The effect of preclotting on surface thrombogenicity and thromboembolic complications of Dacron grafts in the canine thoracic aorta

J

THoRAc CARDIOVASC SURG

88:253-258, 1984

The effect of preclotting on surface thrombogenicity and thromboembolic complications of Dacron grafts in the canine thoracic aorta Previous reports bave advocated preclotting and autoclaving of Dacron grafts to prevent graft hemorrhage, yet no data delineate possible changes in surface thrombogenicity or thromboembolic risks. To assess these factors, preclotted and preclotted-autoclaved woven Dacron prostheses were implanted in the thoracic aorta of 31 dogs. Grafts were harvested 1 week to 1 month following implantation, and the thrombus-free surface of each graft was calculated by computerized morphometry. Values for thrombus-free surface were lowest in autoclaved grafts preclotted in nonheparinized blood (72 %) or in heparinized blood (78 % ~ Grafts preclotted in platelet-rich plasma and autoclaved bad a thrombus-free surface value of 85 %. Regular preclotted grafts or grafts without preclotting bad the greatest thrombus-free surface value (97 % and 99 % ~ Distal emholization to the kidneys was detected in seven of 31 animals (23 % ~ No emboli were found in animals with grafts that were not preclotted or if grafts were preclotted with platelet-rich plasma and autoclaved. Thrombus-free surface values were lower (p < 0.01) in grafts with embolic complications (71 % versus 91 % ~ These data suggest that, if the risk of severe bleeding in fully heparinized patients precludes the use of a woven Dacron graft without preclotting, autoclaving a graft preclotted with platelet-rich plasma appears to be the method of choice.

Peter Gloviczki, M.D., Larry H. Hollier, M.D., Eric A. Hoffman, Ph.D.,· Gunnar Plate, M.D., Victor F. Trastek, M.D., and Michael P. Kaye, M.D., Rochester, Minn.

"rious preclotting techniques have been proposed to avoid bleeding through the interstices of tightly woven Dacron grafts implanted in the thoracic aorta of patients requiring full heparinization.!' Bethea and Reemtsma I described autoclaving a graft previously soaked in heparinized blood, and Cooley and associates- suggested the use of autologous platelet-rich plasma (PRP) for preclotting before autoclaving. Though both techniques have been successfully applied in a limited number of patients, experimental evaluation of autoclaved grafts, other than leak rate determinations,' has not been previously reported. Although intraoperative bleeding through the graft can be prevented successfully by autoclaving the preclottea prosthesis and decreasing permeability,' changes in From the Department of Surgery and the Department of Physiology and Biophysics, Mayo Clinic and Foundation, Rochester, Minn. Received for publication June 8, 1983. Accepted for publication Sept. 1, 1983. Address for reprints: Larry H. Hollier, M.D., Department of Surgery, Mayo Clinic, Rochester, Minn. 55905. *Department of Physiology and Biophysics.

surface thrombogenicity may be expected. Although long-term patency of these grafts will hardly be affected because of the high flow situation and the large caliber of the replaced vessels, the risk of early thromboembolic complications cannot be underestimated.v" This study was designed to evaluate the surface thrombogenicity of woven Dacron grafts preclotted by various techniques and implanted in the canine thoracic aorta. Distal embolization of both kidneys was studied, and the correlation between increased early thrombus formation and peripheral embolization was examined. Materials and methods Thirty-one randomly selected, healthy mongrel dogs of either sex, weighing 25 to 30 kg, were used for the experiments. The animals were anesthetized with methohexital sodium (Brevital), 11 mg/kg, and anesthesia was maintained with halothane. Gentamicin, 80 mg intravenously, and penicillin G benzathine, 1.2 million units intramuscularly, were given to all animals, and they were prepared and draped in sterile fashion. The left carotid and right femoral arteries were exposed and a left thoracotomy was performed through the fourth 253

The Journal of

254

Gioviczki et ai.

Thoracic and Cardiovascular Surgery

Fig. 1. Scanning electron microscopy of tightly woven Dacron grafts immediately following preclotting. A. Standard preclotting. B. Graft preclotted in nonheparinized whole blood and autoclaved. C. Graft preclotted in heparinized whole blood and autoclaved. D. Graft preclotted in platelet-rich plasma and autoclaved. (Original magnifications: A. B. and C. x50; D. XlOO.)

