Comparison of the Bleeding Potential of Vampire Bat Salivary Plasminogen Activator Versus Tissue Plasminogen Activator in an Experimental Rabbit Model
Background Vampire bat salivary plasminogen activator (Bat-PA) has significantly greater fibrin specificity than any of the fibrinolytic agents currently in clinical use. This study tests the hypothesis that avoiding fibrinogen depletion may protect against the hemorrhage induced by plasminogen activator treatment.
Methods and Results Bat-PA was compared with tissue-type plasminogen activator (TPA) in a randomized, prospective, and blinded study using a rabbit ear puncture model of fibrinolytic bleeding. The two agents were used at equimolar dosages (42 nmol/kg) that yielded similar thrombolytic efficacies in a rabbit femoral artery thrombosis model. Both Bat-PA and TPA prolong primary bleeding to double the baseline values, from between 2.1 and 2.3 minutes to between 4.8 and 5.2 minutes. Rebleeding from hemostatically stable sites during the 3-hour observation period occurred equally often with Bat-PA and TPA, 31% from preinjection sites and 23% to 25% from postinjection sites. The lag time between the time of plasminogen activator injection and the onset of rebleeding was likewise the same for both agents, most occurring at 41 to 57 minutes. However, a greater number of prolonged primary or rebleeding occurrences continued for longer than 10 minutes (63% versus 36%) or longer than 30 minutes (30% versus 10%) after Bat-PA than TPA injection. Animals treated with TPA showed a dramatic decrease in plasma fibrinogen and factor VIII concentrations, but those in the Bat-PA treatment group showed only a slight decrease from control values.
Conclusions The results indicate that fibrinolytic bleeding after plasminogen activator infusion into rabbits did not correlate with the intensity of the plasma proteolytic state. If anything, Bat-PA usage was associated with a higher proportion of more protracted fibrinolytic bleeding episodes, despite the relatively mild lytic state in comparison with that induced by TPA.
Plasminogen activators such as streptokinase, urokinase, prourokinase, tissue-type plasminogen activator (TPA), and anistreplase all promote beneficial thrombolytic revascularization but at the risk of detrimental fibrinolytic hemorrhage.1 Attempts to improve the efficacy of this class of antithrombotic agents, either by maximizing the rate and/or degree of clot lysis or by minimizing the occurrence or severity of bleeding, has led to the development of a variety of novel plasminogen activators. These new agents include recombinant mutants of TPA and urokinase,2 chemically synthesized hybrid activator species,3 and monoclonal antibody fusions that direct activators to specific substrates.4 While some of these novel plasminogen activator species exhibit improved biochemical properties of longer plasma half-life or more potent thrombolysis, none have yet been shown to provoke less bleeding. It has been postulated that selectivity for fibrin-bound plasminogen (“fibrin specificity”), manifest as plasma fibrinogen sparing such as that exhibited by TPA relative to streptokinase, urokinase, or anistreplase, would decrease hemorrhagic potential.5 This hypothetical safety advantage of TPA has not been realized in controlled, blinded trials,1 6 7 perhaps because the high dosage of TPA required to effect rapid thrombolysis elicits greater systemic proteolysis than anticipated and thereby reduces or negates the advantage of fibrin specificity.8 9 The availability of vampire bat salivary plasminogen activator (Bat-PA),10 11 12 which has significantly greater fibrin specificity than TPA,13 allows for more definitive testing of the hypothesis that avoiding fibrinogen degradation may protect against fibrinolytic-associated bleeding. Bat-PA promotes clot lysis equivalent to TPA but with minimal consumption of fibrinogen, plasminogen, and α2-antiplasmin.14 15
In a prior report that compared Bat-PA and TPA for thrombolytic effect in a rabbit femoral artery thrombus model, the primary template bleeding time in the ear was prolonged equally after bolus injection of equimolar Bat-PA or TPA.16 Nevertheless, the authors speculated that the avoidance of plasminemia and resultant fibrinogen degradation “could result in fewer and less severe bleeding complications” when encountered in a clinical trial. This prediction was seemingly supported by the results of a subsequent bleeding study in which TPA, but not Bat-PA, provoked severe bleeding from transected rabbit cuticles.17 Other in vivo evaluations of Bat-PA have emphasized thrombolytic potency without fibrinogen degradation but have not specifically explored hemorrhagic potential.18 19 In the present study, we extend the assessment of the potential bleeding risk of Bat-PA and TPA using an established rabbit ear puncture model of fibrinolytic hemorrhage.20 This animal model measures the effect of antihemostatic agents on the primary bleeding time and provides information on the incidence and duration of fibrinolytic bleeding from hemostatically stable trauma sites. Observations in this study may provide insight into the potential for bleeding in patients by use of a thrombolytic agent with strict fibrin specificity.
