Influence of endothelial cell seeding on platelet deposition and patency in small-diameter Dacron arterial grafts

Influence of endothelial cell seeding on platelet deposition and patency in small, diameter Dacron arterial grafts B r e n t T . Allen, M.D., Julie A. L o n g , M.D., R i c h a r d E. Clark, M.D., Gregorio A. Sicard, M.D., Kevin T. Hopkins, B.S., and Michael J. Welch, Ph.D.,

St. Louis, Mo. Serial platelet deposition, surface topography, and patency were evaluated in control (N = 28) and endothelial cell-seeded (N = 28) small-diameter (4 mm inner diameter) USCI Dacron grafts implanted in the carotid and femoral arteries of dogs. All dogs received aspirin (325 mg) daily for 2 weeks starting 24 hours prior to graft implantation. Endothelial cell seeding was performed by mixing suspensions of autologous endothelial cells that had been enzymatically harvested from segments of external jugular vein with blood that was used to predot the prostheses. The platelet deposition on each graft was quantitated by means of indium 111 - labeled platelets and technetium 9 9 m - labeled red cells in a dual-isotope platelet-imaging technique. Platelet deposition on seeded grafts 24 hours after implantation was significantly higher than on the controls (p < 0.05). Two weeks after implantation platelet deposition on seeded prostheses had decreased to a level significantly lower than that on the controls and continued to decline on serial studies up to 7 months. In contrast to seeded grafts, platelet accumulation on control grafts dramatically increased after the withdrawal of aspirin therapy and was associated with a sharp rise in control graft thromboses. Gross and scanning electron microscopic evaluation of endothelial cell-seeded grafts after 1 month indicated complete neointimal coverage, whereas none of the control grafts explanted at 1 month or later exhibited a continuous neointimal lining. Cumulative 7-month patency for seeded prostheses was significantly higher than for the controls (96% and 29%, respectively; p < 0.001). We conclude that endothelial cell seeding in combination with short-term aspirin therapy is a simple, reliable method of reducing platelet deposition and significantly improving patency in smalldiameter Dacron prostheses. Abrupt withdrawal of aspirin therapy may be contraindicated in nonseeded control grafts because it results in increased platelet deposition and thrombosis. (J VASC SURG 1984; 1:224-33.)

Autologous vein, usually from the saphenous system, is the prosthesis o f choice for cardiac and peripheral arterial bypass procedures. However, many patients needing surgical revascularization have saphenous veins unsuitable for use in the arterial system. Various synthetic prostheses are available as alternatives to vein grafts, but their use is

From the Division of Cardiothoracic Surgery, Department of Surgery, and the Division of Radiation Sciences,The Edward Mallinckrodt Institute of Radiology, Washington University School of Medicine, St. Louis, Mo. Presented at the Thirty-first Scientific Meeting of the North American Chapter, International Society for Cardiovascular Surgery, San Francisco, Calif., June 16-17, 1983. Supported in part by NIH Grant No. T32 KGMO7602 and SCOR in Thrombosis No. HL14147. Reprint requests: Brent T. Allen, M.D., Division of Cardiothoracic Surgery, Suite 3108 Queeny Tower, 4989 Barnes Hospital Plaza, St. Louis, MO 63110. 224

limited by low patency rates when implanted as small-diameter arterial grafts. 1 The platelet plays a n important role in the genesis o f graft failure, and several investigators have demonstrated in animals that oral antiplatelet therapy can reduce platelet deposition on vascular grafts and improve patency. 2-4 Although several reports suggest that similar observations may be made in humans, ~'" prolonged administration o f platelet inhibitors can be expensive and requires a high degree o f patient compliance. Additionally, Bomberger et al. 7 have demonstrated that antiplatelet agents such as aspirin and dipyridamole may retard endothelial healing in deendothelialized arteries and suggest that these drugs should be used with caution in vascular surgery. Stanley et al. 8 have recently reported that endothelial cell seeding used in conjunction with temporary platelet inhibition is a feasible and effec~~)e

Volume 1 Number 1 lanuary 1984

Endothelial cell seeding

225

Fig. 1. Scanning electron micrograph of luminal surface of a nondistended external jugular vein after endothelial cell harvesting. All but one (arrow) of the endothelial cells were removed, which left a surface composed of collagen and smooth muscle. (×640.)