and seventh intercostal spaces. The descending thoracic aorta was isolated proximally between the left subclavian and first intercostal arteries and distally along a short segment just above the diaphragm. Heparin, 3 mg/kg, was given intravenously to the animals. A 30 ml blood sample was withdrawn for preclotting before or after heparinization, depending upon the type of implanted graft. Both the left carotid artery and the right femoral artery were cannulated, and a carotid-femoral shunt was established. Cooley Veri-Soft tightly woven low-porosity grafts (Meadox Medicals, Inc., Oakland, N. J.), 12 to 16 em long with an internal diameter of 14 to 16 mm, were used to bypass the thoracic aorta. An end-to-side anastomosis was performed proximally and an endto-end anastomosis distally, with 4-0 Prolene running sutures. The proximal stump of the transected lower thoracic aorta was ligated. After completion of the

anastomoses, the clamps were removed, the carotidfemoral shunt disconnected, and heparin was reversed by intravenous infusion of an equal amount of protamine sulfate. After hemostasis was obtained, a chest tube was placed in the left chest and the wounds were closed in three layers in standard fashion. The animals remained on the table for 2 hours, at which time the chest drains were removed and the animals were returned to the kennel. The implanted grafts were prepared in the following ways (Fig. 1). Group I (six animals). No preclotting was used. Group II (seven animals). Standard preclotting was done with 30 ml of nonheparinized whole blood, which was injected into a sterile metal pan, wetting the Dacron graft. One end of the graft was cross-clamped and, by means of a glass syringe, blood was injected into the graft through the open end as long as bleeding occurred.

Volume 88 Number 2 August, 1984

Preclotting of Dacron grafts

255

Fig. 2. Top, Tightly woven Dacron graft implanted without preclotting. Note smooth surface I week after operation. Bottom, Dacron graft with standard preclotting 1 week after implantation.

If no bleeding was noted, excess blood was stripped from the graft. After preclotting, the lumen of the prosthesis was carefully observed and any intraluminal fibrin septa were removed by forceps. Group III (six animals). The graft was preclotted as in Group II and subsequently autoclaved for 10 minutes at 130° C. After autoclaving, the graft was repeatedly flushed with saline, and all visible blood clot was removed before implantation. Group IV (six animals). These dogs were treated like those in Group III, but the graft was preclotted with whole blood withdrawn 5 minutes after intravenous injection of heparin, 3 mg/kg, to the animal. Group V (six animals). The graft was autoclaved in PRP similar to the protocol of Cooley and associates.' Nonheparinized blood 40 ml was withdrawn from the animal into a syringe containing 10 mg of heparin. The blood was injected into a sterile tube, centrifuged at 1,200rpm for 10 minutes, and the PRP suctioned off the red blood cells. Protamine sulfate (10 mg) was added to the plasma and the plasma was poured onto the Dacron graft. The graft was thoroughly soaked in the plasma and placed in a steam autoclave at 130° C for 5 minutes. All grafts except four were harvested 1 week following implantation. Two grafts from the seven animals in Group II were followed up to I month, but two animals at the beginning of the study had to be killed because of paraplegia before 7 days (first and fifth postoperative days). In both of these cases the distal anastomosis was performed first, so that a longer segment of the thoracic

Fig. 3. Dacron graft 1 week after operation. Top, Preclotted in whole blood and autoclaved. Bottom, Preclotted in heparinized blood and autoclaved.

aorta was excluded from the circulation for about 20 minutes. All subsequent implantations were started with the proximal anastomosis and no more complications were observed. The animal was given 100 mg of heparin intravenously and then, 5 minutes later, it was put to death with pentobarbital. The chest was opened and the graft was removed and opened longitudinally. The excess blood was washed off with saline. All grafts were immersed in Trump's fixative. Both kidneys were removed and opened longitudinally, and 10 transverse sections of each half were made. Each section was evaluated for the presence of emboli under 2.5X magnification. All areas suspected of containing emboli were fixed in Bouin's fixative, embedded in paraffm, cut into 5 J.Lm thick sections, and stained with hematoxylin and eosin. The fmal diagnosis of embolization was confirmed by light microscopy. Arteriography with cannulation of the renal artery to verify an embolus was used in representative samples only. After fixation, the grafts were photographed and magnified prints were reviewed by one of us (E. A. H.), who was not aware of the type of preclotting, for evaluation of the thrombus-free surface (TFS), as suggested by Sauvage and colleagues." A computer program was used for morphometric measurements rather than a grid. TFS area was determined as a

2 5 6 Gloviczki et al.

.

The Journal of Thoracic and Cardiovascular Surgery

"

Fig. 4. Kidney infarction 1 week following implantation of graft preclotted in whole blood and autoclaved. Top, Microscopic picture. Bottom, Section from the same with hematoxylin and eosin stain. (Original magnification X16.)

percentage of the total area of each graft via an operator-interactive computer-based (PDP 11/70) video cursor-track ball system. The Student's t test was used for statistical analysis of the data. All animals received humane care in compliance with the "Principles of Laboratory Animal Care" formulated by the National Society for Medical Research and the "Guide for the Care and Use of Laboratory Animals" prepared by the National Academy of Sciences and published by National Institutes of Health (NIH Publication #80-23, revised 1978). Results After autoclaving, the graft material became stiffer and handling and suturing were more difficult than in nonautoclaved grafts. These changes were more apparent in the grafts preclotted in whole blood rather than with PRP. Bleeding through the graft wall during implantation was noted in Group I only; this was controllable in every case, however, following cautious, staged declamping and infusion of protamine sulfate. Median flow measured in the grafts at the end of the procedure was 770 mI with ranges from 600 to 1,400 mI.

