Abstract
The antithrombotic activity of recombinant, human activated protein C (rh-APC, LY203638) was examined in a model of canine coronary artery thrombosis. Three doses of rh-APC (0.5, 1.0, and 2.0 mg/kg/h) were administered intravenously for 2 h. Whole blood clotting times (thrombin time, activated partial thromboplastin time), ex vivo platelet aggregation, and template bleeding times were determined. Activated partial thromboplastin time significantly increased 2- and 3.7-fold during the 2-h infusion of rh-APC (1.0 and 2.0 mg/kg/h, respectively); thrombin time did not change. Intravenous infusions of rh-APC (1.0 and 2.0 mg/kg/h) produced significant prolongations to occlusion, 186 ± 21 and 190 ± 22 min, respectively, compared with the vehicle and the 0.5 mg/kg/h group (86 ± 12 and 93 ± 17 min, respectively). Vessel patency was better at the end of the experiment in the intermediate- and high-dose groups (3 of 6 and 3 of 5 vessels, 1.0 and 2.0 mg/kg/h, respectively) compared with the vehicle and 0.5 mg/kg/h groups (0/5 and 0/6, respectively). Only the 1.0 mg/kg/h group was found to have significantly elevated template bleeding times, with peak increases seen 60 min into the drug infusion. All groups had returned to baseline values by the end of the study. There was no observed inhibition of platelet aggregation. These data demonstrate that recombinant, human activated protein C is an effective anticoagulant and antithrombotic agent in the dog.
Upon vascular injury both extrinsic and intrinsic coagulation cascades are activated, ultimately resulting in the formation of thrombin. Thrombin converts fibrinogen into fibrin and activates factor XIII to factor XIIIa, which promotes the cross-linking of this fibrin. Thrombin may also activate circulating platelets, causing the exposure of surface glycoprotein IIb-IIIa receptors that bind fibrinogen, thus promoting platelet aggregation and clot formation. Although being the mainstay for the treatment of thromboembolic disorders, heparin is not an ideal anticoagulant in the presence of either activated platelets or an existing thrombus mass. In addition to the procoagulant activity of thrombin, this proteolytic enzyme also has an anticoagulant property. Thrombin forms a bimolecular complex with thrombomodulin on the surface of endothelial cells and in so doing reduces its ability to promote clot formation and, at the same time, makes thrombin a potent activator of the zymogen, protein C (Esmon, 1987). Activated protein C is a serine protease that produces a potent endogenous anticoagulant activity by inactivating clotting factors Va and VIIIa (Marlar et al., 1982). These two clotting factors are essential for the function of two other coagulation proteases, IXa and Xa. Anticoagulant/antithrombotic studies have been conducted in animals with human plasma-derived protein C (Emerick et al., 1987; Gruber et al., 1989; Romisch et al., 1991; Sakamoto et al., 1994) and with genetically engineered, recombinant human protein C (Gruber et al., 1990; Hahn et al., 1996) activated in vitro with thrombin.
In the present study we examined the anticoagulant and antithrombotic pharmacology of rh-APC (LY203638) in a canine model of electrolytic injury-induced coronary artery thrombosis (Jackson et al., 1992). This model of arterial thrombosis involves the formation of a platelet/fibrin-rich thrombus (Romson et al., 1980).
Materials and Methods
Instrumentation.
The instrumentation has been described previously (Jackson et al., 1992). Briefly, 22 6- to 7-month-old mixed breed hounds of both sexes (15.7–22.2 kg, Butler Farms, Clyde, NY) were anesthetized with sodium pentobarbital (30 mg/kg i.v.) and ventilated with room air. The carotid artery, jugular vein, and femoral vein were isolated and cannulated for recording of arterial pressure (MPC-500 transducer, Millar, Houston, TX), blood sampling, and drug administration, respectively. Limb leads were placed subcutaneously for monitoring of a lead II ECG. A left thoracotomy was performed, and the heart was suspended in a pericardial cradle. The left circumflex coronary artery (LCCA) was isolated and cleaned of all adventitia and fat proximal to the first main branch. An electromagnetic flow probe was placed around the LCCA for the direct measurement of blood flow. A second MPC-500 Millar transducer was inserted into the ventricle via the left atria for the measurement of left ventricular pressure. A stimulating anode (26-gauge needle-tipped, 30-gauge silver-coated/copper wire) was inserted into the LCCA distal to the flow probe and in contact with the intimal surface of the artery. Placing the cathode in a subcutaneous site completed the circuit. A plastic screw occluder was placed around the LCCA over the area of the electrode, and a critical stenosis (sufficient to produce a 40–50% reduction in the hyperemic response to a 10-s total occlusion of the LCCA) was applied. All hemodynamic and ECG measurements were recorded and analyzed with a data acquisition system (M3000, Modular Instruments, Inc., Malvern, PA).
