- Email: [email protected]
Prevention of venous thrombosis in microvascular surgery by transmural release of heparin from a polyanhydride polymer
Prevention of venous thrombosis in microvascular surgery by transmural release of heparin from a polyanhydride polymer Lisa A. Orloff, MD, Michael G. Glenn, MD, Abraham J. Domb, PhD, and Ramon A. Esclamado, MD, San Diego, Calif., Seattle, Wash., Jerusalem, Israel, and Ann Arbor, Mich. Background. The effects of transmurally relased heparin on the patency of microvenous anastomoses were studied by using a bioerodible polymer delivery system in a rat microvascular thrombosis model. Methods. A polyanhydride carrier with heparin was wrapped around the outside of a highly thrombogenic venous inversion graft in 14 animals, and patency rates were compared with those of 17 control animals. Results. Anastomotic patency was significantly greater in the groups treated with transmurrally released heparin, measured both at 2d hours (86% versus 16%; p < 0.02) and at 7 days (86% versus 36%; p < 0.05) after operation. No significant complications occurred. Conclusions. Controlled release of heparin by transmural delivery is an effective and safe form of local antithrombotic therapy and may have applications both in microvascular and large vessel surgery. (SuRe;ERr 1995;1 / 7:554-9.) From the Division of Otolaryngology-Head and Neck Surgery, University of Califi)rnia, San Diego Medical Center, San Diego, Calif.; Department of Otolaryngology-Head and Neck Surgery, University of Washington, Seattle, Wash.; Department of Pharmaceutical Chemistry, Hebrew University of Jerusalem, Jerusalem, Israel; and the Department of Otolaryngology-Head and Neck Surgery, University of Michigan, Ann Arbor, Mich.
As THE APPLICATION OF microvascular techniques to reconstructive surgery has become widespread, it has become increasingly apparent that the prevention of anastomotic complications is the most important determinant of success. Vascular thrombosis may occur in up to 10% of microvascular anastomoses and is by far the most significant complication of microvascular surgeryJ "3 Surgical technique is the single most critical factor in determining anastomotie patency. Other variables such as intimal trauma, precision of suturing, and duration of vascular clamp application are also important. Thrombosis occurs as a result of deposition and aggregation of platelets on damaged endothelium. 4 Because thrombosis is accelerated in areas of stasis or low
Supported by National Institutes of Health grant T-32-DC00018, BSF #910021, Presented in a preliminary report at the Research Forum of the American Academy of Otolaryngology-Head and Neck Surgery, Sept. 9-13, 1990, in San Diego, Calif. Reprint requests: Lisa A. Orloff, MD, Division of OtolaryngologyHead and Neck Surgery, University of California, San Diego Medical Center, San Diego, CA 92103-8891, Copyright | 1995 by Mosby-Year Book, Inc. 0039-6060/95/$3.00 + 0 11/56/61214 554
SURGERY
blood flow, venous thrombosis is more common than arterial thrombosis in the clinical setting, s Understandably, there has been an ongoing search for pharmacologic adjuvants to technical expertise. Various agents with either prophylatic or therapeutic antithrombotic activity have been studied. Heparin has been studied more than any other agent. Recently, local administration of heparin to microvascular arterial anastomoses 6 and to arteries with intimal damage 7 has been investigated. Transmural delivery of heparin from a polyvinyl alcohol carrier in these experimental arterial thrombosis models was found to have a significant antithrombotic effect without causing systemic anticoagulation. However, use of polyvinyl alcohol as a vehicle for transmural delivery of heparin was associated with an unacceptably high incidence of local hematoma. Other bioerodible carrier matrices have been evaluated for use in controlled delivery applications, s'lx Polyanhydrides are versatile polymers of aromatic and aliphatic decarboxylic acids that have high hydrolytic reactivity based on their anhydride linkage. The labile hydrophilic anhydride group ensures biodegradability, whereas a variety of hydrocarbon backbones can be incorporated to modify the degradation rate of the polymer. Hydrolytic surface erosion follows nearly zero-or-
Surgery
Orloff et al.
555
Volume 117, Number 5
der kinetics and leads to a decrease in size of the delivery vehicle while physical integrity is maintained. Release of biologically active compounds from polyanhydride-drug matrices parallels polymer erosion, with similar kinetics. Such polymers containing chemotherapeutic drugs have already received U.S. Food and Drug Administration approval for clinical testing in human cancers. Similar polymers would seem to be promising vehicles for local delivery of heparin to anastomotic sites in blood vessels. The arterial inversion graft has been described as a uniformly thrombogenie model for the study of thrombosis in microvascular arterial anastomoses)' 12 Either systemic or local heparin therapy results in a significant increase in patency rates in arterial inversion grafts. Studies in our laboratory have shown that a similar vein inversion graft in rats will also reliably develop thrombosis. Because venous thromboses predominate in the clinical setting, the present study was undertaken to determine the efficacy of heparin released transmurally from a polyanhydride delivery system in such a venous thrombosis model.