The experimental system used the rabbit ear puncture model based on the “rebleed phenomenon.”20 Twelve New Zealand White rabbits (Hazelton Research Products) weighing 2.3 to 3.1 kg were randomized to the three treatment groups in a blinded, prospective manner (see below). The experimental protocol was approved by the University Committee on Animal Research. The animals were euthanatized at the end of the portion of the experiment in which they were under anesthesia.
Before administration of anesthesia, the skin of both ears and of the neck was shaved and a depilatory (Nair, Carter-Wallace, Inc) was applied. A single, 3.5-mm full-thickness puncture was made in each ear of the unanesthetized rabbit with a number 11 surgical blade (Bard-Parker, Becton Dickinson AcuteCare) at 30, 21, 5, and 3 hours before initiation of plasminogen activator or vehicle injection. The duration of bleeding was measured by absorption of blood onto 11-cm circular filter paper (Whatman) at 30-second intervals, carefully avoiding disruption of or trauma to the bleeding site.
Animals were anesthetized with 40 mg/kg of ketamine HCl (Ketaset, Aveco Co) and 5 mg/kg xylazine (Rompun, Mobay Corp) by intramuscular injection, 1.6 to 2.2 mL total volume injected. After anesthesia was achieved, the left external jugular vein was exposed surgically, catheterized with a 30-cm 7F three-lumen catheter (Arrow), and attached to syringes containing saline, anesthesia mixture, and experimental infusion. Anesthesia was maintained by intermittent injections of the ketamine HCl and xylazine mixture approximately every 3 to 10 minutes as required. Saline was infused to keep one port open for blood sampling, the first sample being obtained just before initiation of plasminogen activator or vehicle injection. After initiation of anesthesia, two additional bleeding times were performed at 1 hour and 0.25 hour before plasminogen activator or vehicle injection.
The plasminogen activator injections were performed manually as 42 nmol/kg body wt (3 to 4.5 mL) given over 1 to 1.5 minutes. Each rabbit received 1 of 12 coded samples prospectively randomized and blinded to the observer. Of the 12 infusates, 4 contained TPA (Activase, Genentech), 4 contained Bat-PA (Merck Research Laboratories), 2 contained vehicle for TPA (0.32 mol/L arginine, 0.2 mol/L phosphoric acid, 0.01% Tween 80), and 2 contained vehicle for Bat-PA (50 mmol/L sodium acetate, 0.01% Triton X-100, pH 5.0). All infusates were prepared at Merck Research Laboratories, stored at −70°C, and thawed at room temperature 30 minutes before injection. While the primary focus of this study was a comparison of bleeding induced by Bat-PA versus TPA, the vehicle solutions were included as controls to confirm our assumption that the plasminogen activators were the only ingredients of the administered material that induced bleeding.
After injection of plasminogen activator or vehicle, bilateral ear bleeding times were performed at 0.25, 0.5, 1, 1.5, 2, and 3 hours. During the 3-hour postinjection observation period, all lesions were carefully monitored for rebleeding that occurred at previously clotted sites. Additional blood samples were collected from the central line at 0.5, 2, and 3 hours after initiation of the infusion. Of the 12 animals that were injected, 3 died spontaneously near or at the end of the planned observation period, 2 at 2.5 hours after injection and 1 at 3 hours. Of these 3 animals, 2 had received Bat-PA and 1 had received TPA; none showed excessive bleeding as the cause of death. The remaining animals were euthanatized at 3 hours with an intravenous injection of 1 mL (24 mg) of sodium pentobarbital (Socumb, Butler).