method of improving patency in small-diameter synthetic grafts. Endothelial cell-seeding techniques ar~attractive because of the rapid development of a nonthrombogenic flow surface, which has morpho.logic and biochemical characteristics similar to those of normal endothelium. The purpose of these investigations was to study platelet deposition, surface topography, and patency in small--diameter endothelial cell-seeded and control grafts implanted in dogs temporarily treated with aspirin (A.S.A.). MATERIAL A N D M E T H O D S

Fourteen mongrel dogs (eight males and six females) weighing 20 to 30 kg were given 325 mg of aspirin daily for 2 weeks starting 24 hours prior to graft implantation. Pentobarbital was administered intravenously as an anesthetic agent at an initial dose of 30 mg/kg, with additional small doses infused periodically as needed.,Intraoperatively each animal was ventilated through an endotracheal tube with a

Harvard respirator and given lactated Ringer's solution containing 5% dextrose intravenously at 15 ml/kg/hr. After the neck and groins were shaved and cleansed with a sterile solution, the external jugular veins were exposed bilaterally through a midline incision. Endothelial cells were enzymatically harvested from 10 cm segments of both veins as described by Stanley et al.s Following the cell derivation procedure samples of the endothelial cell suspensions were microscopically inspected and counted, and cell viability was determined by trypan blue exclusion. The cells were concentrated in 1.0 ml of cell culture medium (Dulbecco's modified essential media) and added to 10 ml of venous blood. Nonsupported, weft-knitted, external velour USCI Dacron prostheses (internal pile height 0.12 mm, external pile height 0.25 mm, wall thickness 0.5 mm, mean water porosity 1900 m l / m i n / c m 2) (C. R. Bard, Inc., Billerica, Mass.) 15 cm long and 4 mm in internal diameter were preclotted according

226

journal of VASCULAR SURGERY

A l l e n et al.

Fig. 2. Scanning electron micrograph of luminal surface of a freshly seeded Dacron graft. A cluster of endothelial cells (arrow) is seen among Dacron fibers, red cells, platelets, and fibrin. (x640.) to the four-step method of SauvageY Endothelial cell seeding was performed with the 10 ml aliquot of blood containing the endothelial cells in the first preclotting step. Control prostheses were handled in the same manner as seeded grafts except that 1 ml of culture medium without endothelial cells was added to the sample of blood used in the first preclotting step. Blood irrigations of both control and seeded prostheses were performed through alternating ends of the graft. After preclotting, 1.5 cm was removed from each end of the graft and the remaining segment cut in two 6 cm lengths. Control and seeded grafts were placed in separate solutions of autologous blood containing heparin (500 U/ml) prior to implantation. Surgical technique. While the endothelial cell's were being harvested, a second surgeon exposed the carotid and femoral arteries bilaterally. Following intravenous heparin administration (100 U/kg) seeded (n = 28) and contralateral control grafts

(n = 28) were implanted end to end in the carotid and femoral arteries with 7-0 Prolene (Ethicon, Inc., Somerville, N.J.) by two surgeons using the triangulation technique of Carrel. 1° The order in w h i c h ~ e grafts were implanted was randomly alternated. Blood flow was restored to all arteries simultaneously after injection of indium 111 - labeled platelets. Cephamandole, 250 mg (Eli Lilly & Co., Indianapolis, Ind.) was adminstered preoperatively (intravenously), postoperatively (intramuscularly), and once daily for two days (intramuscularly). Patency was assessed weekly by palpation (femoral grafts) or Doppler (Parks model 1010, Parks Electronics Laboratory, Beaverton, Ore.) examination (carotid grafts). Chi-square tests with the Yates (1934) modification were used for comparisons of patency rates. Platelet deposition and surface morphology studies. Following graft implantation serial in vivo indium 111- labeled platelet deposition studies