MATERIAL AND METHODS Adult male Sprague-Dawley rats (390 to 490 grams; Simonson, Gilroy, Calif.) were anesthetized with intraperitoneal equithesin (each milliliter of which contains chloral hydrate, 0.042 gm; pentobarbital, 0.0097 gm; and magnesium sulfate, 0.021 gm) by using a 3 ml/kg initial dose with 0.2 ml increments as needed. Care was taken to ensure humane treatment of the animals in accordance with the guidelines set forth by the National Society for Medical Research and the National Academy of Sciences. 13 Both groins were shaved and treated with a depilatory agent. The right femoral vein was exposed through a 3 cm linear incision. The profunda femoris vein was ligated and divided. As much adventitia as possible was removed from the femoral vein just distal to the profunda branch, taking care not to perforate the vein. An approximating clamp was applied, and a 1 mm segment of vein was removed and inverted. This segment was reanastomosed by microsurgical technique by using 10 sutures per anastomosis of 10-0 monofilament nylon suture on a 75 #m needle (MS/9; Ethieon, Inc., Somerville, N.J.). The polymer, poly(dimer erucic acid-cosebaeic acid) 1:1, was prepared by melt condensation as previously described. 14 Heparin-loaded polyanhydride films were prepared by melt mixing heparin powder in the polymer. Four groups of rats were studied. In all groups a single preoperative dose of 400 units/kg heparin sodium was administered by subcutaneous injection immediately after induction of anesthesia. In the control groups
Fig. 1. Illustration of venous inversion graft with polymer positioned in three strips around vessel at anastomotic site.
(A and B) the segmental inversion graft was performed, and 1 to 2 mg of 0.12 mm polyanhydride film without heparin was laid in three strips around the vessel at the anastomotic site (Figs. 1 and 2). In the experimental groups (C and D) a similar amount of 0.10 mm polyanhydride film containing 5 units/rag heparin (i.e., 5 to 10 units of heparin) was applied. Animals were randomized by assigning them alternately to group A, B, C, or D consecutively. The degree of systemic anticoagulation was assessed in each rat by measuring the partial thromboplastin time ( P T T ) in blood drawn from the left femoral vein 2 hours after the subcutaneous injection of heparin and again when the animal was killed. Rats in group A (n = 6) and group C (n = 7) were killed 24 hours after operation, whereas those in group B (n = t 1) and group D (n = 7) were killed at 7 days. Immediately before death the venous inversion graft was reexposed under general anesthesia. Flow was assessed by direct visualization through the operating microscope and then by a "milking test" cephalad to the anastomosis. Vessels were classified as either patent or thrombosed. Gross determination proved to be consistently accurate when confirmed by dividing the femoral vein cephalad to the inversion graft and observing free blood flow or lack thereof. Patency was essentially an
556
Surgery
Orloff et al.
May 7995
Fig. 2. Photograph of venous inversion graft with polyanhydride polymer in place (original magnification X 16).
t Fig. 3. Patent vein at 7 days after operation exhibits no inflammation (trichrome stain; original magnification X400).
all-or-nothing phenomenon; vessels were either fully patent or completely thrombosed. T h e vein inversion segment was excised and fixed in 10% buffered formalin. Specimens were embedded in paraffin, and representative 8 #m sections were stained with hemotoxylin-eosin and trichrome stains for histologic study by light microscopy.
RESULTS Patency results are shown in the T a b l e . In the anastomoses studied at 24 hours, only one of six control yes-
sels was patent, whereas six of seven veins treated with transmurally released heparin were patent. This difference is statistically significant by Fisher's exact test (p < 0.02). In animals studied at 7 days, four of 11 control vessels were patent, whereas six of seven heparintreated veins remained open. This difference is also statistically significant by the same test (p < 0.05). No major complications occurred. Gross inflammation was absent or minimal in all specimens and was only seen around vessels with thrombosis and local venous stasis. No hematomas were present around the
Surgery Volume 117, Number 5
Orloff et al.