Fig. 5. Arteriography of an infarcted kidney of an animal with graft preclotted in whole blood and autoclaved before implantation.

Group I: No preclotting. The mean TFS score was 99% (Fig. 2, top). No renal emboli were detected. Group II: Standard preclotting in whole blood. The mean TFS value was 97% (Fig. 2, bottom). One animal had isolated, large infarcted areas in both kidneys, and another presented with unilateral, multiple small emboli. The remaining five animals had no evidence of renal emboli. Group III: Grafts preclotted in whole blood and autoclaved. The mean TFS score was 72%, the lowest among the five groups (Fig. 3, top). One animal presented with massive multiple embolization in the right kidney and had multiple small emboli in the left (Fig. 4); the TFS score was 39%, the lowest in the study. Another animal had unilateral massive embolization, and the angiogram revealed the infarcted upper third of the kidney (Fig. 5). Four animals in this group had no evidence of renal emboli. Group IV: Grafts preclotted in heparinized blood and autoclaved. The TFS scores were generally low (78%) (Fig. 3, bottom); however, there was one graft that had no thrombus on the surface. Three of six animals presented with bilateral kidney embolization;

Volume 88

Preclotting of Dacron grafts 2 5 7

Number 2 August. 1984

~

10 0

UJ

en

...

80

QJ

o

~ ~

'"

60

QJ

~

.D "'"

40

a

Fig. 6. Graft preclotted in platelet-rich plasma and autoclaved; harvested at I week.

one had small, isolated emboli, one had small, multiple emboli, and one animal had massive multiple infarctions in the left kidney and small, multiple emboli in the right kidney. Group V: Grafts preclotted in PRP and autoclaved. Mean TFS score was 85%, and the TFS score was above 95% in three of six grafts (Fig. 6) . Thrombus was well adherent and more uniform in every case than in Groups II, III, and IV. No embolization could be detected in the kidneys. The TFS scores for Groups I and II were significantly (p < 0.(01) higher than for any other group. Differences between Groups III, IV, and V were statistically not significant (Fig. 7). There was a definite correlation between the TFS scores and the incidence of distal embolization irrespective of the type of preclotting used to prepare the graft (p < 0.01) (Fig. 8). Of four of the grafts with TFS scores under 70%, all had thromboembolic complications.

~ 't.

20 0 L...--"---L.._-'----'_---'_-"--_-"---L.._-'----'.....

NP

PWB

PWB + A

PHB + A

PP + A

Fig. 7. Thrombus-free surface area of canine thoracic Dacron grafts I week following implantation. NP, No preclotting. PWB, Preclotted in whole blood. PWB + A, Preclotted in while blood and autoclaved. PHB + A, Preclotted in heparinized blood and autoclaved. PP + A, Preclotted in platelet-rich plasma and autoclaved.

::E

100

ui

en

...

80

QJ

o

~ ~

60

'" QJ QJ

'""

.D

40

E

20

't.

0

e ::

No Embol i (n= 2 4)

Emboli (n = 7)

Discussion

Fig. 8. Thrombus-free surface area of canine thoracic Dacron grafts with and without distal embolization (p < 0.01).

Hemorrhage through the wall of porous fabric grafts can be life threatening, especially in patients requiring full heparinization for cardiopulmonary bypass. Operative blood loss can be diminished effectively by using tightly woven, low-porosity Dacron grafts. Even if these grafts are preclotted with the standard technique, bleeding through the wall can occur in heparinized patients. Cooley and associates ' call this phenomenon "declotting" when circulating, heparinized blood removes the loosely, irregularly attached fibrin layer from the graft surface. While leak rates of tightly woven grafts decrease from III ± 7 ml/cm'ymin to 23 ± 4 mIj cm 2jmin with standard preclotting techniques, autoclaving has been shown to diminish leak rates to 3 ± 4 mIjcm2jmin. Autoclaving successfully prevents "declotting" of vascular grafts as they become more resistant to fibrinolysins than regularly preclotted grafts.' In our

study, no bleeding through the graft wall was detected in grafts preclotted in either whole blood or PRP and autoclaved. Early and late thromboembolic complications of Dacron grafts have been described previously in experimental animals and in patients.r " Distal embolization in the form of microembolization probably is overlooked frequently, since the symptoms are not clinically significant. Even microemboli, however, can cause damage in important organs, such as the brain and kidney. Early embolization from the inner surface of the graft could be the result of preclotting techniques or early thrombus formation at the anastomotic area or the thrombogenic surface of the Dacron graft. Late embolization may result from secondary changes within the neointima of the graft (necrosis, dissection , or ulcerationj.v"

2 5 8 Gloviczki et al.