Volume 1 Number 1 January 1984

Endothelial cell seeding

227

Fig. 3. Serial indium 111 - labeled platelet scintigrams from femoral region of a dog studied at 24 hours, 2 weeks, 1 month, and 2 months postoperatively. Platelet deposition on control grafts (large arrows) decreased from 24 hours to 2 weeks but markedly increased after withdrawal of aspirin therapy. Platelet accumulation on endothelial cell-seeded grafts (small arrows) appeared less than that on controls by 2 weeks and continued to decrease after withdrawal of aspirin therapy. were performed on carotid and femoral grafts at 24 hours, 2 weeks, 1 month, and then at intervals to 7 months. Studies done at 2 weeks or later were perfo~aed 24 hours after labeled platelet infusion. In vivo quantitation of platelet deposition was accomplished by means o f a dual-isotope subtraction method. This technique uses both indium 1 1 1 - labeled platelets and technetium 9 9 m - l a b e l e d red cells, and platelet deposition is expressed as percent of indium excess (%InExe). We have previously documented the ability o f this technique to image and accurately quantitate platelet deposition on vascular grafts by correlating %InExc measured in vivo with %InExe measured directly on explanted grafts (r = 0.94; p < 0.01). 11 For control or seeded grafts, analysis o f %InExe data was performed after values from the carotid and femoral regions were combined. The Mann-Whitney rank test 12 was used for all statistical comparisons of %InExe data. Two dogs died following anesthesia for platelet-

Table I. Time of graft thrombosis Interval ofgraft thrombosis 0-24 hr 24 hr-2 wk 2 wk-1 mo 1-2 mo 2-3 mo 3-7 mo Cumulative patency

No. of control gra3~s

No. of seededgrafts

2 2 11 2 3 --

0 1 0 0 0 0

29% (8/28) 96% (27/28) p < 0.001

imaging procedures: one at 24 hours and the other at 2 weeks. Two dogs w e r e sacrificed at 1 month, two dogs at 2 months, and five dogs at 3 months; three dogs survived beyond 6 months. At time o f sacrifice the dogs were anesthetized and given heparin (100 U / k g ) intravenously, and the grafts were exposed. Six patent grafts (five seeded and one control) were perfused in vivo at 100 mm H g with

228

Journal of VASCULAR SURGERY

A l l e n et al.

~----ASA FL., <,05 n:28

~

400

ENDOTHELIAL CELL SEEDED GRAFTS CONTROL

360

520 -

n:28 <.0005

i,i o9 4IX

n =9

280-

<.0005 n =7

240-

x hi ~- 2 0 0 -

160 -

120 -

<.025 n:22 I-seeded onlyH

40-

24 hrs

2 wks

I mo

2 mos

n=15

n=6

5 mos

6-7 mos

TIME Fig. 4. There was no significant difference between carotid and femoral regions in %InExc for similar grafts (control or seeded). Analysis of %InBx~data for each type of graft was performed after results from carotid and femoral regions were combined. Statistical comparisons (p values) between control and seeded grafts at each time interval are indicated above values for n.

normal saline solution followed by 2.5% glutaraldehyde in cacodylate buffer for 15 minutes. These prostheses were then explanted, opened longitudinally, photographed, and postfixed in 1% osmium tetroxide for i hour. Each specimen was dehydrated in ascending grades of alcohol and critical point dried. The grafts were mounted and sputtcrcoated with gold (125 ~l thick), and a Philips 501 scamaing electron microscope (Eindhoven, Netherlands) was used to evaluate luminal topography. Similar studies were performed on tissue samples from jugular veins after endothelial cell harvesting and on sections of freshly seeded grafts. Thirteen thrombosed control grafts were rinsed gently in normal saline solution after removal and fixed in 2.5% glutaraldehyde before the luminal surface was photographed. Twelve seeded and five control prostheses were

explanted and opened immediately after systemi~ heparin administration. These grafts were submitted for biochemical assays and will be the subject of another report. RESULTS The enzymatic harvesting of endothelial cell~ from two external jugular veins, each 10 cm long, required 45 to 60 minutes and yielded suspensions containing 5 x 106 to 1 x 107 endothelial cells. Following cell harvesting many of the cells appeared to be attached to one another by intimal connective tissue and were typically found in clusters of 20 tQ 50 cells. Microscopic studies indicated that 90% to 95% of the harvested cells were viable as measured by trypan blue exclusion. Scanning electron micrographs of the luminal surface of the excised ~i~ns