557
Table. Results
Group
n
Preoperative subcutaneous heparin
A B C D
6 11 7 7
Yes Yes Yes Yes
Heparinpolyanhydride polymer
Time
(0
No. patent
% patent
p vs control
P T T at death (sec)
No No Yes Yes
24 hr 7 days 24 hr 7 days
1 4 6 6
16 36 86 86
--<0.01 <0.05
38.6 29.0 34.7 24.1
venous anastomoses. Histologic analysis of patent veins by light microscopy showed walls that were not thickened and had little if any acute inflammatory infiltrate (Fig. 3). By contrast, thrombosed veins showed intramural polymorphonuclear leukocytes and intraluminal platelet and fibrinous debris (Fig. 4). Measurements of P T T confirmed the development of systemic anticoagulation in all animals after the initial preoperative subcutaneous dose of heparin. Mean P T T for all rats at t = 2 hours was 117.5 seconds. At the time of death the P T T had returned to a level that approximated the normal laboratory mean of 31 seconds. P T T was 38.6 seconds in control animals versus 34.7 seconds in study animals at 24 hours and 29 seconds in control animals versus 24.1 seconds in study animals at 7 days. This normalization of P T T showed that ongoing transmural release of heparin did not cause systemic anticoagulation. DISCUSSION The consequences of microvascular thrombosis are often catastrophic, resulting in tissue or flap necrosis from ischemia or venous eongestion. Even in experienced hands, thrombosis can result from alterations in blood flow, endothelial damage, or hypercoagulability. Blood flow is temporarily altered during performance of microvascular anastomoses, and unless adequate flow is restored promptly, thrombosis may occur almost immediately. Exposure of subendothelial elements such as collagen, elastin, and microfibrils 4 or foreign material such as suture at the anastomotic site stimulates platelet aggregation and may trigger the clotting cascade. Although absolute hypercoagulability is not often present, the interplay of stasis and endothelial damage may create a relatively thrombogenic or hypercoagulable setting. Heparin has long been used to offset the potential coagulation that follows vascular manipulation. Heparin is a mucopolysaccharide that prevents new fibrin formation and the subsequent incorporation of fibrin into a thrombus. Heparin binds the plasma protein antithrombin III and enhances the ability of antithrombin III to neutralize the activated serine protease clotting factors (XII, XI, IX, X). Heparin also reduces platelet adhesiveness and directly inhibits individual clotting
factors, such as Xa, at concentrations significantly lower than those at which interaction with antithrombin III occurs. 3
Bleeding is the major complication of systemic heparin therapy, with an incidence of 1% to 10%. 15 Thrombocytopenia, lipolysis, hypersensitivity reactions, rebound hypercoagulability, and bone disease are less common side effects of heparin therapy. 16 Although prophylactic systemic anticoagulation with heparin reduces the incidence of thrombosis, the risks are generally thought to outweigh the benefits in the routine clinical practice of microvascular surgery. Heparin has been shown to diffuse from the extravascular to the intravascular space within 30 minutes of application. Mayberg 17 identified radiolabeled heparin, released from locally applied polyvinyl alcohol, within the cytoplasm of the cells of the vessel wall within hours after application and found that it remained in high local concentrations for 24 to 48 hours. Experimental studies of thrombosis prevention require a reliable animal model in which the rate of thrombosis is consistently high. Simple transection and microvascular reanastomosis of blood vessels are not suitable for such studies because the thrombosis rate is low and prohibitive numbers of animals are required to show significant antithrombotic benefits. The arterial inversion graft model described by Greenberg et al. 3 was an important development that represented a reliable model of microvascular thrombosis. The inversion graft technique is very reproducible, and its thrombogenicity can be precisely controlled by varying the length of the inverted segment. The inverted segment exposes type III fibrillar collagen to the vessel lumen, thereby mimicking clinical situations in which strands of adventitia are trapped in the anastomosis. This adventitial collagen is a potent platelet activator. Because venous thromboses are more common than arterial thromboses, we chose to study a venous rather than arterial inversion graft. The thrombosis rate for this vein inversion graft, in the absence of any anticoagulant therapy, was found to be 100%. In fact, in our preliminary studies the venous inversion graft was so viciously thrombogenic that without a preoperative dose of heparin the graft thrombosed immediately in 10 of 10 rats, before the heparin-containing polymer could even
558
Orloff et al.
Surgery May 1995
Fig. 4. Thrombosed vein in group B at 7 days after operation exhibits intraluminal inflammation (trichrome stain; original magnification Xl00). be applied and its effect observed. To circumvent this initial thrombosis, a brief period of systemic anticoagulation was induced before operation to create a window in time in which to complete the anastomoses, confirm immediate anastomotic patency, and apply the polymer. The single preoperative subcutaneous dose of heparin allowed for an initial period of pateney, but vessels that received no further heparin later thrombosed. The purely temporary nature of the systemic anticoagulation was confirmed by measurement of P T T 2 hours after the preoperative heparin injection and at the time of death either 24 hours or 7 days later. All animals exhibited normalization of P T T by the time of death. Because both treated and control animals received preoperative subcutaneous heparin and because locally delivered heparin was not found to have any measurable systemic anticoagulant effect, any differences in final patency rates were ascribed to the heparin released locally from the polyanhydride-heparin polymer. We postulate that while the effect of the preoperative dose of systemic heparin was being lost, controlled release of heparin was occurring in an ongoing fashion, resulting in a sustained antithrombotic effect. Drugs are released from the matrices of polyanhydride polymer vehicles at the same rate at which the polymers' hydrocarbon backbone is hydrolyzed. Drug release can occur over minutes, hours, weeks, or months, depending on the hydrophilicity of the polymer within which the drug is embedded. The relatively hydrophilic polyanhydride polymer that we used had an effect almost immediately after its application, with an ongoing effect during the following week. Further studies may help to characterize the precise release kinetics of
this particular heparin-polyanhydride polymer, but we presume that the locally released heparin reached an effective level within several hours. It is during these initial hours that the risk of microvascular thrombosis is highest, especially in venous anastomoses. Continued heparin release at the anastomotic site by ongoing hydrolysis of the polyanhydride film provided a persistent antithrombotic effect. The polyanhydride polymer itself, like commonly used surgical materials such as absorbable cellulose film and absorbable cellulose sponge, is inert and biocompatible. Polyanhydrides are nonmutagenic and noncytotoxic and do not react with the drugs embedded in their matrices. Breakdown products of polyanhydrides inelude carboxylic acids, aldehydes, water, and Krebs cycle intermediates that are excreted through direct elimination or through conjugation with amino acids before elimination. Previous biocompatibility studies have documented a similar lack of tissue reactivity or toxicity.9 Furthermore, the lightweight translucent polymeric film had no gross adverse effects such as compression or distortion of the anastomotic site. Although the polyanhydride-heparin polymer did not cause a measurable difference in P T T , it was associated with a statistically significant difference in immediate and long-term patency rates. Treated animals had significantly higher patency rates, at both 24 hours and 7 days after operation. Interestingly, the efficacy of the heparin is attributable to an average of only 6.5 units (1.3 mg of film, with 5 units heparin/mg of film.) Given such a small dose of heparin, the difference in patency rates in the polymer-treated groups is best explained by a highly localized and efficient effect of controlled-
Surgery Volume 777, Number 5
Orloff et al.