Thrombus Formation.
Thrombogenesis was initiated by applying 100 μA of direct current to the anode, producing endothelial cell injury. The current was maintained for 60 min and then stopped whether or not the vessel had occluded. Thrombus formation proceeded in a spontaneous manner until the LCCA was totally occluded (determined as zero coronary blood flow). In the experimental protocol, rh-APC or the vehicle was administered 15 min before anodal stimulation had begun. Drug was administered as a 2-h infusion and studied over the dose range [0.5 (n = 5), 1.0 (n = 6), and 2.0 (n = 5) mg/kg/h]. The vehicle-treated group (n = 6) received 20 ml of a solution containing 150 mM NaCl, 20 mM Tris-HCl, and 5 mM CaCl2 in sterile water (pH = 7.2) infused over a 2-h period. The animals were monitored for 2 h after cessation of drug infusion, at which time the animals were sacrificed by electrical ventricular fibrillation. A 2-cm segment of the LCCA containing the thrombus was removed and dissected longitudinally and the thrombus removed and weighed.
Coagulation Assay, Template Bleeding Time, and Hematology.
Whole blood samples (3.0 ml, nonanticoagulated) were drawn immediately before drug administration, 60 and 120 min during drug administration, and 60 and 120 min after cessation of drug. The clotting times, thrombin time, and APTT were determined using a Hemochron 801 (International Technidyne Corp., Edison, NJ) analyzer. Immediate assessment of coagulation was necessary because of the rapid inhibition of rh-APC by its physiological plasma protease inhibitors.
Template bleeding times were performed on the gingiva of the upper and lower jaw of the dogs using a Simplate II bleeding time device (Organon Tecknika, Durham, NC). Template bleeding times were determined immediately before drug administration, 60 and 120 min during drug administration, and 120 min after cessation of drug.
Whole blood cell counts and analysis of hemoglobin content were determined on a 40-μl citrated blood sample processed with a hematology analyzer (Cell-Dyn 900, Sequoia-Turner, Mountain View, CA). Samples were taken before drug administration, 60 min into the drug infusion, and 120 min after cessation of drug
Platelet Aggregation Tests.
In addition, ex vivo and in vitro platelet aggregation (Bio/Data, PAP-4, Hatboro, PA) was performed. Platelet-rich plasma (PRP) drawn immediately before drug administration was prepared from citrated blood samples (3.8%; 1:9 parts blood) using centrifugation at 150g for 10 min. Platelet aggregation (37°C) was induced by the addition of either thrombin (0.1–0.35 U/ml), PAF (0.3 μM), ADP (10 μM), or arachidonic acid (AA, 0.625 mM). ADP and AA-induced platelet aggregation responses were primed with a nonaggregating concentration of epinephrine (0.85 μM). In the in vitro experiments, 5 μg/ml rh-APC was added to normal citrated dog plasma 1 min before the agonist. All agonists and conditions were the same as those for the ex vivo experiments.
Drug and Data Analysis.
LY203638, the Lilly designation for rh-APC, was obtained as a lyophilized powder (lot 308EM1, in which rh-APC represented 17.9% of the powder and the remainder was a Tris salt; the specific activity of rh-APC was 300 U/mg). Thrombin (bovine) and ADP were obtained from Sigma Chemical Co. (St. Louis, MO). PAF was obtained from Bachem (Torrance, CA) and prepared according to the method of Jackson et al. (1986). All data were analyzed using ANOVA for group comparisons and for repeated measures followed by a Student-Newman-Keuls post hoc t test to determine the level of significance. Values were determined to be statistically different at least at the level of P < .05. All values represent mean ± S.E. All studies were conducted according to the “Guide for the Care and Use of Laboratory Animals” adopted by the National Institutes of Health.
Results
Anticoagulant and Antiplatelet Activity.