Volume 1 Number 1 January 1984

after harvesting revealed that virtually all the endothelial cell layer was removed (Fig. 1). Clusters of endothelial cells were only occasionally seen on scanning electron micrographs of the inner surface of freshly seeded grafts (Fig. 2). Patency. Twenty of the 28 control grafts thrombosed during the 7-month study period (Table I). Two control grafts thrombosed during the first 24 hours following implantation, and two occluded at 2 weeks (one was infected). The majority of control graft thromboses (80%) occurred after withdrawal of aspirin therapy. Eleven prostheses thrombosed from 2 weeks to 1 month, two failed from I to 2 months, and the three control grafts that were open at 2 months had thrombosed by the 3-month examination. One endothelial cell-seeded graft thrombosed during the entire study. This graft was from a dog with an infected neck incision that caused the occlusion at 2 weeks of both the control and seeded grafts in the carotid positions, The cumulative 7-month patency rate for seeded grafts was 96% and 29% for control grafts (p < 0.001). Indium 11 i-labeled platelet studies. Platelet scintiphotographs suggested that deposition on both control and seeded grafts was elevated 24 hours after implantation (Fig. 3). Two-week studies indicated a marked reduction in platelet accumulation on both grafts and was generally lower on seeded grafts than on controls. At 1 month, 2 weeks after withdrawal of aspirin therapy, platelet scintigrams demonstrated a dramatic increase in platelet deposition on control grafts that appeared to be symmetric over the entire length of the prosthesis. Control grafts continued to accumulate large numbers of platelets at 2 months. Platelet scintiphotographs of seeded grafts at 1 month indicated minimal accumulation of platelets and were not noticeably different from those images of seeded grafts at 2 weeks. Scintigrams obtained 2 to 7 months following graft placement showed minimal evidence of increased platelet deposition. Platelet deposition as measured by %Insxc for the entire group of dogs is shown in Fig. 4. Seeded grafts accumulated significantly more platelets than did the controls (p < 0.05) in studies done 24 hours following graft implantation. As suggested by the platelet scintiphotographs, the %InExchad markedly declined for both types of grafts at 2 weeks, with platelet deposition on seeded grafts decreasing to a value that was statistically less than that of the controls (p < 0.025). Control grafts at 1 month, 2 weeks after cessation of aspirin administration, demonstrated levels of %Inrxc that were four to five times greater than those of control prostheses stud-

Endothelial cell seeding

229

ied at 2 weeks and were six to seven times greater than %InEx0 values for seeded grafts at 1 month (p < 0.0005 for both comparisons). Platelet deposition increased slightly on seeded grafts from 2 weeks to 1 month, but the difference was not statistically significant. Mean values of %InExc for controis decreased by approximately 20% from 1 to 2 months but remained separated from mean values of %InExc for seeded grafts by a factor of 10 or more. No control grafts were available for comparison after 2 months. Platelet deposition on seeded grafts steadily declined in studies performed from 2 to 7 months after implantation. Postimplantation studies. Examination of patent control grafts explanted at 1 month demonstrated neointimal pannus ingrowth across each anastomosis toward the midportion of the grafts (Fig. 5). The neointimal coverage was not complete in any of the controls, and the areas of the grafts beyond the pannus regions were covered by a carpet of red microthrombi. In contrast, all seeded grafts explanted at 1 month or later revealed a continuous, white, glistening neointimal lining, the luminal surface of which was devoid of adherent microthrombi. The neointima of seeded grafts appeared on gross examination to be thicker on grafts explanted at 3 months than that of those explanted at i month, but the knitted pattern of the Dacron fabric was always easily distinguishable through the neointima. Scanning electron micrographs taken from the central region of the control prostheses indicated a surface composed of red cells, fibrin, and platelet aggregates, whereas scanning electron micrographs from the midportion of seeded grafts revealed a smooth lining of endothelial cells with no thrombotic debris. Gross inspection of control grafts that thrombosed within 1 month of placement suggested that excessive accumulation of thrombotic material, primarily at the anastomoses, was the predominant cause of graft failure. Grafts that occluded after 1 month also had evidence of intraluminal thrombi but displayed, in addition, varying degrees of subintimal thickening near the anastomoses, which may have contributed to graft thromboses (Fig. 6). DISCUSSION

The thrombogenic response to most implanted foreign materials has been a major obstacle in the development of a satisfactory synthetic arterial substitute. We and others have attempted to artificially modify this response by coating the luminal surface of synthetic grafts with compounds containing anticoagulants or platelet inhibitors. 13,14Although these

230

Journal of VASCULAR SURGERY

Allen et al.