release heparin. A l t h o u g h both passive diffusion and active transport m e c h a n i s m s of t r a n s m u r a l h e p a r i n delivery can be postulated, the exact contribution by either of these m e c h a n i s m s is as yet unproven. Similarly, w h e t h e r the h e p a r i n passed t h r o u g h the intact vessel wall tissues or between tissues t h r o u g h the anastomotic site is uncertain. Investigations of the m o d e of t r a n s p o r t and the site of action of locally released h e p a r i n form the basis of future studies w i t h p o l y a n h y d r i d e - h e p a r i n . Because anastomotic patency rates w e r e significantly improved in the absence of any a p p a r e n t side effects, the use of this p o l y a n h y d r i d e - h e p a r i n p o l y m e r for further e x p e r i m e n t a l trials and for u l t i m a t e use in h u m a n beings is very promising. Indeed, t r a n s m u r a l p o l y a n h y d r i d e - h e p a r i n t h e r a p y m a y have clinical applications in vascular surgery involving large vessels, as well as in microvascular surgery. REFERENCES 1. Shaw WW, Converse JM. A survey of 2,680 free flaps: survival, donor site and applications among experienced microvascular surgeons-proceedings of the Fiftieth Annual Convention of the American Society of Plastic and Reconstructive Surgery. Plast Surg Forum 1981;4:93. 2. Davies DM. A world survey of anticoagulation practice in clinical microvascular surgery. Br J Plast Surg 1982;35:96-9. 3. Greenberg BM, Masem M, May JW. Therapeutic value of intravenous heparin in microvascular surgery: an experimental vascular thrombosis study. Plast Reconstr Surg 1988;82: 463-9. 4. Baumgartner HR. Platelet interaction with vascular structures. Thrombos Diathes Haemorrh 1972;51:161-76. 5. Elcock HW, Fredrickson JM. The effect of heparin on throm-
559
bosis at microvenous anastomatic sites. Arch Otolaryngol 1972; 95:68-71. 6. Jones NS, Glenn MG, Orloff LA, Mayberg MR. Prevention of microvascular thrombosis with controlled-release transmural heparin. Arch Otolaryngol Head Neck Surg 1990;116:779-85. 7. Deykin D. Current status of anticoagulant therapy. Am J Med 1982;72:659-64. 8. Domb AJ, Maniar M. Implantable biodegradable polymers as drug-carrier systems. In: Dumitriu S, ed. Polymeric biomaterials. New York: Marcel Dekker, 1994:399-433. 9. Brem H, Domb AJ, Lenartz D, Dureza C, Olivi A, Epstein JI. Brain biocompatibility of a biodegradable controlled release polymer consisting of anhydride copolymer of fatty acid dimer and sebacic acid. J Controlled Release 1992;19:325-30. 10. Chasin M, Domb AJ, Ron E, et al. Polyanhydrides as drug delivery systems. In: Langer R, Chasin M, eds. Polymers as drug delivery systems. New York: Marcel Dekker, 1990. 11. Langer R, Brown L, Edelman E. Controlled release of magnetically modulated release systems for macromolecules. Methods Enzymol 1988;142:399-422. 12. Kersh RA, Handren J, Hargreuter C, May JW. Microvascular surgical experimental thrombosis model: rationale and design. Plast Reconstr Surg 1989:83:866-74. 13. National Academy of Sciences. Guide for the care and use of laboratory animals. National Institutes of Health publication no. 80-23, revised 1978. 14. Domb AJ, Maniar M. Absorbable biopolymers derived from dimer fatty acids. J Polym Sci 1993;31:1275-85. 15. Hayden RE, McLear PW, Phillips JG, Dawson SM. Thrombolysis with systemically administered t-PA in a new venous thrombosis model. Laryngoscope 1989;99:100-4. 16. Okada T, Bark DH, Mayberg MR. Local anticoagulation without systemic effect using a polymer heparin delivery system. Stroke 1988;19:12-8. 17. Mayberg MR. Controlled release of perivascular heparin. In: Barrow D, ed. Perspectives in neurological surgery. St Louis: Quality Medical Publishing, 1990:77-95.