Figure1 illustrates that intravenous infusion of rh-APC produced a dose-dependent increase in APTT. Peak increases for each dose group were 1.6 ± 0.1-, 2.3 ± 0.1-, and 3.8 ± 0.2-fold above baseline and 0.5, 1.0, and 2.0 mg/kg/h rh-APC, respectively. All groups had returned to baseline values by the end of the experiment. Thrombin times were not affected at any dose studied (data not shown). Table 1illustrates that rh-APC had no significant effect on ADP, PAF, AA, or thrombin-induced ex vivo platelet aggregation. In addition, when 5.0 μg/ml rh-APC was added to normal dog plasma (n = 4), there was no significant effect on in vitro platelet aggregation in response to the same agonists (13.2 ± 3.1, 5.0 ± 3.2, 12.8 ± 1.0, and 2.9 ± 2.0% inhibition for ADP, PAF, AA, and thrombin, respectively).
Antithrombotic Efficacy.
Recombinant h-APC produced a significant prolongation in time to occlusion in both the 1.0 and 2.0 mg/kg/h dose groups compared with the vehicle and the 0.5 mg/kg/h treated groups (186 ± 21 and 190 ± 22 min versus 85 ± 12 and 93 ± 17 min, respectively; P < .05) (Fig.2). Quantification of coronary blood flow throughout the experiment demonstrated that the animals receiving 1.0 and 2.0 mg/kg/h rh-APC maintained significantly higher (P < .05) coronary blood flow than did the other groups examined (Fig. 3). There were also more patent vessels (P < .05) at the end of the experiment in the 1.0 and 2.0 mg/kg/h rh-APC groups (3 of 6 and 3 of 5, respectively) versus 0 of 6 (vehicle) and 0 of 5 (0.5 mg/kg/h rh-APC). The two higher dose levels of rh-APC examined produced significant decreases in thrombus mass compared with the vehicle-treated group (Fig. 4).
Effects on Template Bleeding Times and Canine Hematology, Blood Pressure, and Heart Rate.
There was a small, significant (P < .05) increase in template bleeding time during the 1.0 mg/kg/h rh-APC infusion; template bleeding time values had returned to predrug levels by the end of the 2-h washout period (Fig.5). None of the dose groups examined were associated with any effects on red blood cell count, hemoglobin, platelet counts (data not shown), or heart rate and blood pressures (Table 2).
Discussion
Despite the extensive use of heparin in the treatment of acute coronary syndromes, coronary thrombosis still remains the number one cause of morbidity and mortality in the United States today. In experimental animal models of thrombosis, more effective antithrombotic efficacy than that of heparin has been achieved with compounds that are active site inhibitors of thrombin [argatroban, hirudin, and GYKI-14766 (efegatran)] (Eidt et al., 1989; Yasuda et al., 1990;Jackson et al., 1992). Unlike heparin, these new site-specific compounds are not dependent on antithrombin III or heparin cofactor II, are not inactivated by platelet factor 4, and inhibit both free and clot-bound thrombin. The active site thrombin inhibitors do not inhibit thrombin generation. Recombinant human activated protein C, however, does reduce the generation of thrombin (Nicolaes et al., 1997). In this investigation, rh-APC was an effective anticoagulant and was antithrombotic in a canine model of coronary artery thrombosis, demonstrating that suppression of thrombin generation contributes to its antithrombotic activity.
Rh-APC produced a dose-dependent increase in APTT (Fig. 1). Thrombin clotting times were not affected at any of the dose levels examined. As a consequence of its anticoagulant activity, rh-APC demonstrated significant antithrombotic efficacy (i.e., time to occlusion and maintenance of coronary blood flow) at the two highest doses examined and maintained vessel patency in 6 of 11 dogs for the duration of the experiment (Figs. 2 and 3). Thrombi exposed to rh-APC during drug infusion were statistically smaller compared with those of the vehicle-treated group (Fig. 4). In consideration of the short plasma half-life of rh-APC of 8 to 12 min (Gruber et al., 1990), antithrombotic efficacy was limited by the 2-h infusion period in the experimental protocol. One would anticipate that longer infusions of rh-APC would maintain a sustained level of anticoagulation and allow for overall better antithrombotic efficacy. Similar observations were made by Gruber et al. (1989) in an arteriovenous shunt model of thrombosis in the baboon. In their investigation, plasma-derived, human activated protein C was infused for 60 min and demonstrated reductions in platelet deposition through a thrombin-mediated effect. When the drug infusion was stopped, there was a restoration of platelet deposition.
