding and without subsequent aneurysm formation 1 and 3 years after implantation. Each of these thrombolytic efforts resulted in long-term patency. All available patients with patent negatively charged glutaraldehyde-tanned vascular grafts have recently undergone duplex scanning. These nine patients with both large- and small-diameter grafts in place for periods of 2 to 7 years were found to be free of aneurysms*. In three of these patients, arteriograms confirmed these findings. ‘Specific data concerning these patients are available from the senior author (PNS).
The American Journal of Surgery
NegativelyChargedGlutaraldehyde-Tanned Vascular Graft
comments The principal objection to the use of biologic grafts appears to be aneurysm formation. Actually, careful scrutiny of the literature and our personal experience suggest that biologic grafts perform more satisfactorily over the long term, with respect to patency, than any of the polymeric grafts [13,14]. It is of interest that even the autogenous saphenous vein itself is capable of aneurysm formation under certain unsupported conditions, such as aortorenal bypass in the hypertensive patient [14,15]. All homologous arterial grafts, such as the Rosenberg bovine heterograft and the Meadox-Dardick umbilical vein, are reported to have functioned better than their synthetic equivalents, such as the Dacron cloth, polytetrafluoroethylene, and all of the small-diameter prostheses, until aneurysms occurred at 36 months and thereafter. The negatively charged glutaraldehyde-tanned vascular graft (now called the St. Jude Biopolymeric graft) seems to have significantly better resistance to aneurysm formation and long-term patency, even though it has been implanted in a limited number of patients. This graft is currently available as an arterial substitute, and its use for dialysis access, as well as in other applications, has been applied for. Its longterm utility in the hands of the vascular surgical community will only be determined when we know the results in large numbers of patients in a clinical trial. Summary The negatively charged glutaraldehyde-tanned vascular graft was developed to determine the utility of grafts of biologic origin to provide satisfactory revascularization for ischemic limbs with advanced arteriosclerotic peripheral vascular disease. The bovine carotid artery was modified under a number of conditions using a large number of variables and evaluated in the carotid artery and aorta of dogs. Eventually ficin-digested carboxylated glutaraldehyde-tanned grafts were selected as being most antithrombotic and most resistant to aneurysm formation. Under an Investigational Device Exemption with the Food and Drug Administration, 146 grafts were
Volume 154, November 1997
evaluated in 108 patients with patency approaching 67 percent for all grafts at 5 years and 59 percent at 9 years plus. Total aneurysm formation has been 4 percent (six grafts) occurring only in hypertensive patients. The graft is now undergoing clinical use and evaluation worldwide. References 1. Sawyer PN, O’Shaughnessy AM, Sophie Z. Development and 2.
3.
4.
5. 8.
7. 8. 9.
10.
11.
12.
13.
14.
15.
performance characteristics of a new vascular graft: J Biomed Mater Res 1985; 19: 991-1010. Sawyer PN, Stanczewski B, Sivakoff SJ, Lucas TR, Kirchenbaum D. Search for the ideal collagen vascular prosthesis. Trans Am Sot Artif Intern Organs 1977; 23: 288-92. Rosenberg N, Henderson J, Lord GH, Bothwell JW. An arterial prosthesis of heterologous vascular origin. JAMA 1984; 187: 74-85. Sawyer PN, Stanczewski B, Lucas TR, et al. In: Sawyer PN, Kapplitt MI, ed. Experimental and clinical evaluation of a new negatively charged bovine heterograft for use in peripheral and coronary revascularization vascular grafts. New York: Appleton-Century-Crofts, 1977: 282-397. Reddy K, Haque SN, Cohen L, et al. A clinical experience with the NCGT graft. J Biomed Mater Res 1981; 15: 335-41. Rosenberg N. Dialdehyde starch-tanned bovine heterograft. In: Stanley JC, ed. Biologic and synthetic vascular prostheses. New York: Grune and Stratton, 1982; 423-32. Cutler BS, Thompson JE, Patman DE. The modified bovine arterial graft: a clinical study. Surgery 1974; 36: 431-42. Szilagyi DE, McDonald RT, Smith RF. Biologic fate of human arterial homografts. Arch Surg 1957; 75:506-14. Baier RE. Phvsical chemistrv of blood. In: Stanlev JC. ed. Biologic and synthetic vascular prostheses. New York: Grune and Stratton, 1982: 83-99. Hasson E, Newton WD, Waltman AC, et al. Mural degeneration in the glutaraldehyde-tanned umbilical vein graft: incidence and implications. J Vast Surg 1986; 4: 243-50. Sawyer PN, Adamson R, Butt K, et al. Long-term function of NCGT vascular conduits in a multicenter trial: evaluation of physical chemical parameters. Biomater Med Devices Artlf Organs 1980; 8: 345-67. Lee E, Desu M. A computer program for comparing K samples with right-censored data. Comput Programs Blamed 1972; 2: 315-21. Sottiurdi VS, Batson RC. Autogenous vein graft: experimental studies. In: Stanley JC, ed. Biologic and synthetic vascular prostheses. New York: tine and Stratton, 1982. Stewart SFC, Lyman DJ. Essential physical characteristics of vascular grafts. In: Sawyer PN, ed. Modern vascular grafts. New York: McGraw Hill, 1986: 115-32. Stanley JC, Graham LM, Whitehall WM, Linderman SM. Zerolock GB, Cronenwett JL. Autogenous saphenous vein as an arterial graft: clinical status. In: Stanley JC, ed. Biologic and synthetic vascular prostheses. New York: Grune and Stratton, 1982: 333-49.
537