The blood collection line of the triple-lumen catheter was cleared of saline, after which 5 mL of whole blood was collected, added to a plastic tube containing 50 μL of 40% sodium citrate, and mixed thoroughly by inversion. Samples were placed on crushed ice until centrifugation at 3000 rpm for 10 minutes at 4°C (Beckman GS-6R centrifuge). Plasma was withdrawn with a 3.5-mL bulb-draw plastic transfer pipette (Laboratory Products Sales). Two aliquots each of 100 and 300 μL were added to 2-mL plastic screw-cap microfuge tubes (Laboratory Products Sales) containing 50 μL of d-Phe-Pro-Arg-CH2Cl (PPACK, Calbiochem) at a final concentration of 10 μmol/L. Plasma samples were then stored at −20°C.
Fibrinogen concentration was performed by a modification of the method of Clauss21 (Fibriquik, Organon Teknika Corp), using fibrinogen-calibrated plasma (Verify, Organon Teknika). Frozen plasma samples were shipped to the laboratory of S.J.G. for factor VIII determinations by use of the Coatest VIII:C/4 kit (Kabi Vitrum AB).
After completion of the studies, data were grouped according to type of plasminogen activator or vehicle infused and analyzed without knowledge of the specific agent or vehicle infused. Four animals with entirely normal primary bleeding times and not a single incident of rebleeding were assumed to represent those that received either of the vehicles and were not included in subsequent statistical analysis. Data for the remaining two groups of animals were analyzed by χ2 and t test for comparison of means.
The primary bleeding times and rebleeding occurrences are shown schematically in Fig 1⇓. No rebleeding occurred at ear trauma sites that were incurred at 21 or 30 hours before administration of Bat-PA, TPA, or vehicle. Therefore, these data were not included in the comparison of Bat-PA versus TPA.
Primary Bleeding Time
The mean duration of 32 preinjection primary bleeding times was the same for rabbits receiving Bat-PA and TPA (2.5 and 2.1 minutes, P=.17) (Fig 1⇑ and Table 1⇓). The mean of 48 postinjection primary bleeding times was prolonged to approximately double the preinjection values with both Bat-PA and TPA (5.2 versus 2.5 minutes, P=.06, and 4.8 versus 2.1 minutes, P=.0002, respectively). Bat-PA and TPA also induced similar frequencies of prolonged (>6 minutes) postinjection primary bleeding times, 16.7% versus 20.8%, P=NS. In addition, the proportions of bleeding times greater than 10 or 30 minutes were not significantly different after Bat-PA or TPA infusion (Table 1⇓).
Lag time indicates the elapsed period between cessation of primary bleeding and the onset of spontaneous rebleeding at any given trauma site. The lag time until rebleeding (minutes) was the same from puncture sites that were induced before or after activator infusion, 56.9±11 versus 44.7±10.7, P=.44, for Bat-PA and 46.0±10.1 versus 41.4±13.2, P=.79, for TPA (Table 1⇑). The comparison of Bat-PA and TPA showed no difference in lag time for either preinjection sites (56.9±11.2 versus 46.0±10.1, P=.48) or postinjection sites (44.7±10.7 versus 41.4±13.2, P=.85).
The overall incidence of rebleeding was the same for rabbits treated with Bat-PA or TPA (31.3% versus 31.3% and 25.0% versus 22.9% for preinjection and postinjection sites). However, there was a trend (not statistically significant) for rebleeding episodes induced by Bat-PA to persist longer than after TPA. The strongest trend was evident from the number of rebleeds at preinjection sites lasting for more than 10 minutes (25.0% for Bat-PA versus 9.4% for TPA, P=.19). In addition, there was a trend toward a longer mean duration of rebleeding (minutes) after Bat-PA at both preinjection sites (38.1±10.4 versus 21.6±9.1, P=.12) and postinjection sites (22.9±7.8 versus 13.5±4.9, P=.16).