There are different species sensitivities in regard to the human anticoagulant activity of human APC [human > monkey > dog > rabbit > guinea pig > mouse > rat (Chabbat et al., 1991)]. The dose of rh-APC administered to dogs to double the APTT is approximately 15- to 20-fold higher than that needed to double the APTT in humans. The baboon has a 4 times more sensitive anticoagulant response than the dog [0.25 mg/kg/h versus 1.0 mg/kg/h to double the APTT, respectively (Gruber et al., 1990)]. One possible explanation is that APC requires protein S for maximal anticoagulant activity (Esmon, 1987). Bovine protein S was shown to enhance the antithrombotic activity of bovine protein C in a rabbit model of arterial thrombosis, demonstrating that rabbit protein S was not optimal for bovine protein C (Arnlijots and Dahlback, 1995). It is therefore highly likely that rh-APC does not interact or interacts poorly with dog protein S. Additionally, it may be that the lipid component of the endothelial and platelet membranes also contributes to the differences in species sensitivity for rh-APC, especially for the interaction between protein S and rh-APC (Esmon, 1987).
All dose levels of rh-APC examined were found to have no effect on thrombin-, PAF-, ADP-, or AA-induced ex vivo platelet aggregation (Table 1). This observation could be misleading because rh-APC has a short half-life in plasma such that by the time blood was drawn and centrifuged and PRP was prepared (45–60 min), more than half the functional rh-APC would be gone. Therefore we took four different dogs not exposed to rh-APC and made PRP identical to the preparations used for the ex vivo platelet aggregation test. To this PRP we added 5.0 μg/ml rh-APC and after 1 min of incubation immediately added the agonist used in the ex vivo platelet aggregation test. Table 1 shows that rh-APC has no direct effect on platelet aggregation. Therefore, by inhibiting thrombin generation, rh-APC minimizes the endogenous platelet aggregating effect of thrombin, indirectly reducing platelet aggregation.
Template bleeding times were minimally affected by the 2-h infusion of rh-APC. A small but significant increase was observed with 1.0 mg/kg/h rh-APC at 60 min into the infusion; however, just before the 2-h infusion was terminated template bleeding time was essentially normal (Fig. 5). Template bleeding time returned to baseline values by the end of the study. The irreversible thrombin inhibitord-phenylalanyl-l-prolyl-l-arginyl chloromethylketone (PPACK) has been shown to increase bleeding times in the baboon during infusions of minimally efficacious doses (Hanson and Harker, 1988; Krupski et al., 1990). PPACK was observed to produce increases in template bleeding time of 27 min and >30 min at doses of 1.6 and 2.7 mg/kg/h, respectively, compared with predrug values of 4 to 5 min. Antithrombotic efficacy of recombinant hirudin and argatroban was also associated with changes in template bleeding times (Heras et al., 1990; Jang et al., 1990). If the dose of these site-specific thrombin inhibitors is increased sufficiently, there is a risk of increased template bleeding time. Jackson et al. (1992) showed that in this same model of thrombosis heparin, at a minimally effective dose level (80 U/kg + 30 U/kg/h), produced an efficacy to a bleeding time ratio of 1:1.
The results of the present investigation provide additional support that recombinant, human activated protein C is an effective anticoagulant and antithrombotic agent. Recombinant, human activated protein C is presently in phase III clinical trials for the treatment of sepsis.
Acknowledgments
We thank Dr. Walt Prouty (Lilly Research Laboratories) for the generous supply of recombinant, human activated protein C. In addition, we thank Drs. Brian Grinnell, Betty Yan (Lilly Research Laboratories), and Nils Bang (retired Lilly physician) for support and constant encouragement during this work.
Footnotes
-
Send reprint requests to: Charles V. Jackson, Ph.D., Cardiovascular Research Division, Lilly Research Laboratories, Lilly Corporate Center—MC304, Indianapolis, IN 46285-0524. E-mail:jacksoncv{at}lilly.com
- Abbreviations:
- rh-APC
- recombinant human activated protein C, LY203638
- LCCA
- left circumflex coronary artery
- TBT
- template bleeding time
- APTT
- activated partial thromboplastin time
- PAF
- 1-O-hexadecyl-2-O-acetyl-sn-glycero-3-phosphorylcholine
- AA
- arachidonic acid
- PRP
- platelet-rich plasma
- Received April 17, 2000.
- Accepted July 19, 2000.
- The American Society for Pharmacology and Experimental Therapeutics