Fig. 5. Patent control graft (A) explanted at 1 month from right carotid artery (RC) demonstrates white areas ofneointimal pannus ingrowth from each anastomosis. Neointima was not complete, however, as demonstrated by a scanning electron micrograph (B) taken from dark midportion (squared area inA) of a control graft, which revealed a surface composed of fibrin red cells and platelet aggregates. An endothelial cell- seeded graft (C) explanted at 3 months from left carotid artery (LC) exhibits a complete neointimal lining. A smooth layer of endothelial cells is seen in a scanning electron micrograph (D) taken from squared area of seeded graft in C. (B and D, Original magnification x 1250.)

Fig. 6. A control graft that thrombosed between 2 and 3 months. Blood flow is toward RF. Areas of neointimal pannus ingrowth are indicated by small arrows. Graft occlusion appeared to result from subintimal thickening at distal anastomosis (large arrow).

techniques are technically feasible, they are presently only temporarily effective because of the rapid elution of the active agent(s) from the graft surface. The investigations reported here confirm the observations made by Stanley et al.s that in the presence of temporary platelet inhibition, small-diameter Dacron grafts develop a smooth continuous layer of endothelial cells over the luminal surface when seeded with suspensions of enzymatically derived autologous venous endothelium. The neointimal surface appeared stable over the 7-month study period and may reduce the thrombogenic response to prosthetic material by acting as a nonthrombogenic physical barrier between the blood and thc graft's Dacron fibers. Shareikin et al. 1~ and Clagett et al. TM have provided indirect evidence that the interaction ofplatelets with seeded grafts is reduced by demonstrafing that platelet survival in dogs with endothelial cell- seeded aortic prostheses normalized within

Volume 1 Numbcl- 1 January 1984

8 weeks of implantation (compared with a platelet survival at 8 weeks in dogs with nonseeded grafts that was significantly reduced). The objective of this project was to directly study platelet deposition in vivo on small-diameter endothelial cell-seeded and control Dacron grafts by means of dual-isotope platelet scintigraphy. Having used this technique in previous investigations to quantitate platelet deposition on a variety of vascular grafts, we have noted that early platelet deposition critically influences graft performance and may predict 1-month graft patency, n Burkel et al. 17 have stated that small-diameter synthetic grafts seeded in a manner similar to the method used in our studies have a high rate of early thrombosis when implanted in animals not receiving aspirin and dipyridamole therapy. Their studies indicate that a brief (1 to 2 v ~ k s ) period of systemic platelet inhibition is required to prevent thrombosis while the luminal surface of seeded grafts is being lined with neoendot_helium. Our findings suggest that the endothelial cell seeding procedure acutely increases platelet accumulation and, in the absence of systemic platelet inhibition, this level of platelet deposition would probably lead to graft thrombosis. It appears from our studies that aspirin alone is effective in preventing early thrombosis of seeded grafts. We are uncertain about what factors are responsible for the increased platelet affinity of seeded grafts. Microscopic inspection of freshly harvested endothelial cells indicated that many of the cells were still connected to one another by stromal elements. This subendothelial tissue is known to be very thrombogenic and may stimulate increased platelet accumulation on seeded grafts at 24 hours. 18 Alternat e l y the trypsin and collagenase enzymes used in harvesting may have damaged some of the cells and enhanced their thrombogenicity. Aspirin is known to inhibit cyclooxygenase, an enzyme common to the prostaglandin pathway that leads to the production in platelets of the prothromboric substance thromboxane A2 or in endothelial cells to the synthesis of the antithrombotic agent prostacyclin (PGI2). It is well known that platelets exposed to aspirin are irreversibly inhibited by virtue of their inability to resynthesize cyclooxygenase.'9 In contrast, endothelial cells are capable of synthesizing the enzyme and therefore eventually recover from the inhibitory effects of aspirin. The length of time required before PGI2 production returns and the relative effects of different aspirin doses or administration schedules on the platelet's interaction with the vessel wall are currently controversial subjects a~-~, are beyond the scope of this investigation. 2°