Our data suggest that the risk of embolization depends upon the thrombus formation on the inner surface of the woven Dacron grafts. As graft thrombogenicity increases significantly with autoclaving, the risk of distal embolization increases accordingly. The kidney, as a biologic filter to evaluate distal embolization from implanted prostheses, has been used in earlier experimental studies."!' Del Campo and associates" observed kidney embolization in 20% of all animals with grafts implanted into the suprarenal abdominal aorta. Two of 15 animals with woven Dacron grafts and four of 15 with knitted Dacron grafts presented with infarction in one or both kidneys at 4 to 20 days after operation. Of the seven animals that presented with kidney embolization in our study, three had microembolization with occlusion in the interlobular and cortical branches of the renal artery and, in four animals, massive isolated or multiple kidney embolization could be detected. Woven grafts without preclotting showed excellent characteristics in our study, with low surface thrombogenicity and no complication. A definite correlation could be established between the extent of early thrombus formation in the graft and the incidence of distal embolization in Groups II, III, and IV. Standard preclotting of tightly woven Dacron grafts results in a looselyadherent fibrin coating of the surface with the risk of distal microembolization. Autoclaving may cause denaturation of cellular proteins and aggregated erythrocytes, which can be the sources of microemboli, a risk that seems to be diminished with the use of PRP. No embolization occurred if PRP was used to preclot the graft before autoclaving, and bleeding through the graft interstices was also prevented by this method. Though thrombus deposition at 1 week could be observed on the surface of grafts preclotted in PRP, it was a tightly adherent, white clot in all instances, without apparent danger of embolization. Preparation of the tightly woven Dacron graft in this way, if required because of full heparinization of the patient, is therefore suggested. We appreciate the technical assistance provided by Mr. O. Arlan Hildestad and Mrs. Kathy J. Topham.

The Journal of Thoracic and Cardiovascular Surgery

REFERENCES Bethea MC, Reemtsma K: Graft hemostasis. An alternative to preelotting, Ann Thorac Surg 27:374, 1979 2 Cooley DA, Romagnoli A, Milam JD, Bossart MI: A method of preparing wovenDacron aortic grafts to prevent interstitial hemorrhage. Cardiovasc Dis Bull Tex Heart Inst 8:48-52, 1981 3 Glynn MFX, Williams WG: A technique for pre-clotting vascular grafts. Ann Thorac Surg 29:182-183, 1980 4 Kouchoukos NT, Karp RB, Blackstone EH, Kirklin lW, Pacifico AD, Zorn GL: Replacement of the ascending aorta and aortic valve with a composite graft. Ann Surg 192:403-413, 1980 5 Thurer RL, Hauer JM, Weintraub RM: A comparisonof preclotting techniques' for prosthetic aortic replacement. Circulation 66:Suppl 1:143-146, 1982 6 Weselowski SA, Fries CC, Karlson KE, De Bakey M, Sawyer PN: Porosity. Primary determinant of ultimate fate of synthetic vascular grafts. Surgery 50:91-105, 1961 7 Yates SG II, Nakagawa Y, Berger K, Sauvage LR: Surface thrombogenicity of arterial prostheses. Surg Gynecol Obstet 136:12-16, 1973 8 Sharp WV: The carotid artery. A test site for small vessel prosthetics. J Surg Res 10:41-46, 1970 9 Kusserow B, Larrow R, Nichols J: Observations concerning prosthesis-induced thromboembolic phenomena made with an in vivo embolus test system. Trans Am Soc Artif Intern Organs 16:58-62, 1970 10 Kusserow BK: The use of pathologic techniques in the evaluation of emboli from prosthetic devices. Bull NY Acad Med 48:468-477, 1972 11 Del Campo A, Charoenkul V, Ostreicher R, Peirce EC II: A test of thromboembolisms from arterial grafts. Preliminary report of a new technique. J Cardiovasc Surg 21:219-224, 1980 12 Stratton JW, Hall RV: Pseudointimal embolism from a woven Dacron graft. Surgery 86:772-773, 1979 13 Sauvage LR, Walker MW, Berger K, Robel SB, Lischko MM, Yates SG, Logan GA: Current arterial prostheses. Experimental evaluation by implantation in the carotid and circumflex coronary arteries of the dog. Arch Surg 114:687-691, 1979