AVAILABILITY OF JOURNAL BACK ISSUES As a service to our subscribers, copies of back issues of SURGERY for the preceding 5 years are maintained and are available for purchase from the publisher, Mosby, at a cost of $8.50 per issue. The following quantity discounts are available: 25% off on quantities of 12 to 23, and one third off on quantities of 24 or more. Please write to Mosby-Year Book, Inc., Subscription Services, 11830 Westline Industrial Drive, St. Louis M O 63146-3318, or call (800)453-4351 or (314)453-4351 for information on availability of particular issues. If unavailable from the publisher, photocopies of complete issues are available from University Microfilms International, 300 N. Zeeb Rd., Ann Arbor, M I 48106 (313)761-4700.
As THE APPLICATION OF microvascular techniques to reconstructive surgery has become widespread, it has become increasingly apparent that the prevention of anastomotic complications is the most important determinant of success. Vascular thrombosis may occur in up to 10% of microvascular anastomoses and is by far the most significant complication of microvascular surgeryJ "3 Surgical technique is the single most critical factor in determining anastomotie patency. Other variables such as intimal trauma, precision of suturing, and duration of vascular clamp application are also important. Thrombosis occurs as a result of deposition and aggregation of platelets on damaged endothelium. 4 Because thrombosis is accelerated in areas of stasis or low
Supported by National Institutes of Health grant T-32-DC00018, BSF #910021, Presented in a preliminary report at the Research Forum of the American Academy of Otolaryngology-Head and Neck Surgery, Sept. 9-13, 1990, in San Diego, Calif. Reprint requests: Lisa A. Orloff, MD, Division of OtolaryngologyHead and Neck Surgery, University of California, San Diego Medical Center, San Diego, CA 92103-8891, Copyright | 1995 by Mosby-Year Book, Inc. 0039-6060/95/$3.00 + 0 11/56/61214 554
SURGERY
blood flow, venous thrombosis is more common than arterial thrombosis in the clinical setting, s Understandably, there has been an ongoing search for pharmacologic adjuvants to technical expertise. Various agents with either prophylatic or therapeutic antithrombotic activity have been studied. Heparin has been studied more than any other agent. Recently, local administration of heparin to microvascular arterial anastomoses 6 and to arteries with intimal damage 7 has been investigated. Transmural delivery of heparin from a polyvinyl alcohol carrier in these experimental arterial thrombosis models was found to have a significant antithrombotic effect without causing systemic anticoagulation. However, use of polyvinyl alcohol as a vehicle for transmural delivery of heparin was associated with an unacceptably high incidence of local hematoma. Other bioerodible carrier matrices have been evaluated for use in controlled delivery applications, s'lx Polyanhydrides are versatile polymers of aromatic and aliphatic decarboxylic acids that have high hydrolytic reactivity based on their anhydride linkage. The labile hydrophilic anhydride group ensures biodegradability, whereas a variety of hydrocarbon backbones can be incorporated to modify the degradation rate of the polymer. Hydrolytic surface erosion follows nearly zero-or-
Surgery
Orloff et al.
555
Volume 117, Number 5
der kinetics and leads to a decrease in size of the delivery vehicle while physical integrity is maintained. Release of biologically active compounds from polyanhydride-drug matrices parallels polymer erosion, with similar kinetics. Such polymers containing chemotherapeutic drugs have already received U.S. Food and Drug Administration approval for clinical testing in human cancers. Similar polymers would seem to be promising vehicles for local delivery of heparin to anastomotic sites in blood vessels. The arterial inversion graft has been described as a uniformly thrombogenie model for the study of thrombosis in microvascular arterial anastomoses)' 12 Either systemic or local heparin therapy results in a significant increase in patency rates in arterial inversion grafts. Studies in our laboratory have shown that a similar vein inversion graft in rats will also reliably develop thrombosis. Because venous thromboses predominate in the clinical setting, the present study was undertaken to determine the efficacy of heparin released transmurally from a polyanhydride delivery system in such a venous thrombosis model.
MATERIAL AND METHODS Adult male Sprague-Dawley rats (390 to 490 grams; Simonson, Gilroy, Calif.) were anesthetized with intraperitoneal equithesin (each milliliter of which contains chloral hydrate, 0.042 gm; pentobarbital, 0.0097 gm; and magnesium sulfate, 0.021 gm) by using a 3 ml/kg initial dose with 0.2 ml increments as needed. Care was taken to ensure humane treatment of the animals in accordance with the guidelines set forth by the National Society for Medical Research and the National Academy of Sciences. 13 Both groins were shaved and treated with a depilatory agent. The right femoral vein was exposed through a 3 cm linear incision. The profunda femoris vein was ligated and divided. As much adventitia as possible was removed from the femoral vein just distal to the profunda branch, taking care not to perforate the vein. An approximating clamp was applied, and a 1 mm segment of vein was removed and inverted. This segment was reanastomosed by microsurgical technique by using 10 sutures per anastomosis of 10-0 monofilament nylon suture on a 75 #m needle (MS/9; Ethieon, Inc., Somerville, N.J.). The polymer, poly(dimer erucic acid-cosebaeic acid) 1:1, was prepared by melt condensation as previously described. 14 Heparin-loaded polyanhydride films were prepared by melt mixing heparin powder in the polymer. Four groups of rats were studied. In all groups a single preoperative dose of 400 units/kg heparin sodium was administered by subcutaneous injection immediately after induction of anesthesia. In the control groups
Fig. 1. Illustration of venous inversion graft with polymer positioned in three strips around vessel at anastomotic site.