There was no predisposition for trauma sites that exhibited prolonged (>6 minutes) primary bleeding times to manifest spontaneous rebleeding. For example, 16 of 23 rebleeding recurrences (70%) were at sites of normal primary bleeding times, compared with only 7 of 24 rebleed events (29%) at sites of a prolonged primary bleeding time.
Combined End Points for Excessive Bleeding
Excessive bleeding was arbitrarily scored as bleeding that continued for more than 10 minutes or more than 30 minutes (Table 2⇓). Combining the data for primary bleeding times and rebleeding durations showed a trend toward more prolonged bleeding after Bat-PA infusion compared with after TPA infusion. Bleeding episodes longer than 10 minutes occurred approximately twice as frequently in the Bat-PA treatment group (63.3%) as in the TPA treatment group (35.5%) (P=.06). Bleeding episodes longer than 30 minutes occurred approximately threefold more frequently in the Bat-PA treatment group (30.0%) than in the TPA treatment group (9.7%) (P=.09).
Plasma Fibrinogen and Factor VIII Concentrations
Fig 2⇓ shows the mean plasma fibrinogen and factor VIII concentrations after injection of vehicle, Bat-PA, or TPA. There was little change in the levels of both fibrinogen and factor VIII after vehicle injection. Bat-PA administration caused a modest decrease in the level of plasma fibrinogen to 70% of baseline at 3 hours. In marked contrast, plasma fibrinogen was decreased to unmeasurable levels at 30 minutes after TPA administration. A similar difference in the depletion of factor VIII occurred, with a more profound decrease in the TPA-treated rabbits.
The results of this study show that fibrinolytic bleeding occurs after Bat-PA injection in a rabbit ear puncture model20 despite the presence of only mild systemic proteolysis. Although TPA administration caused more marked decreases in plasma fibrinogen and factor VIII concentrations than did Bat-PA (Fig 2⇑), the primary bleeding times were equally prolonged to double baseline values, as was the number of sites that showed prolonged bleeding to more than 6, 10, or 30 minutes, approximately 17% to 21%, 8.4%, and 0% to 2.1%, respectively (Table 1⇑). In addition, there was no difference in the lag time between injection of TPA or Bat-PA and the onset of rebleeding, either from preinjection sites (46 and 56.9 minutes, respectively) or postinjection puncture sites (41.4 and 44.7 minutes, respectively). Furthermore, the overall incidence of rebleeding was the same for TPA and Bat-PA, 31.3% of preinjection sites for both agents and 22.9% and 25% of postinjection sites, respectively. There was no predictability of rebleeding on the basis of the duration of the primary bleeding time, as suggested by prior work.22 Thus, about two thirds of rebleeding occurrences were at sites of normal primary bleeding times, and only one third of primary bleeding times that were prolonged demonstrated a rebleeding episode. The only distinction noted was a higher incidence of prolonged bleeding in rabbits injected with Bat-PA. Although the number of observations was limited, there was a trend showing more frequent bleeding episodes (both primary and rebleed) lasting for longer than 10 or 30 minutes in the Bat-PA treatment group (Table 2⇑).
Our finding that fibrinolytic bleeding was induced by Bat-PA, despite the virtual absence of the plasma proteolytic state, supports the long-held concept that systemic proteolysis is not imperative for bleeding complications.23 24 According to this explanation, fibrinolytic-induced bleeding occurs principally because of hemostatic plug disintegration and is independent of the extent of activation of the “fluid phase” plasminogen. Typically, a poor correlation between plasma fibrinogen and bleeding has been observed in patients receiving different plasminogen activators and in different clinical situations.9 23 24 25 However, some workers have suggested that a more profound hypofibrinogenemia predisposes to greater bleeding.26 27 28
A reasonable explanation for the apparently more protracted bleeding state in rabbits after Bat-PA than TPA administration is that Bat-PA is cleared more slowly from the circulation after intravenous administration.16 The continued exposure of nascent fibrin strands to the highly fibrin-specific Bat-PA at puncture sites may markedly impair hemostasis despite the absence of significant systemic proteolysis.