Endothelial cell seeding

231

Two weeks after implantation and under the influence of aspirin administration, platelet deposition on seeded grafts was significantly less than that on the controls. Stanley et a12 have demonstrated that seeded grafts at this stage are nearly completely lined by a thin endothelial cell layer. Clagctt et al. '6 have reported that endothelial cell-seeded aortic grafts harvested at 3 months produce PGI2 similar in amount to aortic tissue. We are uncertain if the differences in platelet deposition at 2 weeks between seeded and control grafts result from differences in surface morphology, biochemistry (PGI2 production), or a combination of the two. In dogs not treated with platelet inhibitors, platelet deposition on all types of vascular grafts is highest within hours of implantation and subsequently declines on serial examinations. 11 The withdrawal of aspirin therapy in this project was followed by a marked rise in platelet accumulation on control grafts (approaching the level of deposition seen on a freshly implanted graft) and a concomitant increase in graft thrombosis. We have noted this phenomenon of increased platelet deposition after a period of systemic platelet inhibition in other studies that used aspirin as a platelet inhibitor.14 The mechanisms responsible for this observation are unclear, but it appears that aspirin may alter the process of graft maturation that in dogs not treated with aspirin leads to a progressive decline in platelet deposition over time. The high rate of control graft thrombosis following cessation of aspirin administration emphasizes the important role of the platelet in graft thrombosis and may have clinical implications suggesting that abrupt withdrawal of aspirin from patients with vascular grafts may be contraindicated. Studies of seeded grafts performed after 1 month indicated a progressive decline in platelet deposition to 7 months following implantation. At 2 months, when platelet survival has reportedly normalized in dogs with seeded aortic prosthcses, 15'IG plat_elet deposition was significantly elevated (%InExc[X --+ SEM] = 12 +4). Although these data are not strictly comparable, they suggest that platelet deposition on seeded grafts is elevated after the platelet survival has normalized. In summary, the endothelial cell-harvesting technique used in these investigations proved to be a rapid, efficient means of obtaining suspensions of viable endothelial cells. Endothelial cell seeding of small-diameter Dacron grafts implanted in dogs temporarily treated with aspirin promoted the development of a smooth endothelial lining that significantly reduced the incidence of graft thrombosis.

232

Journal of VASCULAR SURGERY

Allen et al.

Platelet deposition studies indicated a marked increase in platelet accumulation on control grafts after the withdrawal of aspirin therapy. This elevation in platelet deposition was associated with a sharp rise in the incidence of graft thrombosis. In contrast to the control grafts, platelet accumulation on endothelial cell- seeded prostheses at 2 weeks or later was reduced and not influenced by the withdrawal o f aspirin. The low level o f platelet deposition on seeded grafts was associated with a low incidence o f graft thrombosis. We thank C. R. Bard, Inc., Billerica, Mass.; Ethicon, Inc., Somerville, N.J.; and Eli Lilly & Co., Indianapolis, Ind., for providing materials used in this project. We gratefully acknowledge the technical assistance o f Ms. Carla J. Mathias, Ms. Jac Wilson, and Mr. Robert Feldhans; the laboratory supplies provided by Dr. David W. Scharp; thc consulting comments made by Dr. Barry A. Siegel; and the expert secretarial services extended by Mrs. Frances L. Grubbs. Special thanks is given to Dr. James C. Stanley and his associates at the University o f Michigan Medical School for allowing us to observe their endothclia! cell-seeding techniques.

7.

8.

9.

10. 11.

12. 13.

14.

REFERENCES

1. Callow AD. Historical overview of experimental and clinical development of vascular grafts. In: Stanley JC, Burkel WE, Lindenauer SM, Bartlett RH, Turcotte JG, eds. Biologic and synthetic vascular prostheses. New York: Grune & Stratton, Inc, 1982. 2. Fuster V, Dewanjee MK, Kaye MP, Josa M, Merke MP, Chesebro JH. Noninvasive radioisotopic technique for detection of platelet deposition in coronary artery bypass grafts in dogs and its reduction with platelet inhibitors. Circulation 1979; 60:1508-12. 3. Harjola P-T, Meurala H, Frick MH. Prevention of early reocclusion by dipyridamole and aspirin in arterial reconstructive surgery. J Cardiovasc Surg 1981; 22:141-4. 4. Oblath RW, Buckley FO, Green RM, Schwartz SI, DeWeese JA. Prevention ofplatelet aggregation and adherence of prosthetic vascular grafts by aspirin and persantine. Surgery 1978; 84:37-44. 5. Chesebro JH, Clements IP, Fuster V, Elveback LR, Smith HC, Bardsley WT, Frye RL, Holmes Jr DR, Vlietstra RE, Pluth JR, Wallace, RB, Puga FJ, Orszulak TA, Piehler JM, Schaff HV, Danielson GK. A platelet-inhibitor drug trial in coronary artery bypass operations: Benefit of perioperative dipyridamole and aspirin therapy on early postoperative vein-graft patency. N Engl J Med 1982; 307:73-78. 6. Pumphrey CW, Chesebro JH, Dewanjee MK, Wahner HW, Hollier LH, Paii'olero PC, Fuster V. In vivo quantitation of platelet deposition on human peripheral arterial bypass grafts

DISCUSSION Dr. William G. Malette (Omaha, Neb.). We would like to suggest another way to achieve an endothelial surface on Dacron grafts.