(A and B) the segmental inversion graft was performed, and 1 to 2 mg of 0.12 mm polyanhydride film without heparin was laid in three strips around the vessel at the anastomotic site (Figs. 1 and 2). In the experimental groups (C and D) a similar amount of 0.10 mm polyanhydride film containing 5 units/rag heparin (i.e., 5 to 10 units of heparin) was applied. Animals were randomized by assigning them alternately to group A, B, C, or D consecutively. The degree of systemic anticoagulation was assessed in each rat by measuring the partial thromboplastin time ( P T T ) in blood drawn from the left femoral vein 2 hours after the subcutaneous injection of heparin and again when the animal was killed. Rats in group A (n = 6) and group C (n = 7) were killed 24 hours after operation, whereas those in group B (n = t 1) and group D (n = 7) were killed at 7 days. Immediately before death the venous inversion graft was reexposed under general anesthesia. Flow was assessed by direct visualization through the operating microscope and then by a "milking test" cephalad to the anastomosis. Vessels were classified as either patent or thrombosed. Gross determination proved to be consistently accurate when confirmed by dividing the femoral vein cephalad to the inversion graft and observing free blood flow or lack thereof. Patency was essentially an
556
Surgery
Orloff et al.
May 7995
Fig. 2. Photograph of venous inversion graft with polyanhydride polymer in place (original magnification X 16).
t Fig. 3. Patent vein at 7 days after operation exhibits no inflammation (trichrome stain; original magnification X400).
all-or-nothing phenomenon; vessels were either fully patent or completely thrombosed. T h e vein inversion segment was excised and fixed in 10% buffered formalin. Specimens were embedded in paraffin, and representative 8 #m sections were stained with hemotoxylin-eosin and trichrome stains for histologic study by light microscopy.
RESULTS Patency results are shown in the T a b l e . In the anastomoses studied at 24 hours, only one of six control yes-
sels was patent, whereas six of seven veins treated with transmurally released heparin were patent. This difference is statistically significant by Fisher's exact test (p < 0.02). In animals studied at 7 days, four of 11 control vessels were patent, whereas six of seven heparintreated veins remained open. This difference is also statistically significant by the same test (p < 0.05). No major complications occurred. Gross inflammation was absent or minimal in all specimens and was only seen around vessels with thrombosis and local venous stasis. No hematomas were present around the
Surgery Volume 117, Number 5
Orloff et al.
557
Table. Results
Group
n
Preoperative subcutaneous heparin
A B C D
6 11 7 7
Yes Yes Yes Yes
Heparinpolyanhydride polymer
Time
(0
No. patent
% patent
p vs control
P T T at death (sec)
No No Yes Yes
24 hr 7 days 24 hr 7 days
1 4 6 6
16 36 86 86
--<0.01 <0.05
38.6 29.0 34.7 24.1
venous anastomoses. Histologic analysis of patent veins by light microscopy showed walls that were not thickened and had little if any acute inflammatory infiltrate (Fig. 3). By contrast, thrombosed veins showed intramural polymorphonuclear leukocytes and intraluminal platelet and fibrinous debris (Fig. 4). Measurements of P T T confirmed the development of systemic anticoagulation in all animals after the initial preoperative subcutaneous dose of heparin. Mean P T T for all rats at t = 2 hours was 117.5 seconds. At the time of death the P T T had returned to a level that approximated the normal laboratory mean of 31 seconds. P T T was 38.6 seconds in control animals versus 34.7 seconds in study animals at 24 hours and 29 seconds in control animals versus 24.1 seconds in study animals at 7 days. This normalization of P T T showed that ongoing transmural release of heparin did not cause systemic anticoagulation. DISCUSSION The consequences of microvascular thrombosis are often catastrophic, resulting in tissue or flap necrosis from ischemia or venous eongestion. Even in experienced hands, thrombosis can result from alterations in blood flow, endothelial damage, or hypercoagulability. Blood flow is temporarily altered during performance of microvascular anastomoses, and unless adequate flow is restored promptly, thrombosis may occur almost immediately. Exposure of subendothelial elements such as collagen, elastin, and microfibrils 4 or foreign material such as suture at the anastomotic site stimulates platelet aggregation and may trigger the clotting cascade. Although absolute hypercoagulability is not often present, the interplay of stasis and endothelial damage may create a relatively thrombogenic or hypercoagulable setting. Heparin has long been used to offset the potential coagulation that follows vascular manipulation. Heparin is a mucopolysaccharide that prevents new fibrin formation and the subsequent incorporation of fibrin into a thrombus. Heparin binds the plasma protein antithrombin III and enhances the ability of antithrombin III to neutralize the activated serine protease clotting factors (XII, XI, IX, X). Heparin also reduces platelet adhesiveness and directly inhibits individual clotting
factors, such as Xa, at concentrations significantly lower than those at which interaction with antithrombin III occurs. 3