The potential bleeding risk due to Bat-PA administration has been tested with other in vivo bleeding models. Intravenous bolus administration of equimolar Bat-PA and TPA (42 nmol/kg) provoked similar prolongations of the template bleeding times in aspirin-pretreated rabbits.16 In contrast, TPA at 42 nmol/kg caused a marked prolongation of the cuticle bleeding time in rabbits, while equimolar Bat-PA was without effect.17 The present study used the same dosages of Bat-PA and TPA, but the results showed a strong trend toward more excessive bleeding after Bat-PA administration than TPA administration. The reason for the disparate effects of Bat-PA on hemostasis is obscure. One possibility to explain the difference between these bleeding models is that the transected cuticle blood vessel is substantially larger than those that are traumatized by a skin template bleeding time16 or after through-punctures of the rabbit ear20 29 and may depend more critically on plasma factor VIII concentration.
The present comparison of Bat-PA and TPA used equimolar dosages of plasminogen activators (42 nmol each), based on prior experiments in rabbits.16 However, in contrast to TPA, Bat-PA was shown to display similar thrombolytic efficacy with a threefold lower dose using a rabbit femoral artery thrombosis model.16 The greater potency of Bat-PA relative to TPA has also been shown with a rat pulmonary embolism18 and a canine femoral artery19 model. Although a lower dosage of Bat-PA could potentially produce a lesser hemorrhagic effect while maintaining thrombolytic efficacy, the data clearly document that a mild lytic state does not translate directly to less intense hemorrhagic manifestations.
In summary, Bat-PA induces a significantly lesser degree of hypofibrinogenemia than an equimolar dosage of TPA in the rabbit, but fibrinolytic bleeding is at least as evident and perhaps more protracted. These results may not be predictive of experience in patients who receive equivalent or lower dosages of Bat-PA. However, the data do suggest that Bat-PA and other plasminogen activators that display fibrin specificity will cause hemorrhagic complications similar to those that occur with plasminogen activators in current use for human disease.
This work was supported in part by grant HL-30616 from the National Heart, Lung, and Blood Institute, National Institutes of Health, and by a Robert I. Weed Hematology Fellowship from the University of Rochester (Dr Montoney). The authors thank Carol Weed for her help in the preparation of the manuscript and Michael McDermott, PhD, for assistance in the statistical analysis.
- Received July 11, 1994.
- Accepted September 5, 1994.
- Copyright © 1995 by American Heart Association
Haber E, Quertermous T, Matsueda GR, Runge MS. Innovative approaches to plasminogen activator therapy. Science. 1989; 243:51-56.
Collen D, Verstraete M. Systemic thrombolytic therapy of acute myocardial infarction. Circulation. 1983;68:462-465.
The TIMI Study Group. The Thrombolysis in Myocardial Infarction (TIMI) trial: phase I findings. N Engl J Med. 1985; 312:932-936.
ISIS-3 (Third International Study of Infarct Survival) Collaborative Group. ISIS-3: a randomised comparison of streptokinase vs tissue plasminogen activator vs anistreplase and of aspirin plus heparin vs aspirin alone among 41,299 cases of suspected acute myocardial infarction. Lancet. 1993;339:753-770.
Topol EJ, Bell WE, Weissfeldt ML. Coronary thrombolysis with recombinant tissue-type plasminogen activator: a hematologic and pharmacologic study. Ann Intern Med. 1985;103:837-843.
Rao AK, Pratt C, Berke A, Jaffe A, Ockene I, Schreiber TL, Bell WR, Knatterud G, Robertson TL, Terrin ML. Thrombolysis in Myocardial Infarction (TIMI) Trial - Phase I: hemorrhagic manifestations and changes in plasma fibrinogen and the fibrinolytic system in patients treated with recombinant tissue plasminogen activator and streptokinase. J Am Coll Cardiol. 1988;11:1-11.