15.

16.

17.

18.

19.

20.

using indium 111 - labeled platelets: Effect of dipyridamole and aspirin. Am J Cardiol 1983; 51:796-801. Bomberger RA, DePalma RG, Ambrose TA, Manalo P. Aspirin and dipyridamole inhibit endothelial healing. Arch Surg 1982; 117:1459-64. Stanley JC, Burkel WE, Ford JW, Vinter DW, Kahn RH, Whitehousc Jr WM, Graham LM. Enhanced patency of small diameter, externally supported Dacron iliofemoral grafts seeded with endothelial cells. Surgery 1982; 92:994-1005. Yates SG, Barros D'Sa AA, Berger K, Fernandez LG, Wood SJ, Rittenhouse EA, David CC, Mansfield PB, Sauvage LR. The preclotring of porous arterial prostheses. Ann Surg 1978; 188:611-22. Carrel A. Results of the transplantation of blood vessels, organs and limbs. JAMA 1908; 51:1662-7. Allen BT, Mathias CJ, Welch MJ, Clark RE. Platelet deposition on vascular grafts: The accuracy of in vivo quantitation and the significance of in vivo platelet reactivity. Circulation. (In press.) Zar JH. In: Biostatistical analysis. Englewood Cliffs, NJ: Prentice-Hall, Inc, 1974:101-20. Hufnagel CA. Heparin bonded surfaces in vascular grits. In: Stanley JC, Burkel WE, Lindenauer SM, Bartlett RH, Turcotte JG, eds. Biologic and synthetic vascular prostheses. New York: Grune & Stratton, Inc, 1982. Allen BT, Sparks RE, Welch MJ, Mason NS, Mathias CJ, Clark RE. Reduction of platelet deposition on vascular grafts using an anti-platelet graft coating technique. J Surg Res. (In press.) Shareff,in JB, Larker C, Smith M, Cruess D, Clagett P, Rich NM. Early normalization of platelet survival by endothelial seeding of Dacron arterial prostheses in dogs. Surgery 1982; 92:385-93. Clagett GP, Burkel WE, Sharefkin JB, Ford JW, Hufnagel H, Vinter DW, Kalm RH, Graham LM, Stanley JC. Antithrombotic character of canine endothelial cell- seeded arterial prostheses. Surg Forum 1982; 33:471-3. Burkel WE, Ford JW, Vinter DW, Kahn RH, Graham LM, Stanley JC. Endothelial cell seeding of enzymatically derived and cultured cells on prosthetic grafts. In: Stanley JC, Burkel WE, Lindenauer SM, Bartlett RH, Turcott JG, eds. Biologic and synthetic vascular prostheses. New York: Grune & Stratton, Inc, 1982. Harker LA. Platelet mechanisms in the genesis and prevention of graft-related vascular injury reactions and thromboembolism. In: Sawyer PN, Kaplitt MJ, eds. Vascular ~ grafts. New York: Appleton-Century-Crofts, Inc, 1978. Romson JL, Shea MJ, Lucches BR. Pharmacologic control of arterial thrombosis. In: Stanley JC, Burkel WE, Lindenauer SM, Bartlett RH, Turcotte JG. Biologic and synthetic vascular prostheses. New York: Grune & Stratton, Inc, 1982. Preston FE, Greaves M, Jackson CA, Stoddard CJ. Low-dose, aspirin inhibits platelet and venous cyclo-oxygenase in man. Thromb Res 1982; 27:477-84.