Bleeding is the major complication of systemic heparin therapy, with an incidence of 1% to 10%. 15 Thrombocytopenia, lipolysis, hypersensitivity reactions, rebound hypercoagulability, and bone disease are less common side effects of heparin therapy. 16 Although prophylactic systemic anticoagulation with heparin reduces the incidence of thrombosis, the risks are generally thought to outweigh the benefits in the routine clinical practice of microvascular surgery. Heparin has been shown to diffuse from the extravascular to the intravascular space within 30 minutes of application. Mayberg 17 identified radiolabeled heparin, released from locally applied polyvinyl alcohol, within the cytoplasm of the cells of the vessel wall within hours after application and found that it remained in high local concentrations for 24 to 48 hours. Experimental studies of thrombosis prevention require a reliable animal model in which the rate of thrombosis is consistently high. Simple transection and microvascular reanastomosis of blood vessels are not suitable for such studies because the thrombosis rate is low and prohibitive numbers of animals are required to show significant antithrombotic benefits. The arterial inversion graft model described by Greenberg et al. 3 was an important development that represented a reliable model of microvascular thrombosis. The inversion graft technique is very reproducible, and its thrombogenicity can be precisely controlled by varying the length of the inverted segment. The inverted segment exposes type III fibrillar collagen to the vessel lumen, thereby mimicking clinical situations in which strands of adventitia are trapped in the anastomosis. This adventitial collagen is a potent platelet activator. Because venous thromboses are more common than arterial thromboses, we chose to study a venous rather than arterial inversion graft. The thrombosis rate for this vein inversion graft, in the absence of any anticoagulant therapy, was found to be 100%. In fact, in our preliminary studies the venous inversion graft was so viciously thrombogenic that without a preoperative dose of heparin the graft thrombosed immediately in 10 of 10 rats, before the heparin-containing polymer could even
558
Orloff et al.
Surgery May 1995
Fig. 4. Thrombosed vein in group B at 7 days after operation exhibits intraluminal inflammation (trichrome stain; original magnification Xl00). be applied and its effect observed. To circumvent this initial thrombosis, a brief period of systemic anticoagulation was induced before operation to create a window in time in which to complete the anastomoses, confirm immediate anastomotic patency, and apply the polymer. The single preoperative subcutaneous dose of heparin allowed for an initial period of pateney, but vessels that received no further heparin later thrombosed. The purely temporary nature of the systemic anticoagulation was confirmed by measurement of P T T 2 hours after the preoperative heparin injection and at the time of death either 24 hours or 7 days later. All animals exhibited normalization of P T T by the time of death. Because both treated and control animals received preoperative subcutaneous heparin and because locally delivered heparin was not found to have any measurable systemic anticoagulant effect, any differences in final patency rates were ascribed to the heparin released locally from the polyanhydride-heparin polymer. We postulate that while the effect of the preoperative dose of systemic heparin was being lost, controlled release of heparin was occurring in an ongoing fashion, resulting in a sustained antithrombotic effect. Drugs are released from the matrices of polyanhydride polymer vehicles at the same rate at which the polymers' hydrocarbon backbone is hydrolyzed. Drug release can occur over minutes, hours, weeks, or months, depending on the hydrophilicity of the polymer within which the drug is embedded. The relatively hydrophilic polyanhydride polymer that we used had an effect almost immediately after its application, with an ongoing effect during the following week. Further studies may help to characterize the precise release kinetics of
this particular heparin-polyanhydride polymer, but we presume that the locally released heparin reached an effective level within several hours. It is during these initial hours that the risk of microvascular thrombosis is highest, especially in venous anastomoses. Continued heparin release at the anastomotic site by ongoing hydrolysis of the polyanhydride film provided a persistent antithrombotic effect. The polyanhydride polymer itself, like commonly used surgical materials such as absorbable cellulose film and absorbable cellulose sponge, is inert and biocompatible. Polyanhydrides are nonmutagenic and noncytotoxic and do not react with the drugs embedded in their matrices. Breakdown products of polyanhydrides inelude carboxylic acids, aldehydes, water, and Krebs cycle intermediates that are excreted through direct elimination or through conjugation with amino acids before elimination. Previous biocompatibility studies have documented a similar lack of tissue reactivity or toxicity.9 Furthermore, the lightweight translucent polymeric film had no gross adverse effects such as compression or distortion of the anastomotic site. Although the polyanhydride-heparin polymer did not cause a measurable difference in P T T , it was associated with a statistically significant difference in immediate and long-term patency rates. Treated animals had significantly higher patency rates, at both 24 hours and 7 days after operation. Interestingly, the efficacy of the heparin is attributable to an average of only 6.5 units (1.3 mg of film, with 5 units heparin/mg of film.) Given such a small dose of heparin, the difference in patency rates in the polymer-treated groups is best explained by a highly localized and efficient effect of controlled-
Surgery Volume 777, Number 5
Orloff et al.