Gardell SJ, Duong LT, Diehl RE, York JD, Hare TR, Register RB, Jacobs JW, Dixon RAF, Friedman PA. Isolation, characterization, and cDNA cloning of a vampire bat salivary plasminogen activator. J Biol Chem. 1989;264:17947-17952.
Kratzschmar J, Haendler B, Bringmann P, Dinter H, Hess H, Donner P, Schleuning WD. High-level secretion of the four salivary plasminogen activators from the vampire bat Desmodus rotundus by stably transfected baby hamster kidney cells. Gene. 1992; 116:281-284.
Bergum PW, Gardell SJ. Vampire bat salivary plasminogen activator exhibits a strict and fastidious requirement for polymeric fibrin as its cofactor, unlike human tissue-type plasminogen activator: a kinetic analysis. J Biol Chem. 1992;267:17726-17731.
Gardell SJ, Hare TR, Bergum PW, Cuca GC, O’Neill-Palladino L, Zavodny SM. Vampire bat salivary plasminogen activator is quiescent in human plasma in the absence of fibrin unlike human tissue plasminogen activator. Blood. 1990;76:2560-2564.
Gardell SJ, Ramjit DR, Stabilito II, Fujita T, Lynch JJ, Cuca GC, Jain D, Wang S, Tung J, Mark GE, Shebuski RJ. Effective thrombolysis without marked plasminemia after bolus intravenous administration of vampire bat salivary plasminogen activator in rabbits. Circulation. 1991;84:244-253.
Mellott MJ, Ramjit DR, Stabilito II, Hare TR, Senderak ET, Lynch JJ Jr, Gardell SJ. Vampire bat salivary plasminogen activator evokes minimal bleeding relative to tissue-type plasminogen activator, as assessed by a rabbit cuticle bleeding time model. Blood. In press.
Witt W, Baldus B, Bringmann P, Cashion L, Donner P, Schleuning W-D. Thrombolytic properties of Desmodus rotundus (vampire bat) salivary plasminogen activator in experimental pulmonary embolism in rats. Blood. 1992;79:1213-1217.
Mellott MJ, Stabilito II, Holahan MA, Cuca GC, Wang S, Li P, Barrett JS, Lynch JJ, Gardell SJ. Vampire bat salivary plasminogen activator promotes rapid and sustained reperfusion without concomitant systemic plasminogen activation in a canine model of arterial thrombosis. Arterioscler Thromb. 1992;12:212-221.
Garabedian HD, Gold HK, Leinbach RC, Svizzero TA, Finkelstein DM, Guerrero JL, Collen D. Bleeding time prolongation and bleeding during infusion of recombinant tissue-type plasminogen activator in dogs: potentiation by aspirin and reversal with aprotinin. J Am Coll Cardiol. 1991;17:1213-1222.
The Urokinase Pulmonary Embolism Trial: a national cooperative study. Circulation. 1973;47(suppl II):II-1-II-108.
Marder VJ. The use of thrombolytic agents: choice of patient, drug administration, laboratory monitoring. Ann Intern Med. 1979;90:802-808.
Duckert F, Müller G, Nyman D, Benz A, Prisender S, Madar G, Da Silva MA, Widmer LK, Schmitt HE. Treatment of deep vein thrombosis with streptokinase. Br Med J. 1975;1:479-481.
Arnold AER, Brower RW, Collen D, van Es G-A, Lubsen J, Serruys PW, Simoons ML, Verstraete M for the European Cooperative Study Group for RT-PA. Increased serum levels of fibrinogen degradation products due to treatment with recombinant tissue-type plasminogen activator for acute myocardial infarction are related to bleeding complications, but not to coronary patency. J Am Coll Cardiol. 1989;14:581-588.
Agnelli G, Buchanan MR, Fernandez F, Boneu B, Van Ryn J, Hirsh J. A comparison of the thrombolytic and hemorrhagic effects of tissue-type plasminogen activator and streptokinase in rabbits. Circulation. 1985;72:178-182.