During our study o f chitosan as a hemostatic agent in knitted Dacron grafts, we examined the grafts microscopically at 24 hours and at 30, 60, 90, and 120 days. Sections were stained with a trichrome sr-aJn

Volume 1 Number 1 January 1984

th'a~ stains fibrous tissue blue-green and smooth muscle red. At 24 hours the acellular chitosan coagulum lines the grafts and plugs the interstices. After 30 days untreated control grafts showed the usual fibrous healing without endothelium. In contrast, the chitosan-treated graft was encased in vascularized smooth muscle. A closer look re~vealed the abundant vascularity invading the muscular layers and penetrating through the graft--almost to the intimal surface. We believe that chitosan treatment has prevented fibroplasia, allowed the ingrowth of smooth muscle, and allowed the development of a native endothelial intima in Dacron grafts. Dr. Walter M. Whitehouse, Jr. (Ann Arbor, Mich.). We recently repotted a canine study on the use of indium 111-labeled platelet imaging of 6 mm knitted Dacron thoracoabdominal grafts seeded with endothelial cells. Our results were similar to those presented here by Dr. Al,.'~n with progressive diminution in platelet deposition on seeded grafts over an 8-week study period. Indeed, there was no detectable platelet deposition at 8 weeks, and these grafts were found to have 99% luminal coverage with endothelium. In Dr. Allen's study minimal but detectable platelet deposition was present at 6 to 7 months. Such discrepancies in these studies may reflect different sensitivities of the methods used, differences in graft size or endothelial coverage, as well as variations in labeling methods. Perhaps the authors could comment on the doubleisotope methodology used to calculate the percent o f indium excess: specifically, how precisely platelet deposition can be localized, whether by looking at the graft as a whole or by looking within individual segments of the graft. Have the authors carefully documented the extent of endothelial cell coverage in both seeded and control groups? In addition, perhaps they could discuss their labeling techniques. Specifically, what was the amount of rad(gactivity added to the platelets at the time o f labeling? Objective methods for in vivo quantitation of platelet deposition will become increasingly important as experimental and clinical activity in endothelial cell seeding expands. Dr. Malcolm B. Herring (Indianapolis, Ind.). I believe that we need to define enzymatic harvesting. There seems to be a plethora of enzymes used in harvesting cells, and we should begin to document them very accurately. There are at least four types of crude collagenases that are available commercially and one type of purified collagenase of which I am aware. Each of these will undoubtedly have different effects on the endothelium. We should make an effort to identify the enzymes carefully in this type o f study.

Endothelial cell seeding

233

In your study you described improved platelet performance and medium-term patency in seeded grafts. I am curious about the short-term patency problem seen when longer segments of small artery are bypassed and when graft occlusion occurs within 1 or 2 hours of implantation. Do you think that short-term patency can be achieved in such seeded grafts simply by adding platelet drugs? I have one other question, which dovetails with Dr. Whitehouse's question. Were you able in your study to correlate the percent of endothelial healing with the decrease in platelet adhesion? Dr. Allen (closing). Dr. Malette, chitosan coating appears to be a promising technique. We would be interested in evaluating platelet deposition on chitosan-coated grafts. Dr. Whitehouse, we are not certain that the dual-isotope imaging technique used in these studies is more sensitive than other platelet-imaging techniques in determining the presence or absence of platelet thrombi. However, we have previously demonstrated that dual-isotope platelet scintigraphy can accurately quantitate platelet deposition on vascular grafts by correlating the percent of indium excess obtained in vivo with that obtained directly through conventional in vitro methods. The accuracy of other methods of quantitating platelet deposition in vivo has not yet been documented. Not all explanted grafts were submitted to scanning electron microscopy; therefore we were unable to statistically compare differences in endothelial cell coverage between control and seeded prostheses. We have had extensive experience in labeling platelets with indium 111 at our institution, and our techniques are well standardized. In canine studies the platelets obtained from 30 ml o f venous blood are labeled with 300 to 600 /zCi of indium. Red cells are obtained from the same sample of blood and labeled with 1 to 2 mCi of technetium 99m. Dr. Herring, we agree with your comments. Two solutions of enzymes were used in this project: one containing 6a0 U / m l of collagenase (CLS type I) in Hanks' solution (containing magnesium and calcium) and the other containing 0.1% trypsin in Hanks' solution without calcium or magnesium. We have recently completed a project to determine the significance of early platelet deposition on eventual graft failure. We found that the degree of platelet deposition 24 hours following implantation critically influenced 1-month graft patency. It appeared that in the canine model excessive early platelet deposition promoted graft failure within 1 month. Data from the present study suggest that in the presence of systemic platelet inhibition, elevated levels of platelet deposition on seeded grafts at 24 hours do not result in thrombosis.