release heparin. A l t h o u g h both passive diffusion and active transport m e c h a n i s m s of t r a n s m u r a l h e p a r i n delivery can be postulated, the exact contribution by either of these m e c h a n i s m s is as yet unproven. Similarly, w h e t h e r the h e p a r i n passed t h r o u g h the intact vessel wall tissues or between tissues t h r o u g h the anastomotic site is uncertain. Investigations of the m o d e of t r a n s p o r t and the site of action of locally released h e p a r i n form the basis of future studies w i t h p o l y a n h y d r i d e - h e p a r i n . Because anastomotic patency rates w e r e significantly improved in the absence of any a p p a r e n t side effects, the use of this p o l y a n h y d r i d e - h e p a r i n p o l y m e r for further e x p e r i m e n t a l trials and for u l t i m a t e use in h u m a n beings is very promising. Indeed, t r a n s m u r a l p o l y a n h y d r i d e - h e p a r i n t h e r a p y m a y have clinical applications in vascular surgery involving large vessels, as well as in microvascular surgery. REFERENCES 1. Shaw WW, Converse JM. A survey of 2,680 free flaps: survival, donor site and applications among experienced microvascular surgeons-proceedings of the Fiftieth Annual Convention of the American Society of Plastic and Reconstructive Surgery. Plast Surg Forum 1981;4:93. 2. Davies DM. A world survey of anticoagulation practice in clinical microvascular surgery. Br J Plast Surg 1982;35:96-9. 3. Greenberg BM, Masem M, May JW. Therapeutic value of intravenous heparin in microvascular surgery: an experimental vascular thrombosis study. Plast Reconstr Surg 1988;82: 463-9. 4. Baumgartner HR. Platelet interaction with vascular structures. Thrombos Diathes Haemorrh 1972;51:161-76. 5. Elcock HW, Fredrickson JM. The effect of heparin on throm-
559
bosis at microvenous anastomatic sites. Arch Otolaryngol 1972; 95:68-71. 6. Jones NS, Glenn MG, Orloff LA, Mayberg MR. Prevention of microvascular thrombosis with controlled-release transmural heparin. Arch Otolaryngol Head Neck Surg 1990;116:779-85. 7. Deykin D. Current status of anticoagulant therapy. Am J Med 1982;72:659-64. 8. Domb AJ, Maniar M. Implantable biodegradable polymers as drug-carrier systems. In: Dumitriu S, ed. Polymeric biomaterials. New York: Marcel Dekker, 1994:399-433. 9. Brem H, Domb AJ, Lenartz D, Dureza C, Olivi A, Epstein JI. Brain biocompatibility of a biodegradable controlled release polymer consisting of anhydride copolymer of fatty acid dimer and sebacic acid. J Controlled Release 1992;19:325-30. 10. Chasin M, Domb AJ, Ron E, et al. Polyanhydrides as drug delivery systems. In: Langer R, Chasin M, eds. Polymers as drug delivery systems. New York: Marcel Dekker, 1990. 11. Langer R, Brown L, Edelman E. Controlled release of magnetically modulated release systems for macromolecules. Methods Enzymol 1988;142:399-422. 12. Kersh RA, Handren J, Hargreuter C, May JW. Microvascular surgical experimental thrombosis model: rationale and design. Plast Reconstr Surg 1989:83:866-74. 13. National Academy of Sciences. Guide for the care and use of laboratory animals. National Institutes of Health publication no. 80-23, revised 1978. 14. Domb AJ, Maniar M. Absorbable biopolymers derived from dimer fatty acids. J Polym Sci 1993;31:1275-85. 15. Hayden RE, McLear PW, Phillips JG, Dawson SM. Thrombolysis with systemically administered t-PA in a new venous thrombosis model. Laryngoscope 1989;99:100-4. 16. Okada T, Bark DH, Mayberg MR. Local anticoagulation without systemic effect using a polymer heparin delivery system. Stroke 1988;19:12-8. 17. Mayberg MR. Controlled release of perivascular heparin. In: Barrow D, ed. Perspectives in neurological surgery. St Louis: Quality Medical Publishing, 1990:77-95.
AVAILABILITY OF JOURNAL BACK ISSUES As a service to our subscribers, copies of back issues of SURGERY for the preceding 5 years are maintained and are available for purchase from the publisher, Mosby, at a cost of $8.50 per issue. The following quantity discounts are available: 25% off on quantities of 12 to 23, and one third off on quantities of 24 or more. Please write to Mosby-Year Book, Inc., Subscription Services, 11830 Westline Industrial Drive, St. Louis M O 63146-3318, or call (800)453-4351 or (314)453-4351 for information on availability of particular issues. If unavailable from the publisher, photocopies of complete issues are available from University Microfilms International, 300 N. Zeeb Rd., Ann Arbor, M I 48106 (313)761-4700.