Orally Effective CVS-1123 Prevents Coronary Artery Thrombosis in the Conscious Dog
Background We examined the oral efficacy of a direct thrombin inhibitor, CVS-1123 [(CH3CH2CH2)2-CH-CO-Asp(OCH3)-Pro-Arg-CHO; MW, 575]. The object was to determine whether thrombin inhibition could reduce the incidence of occlusive coronary artery thrombosis in response to arterial wall injury.
Methods and Results Arterial wall injury was induced in conscious dogs by a 150-μA anodal current applied to the intimal surface of the circumflex coronary artery 30 minutes after oral CVS-1123 (20 mg/kg every 8 hours for three doses; n=11) or placebo containing diluent (n=10). Dogs were monitored for 8 hours and at 24 hours. The coronary artery remained patent for 24 hours in 8 of 11 CVS-1123–treated dogs. All dogs (n=10) in the placebo group developed a sustained, occlusive arterial thrombus. Two hours after the initial oral dose, the plasma CVS-1123 concentration was 13±1 μg/mL, reaching a maximum of 15±1 μg/mL after the second dose and 4.4±0.5 μg/mL at 24 hours. Ex vivo platelet aggregation to γ-thrombin was inhibited and activated partial thromboplastin time was increased after treatment with CVS-1123 (P<.05).
Conclusions The direct thrombin inhibitor CVS-1123 is effective after oral administration in reducing the incidence of primary thrombus formation in an experimental model of arterial wall injury. Thrombin-specific inhibitors, such as CVS-1123, may be alternative antithrombotic agents in clinical settings in which heparin-associated thrombosis is a complicating factor or when long-term anticoagulation is required.
The glycosylated, trypsin-like serine protease α-thrombin occupies a central role in the coagulation cascade and other host defense mechanisms.1 2 3 Thrombin is recognized for its ability to catalyze the formation of insoluble fibrin from fibrinogen4 and is regarded primarily as the most potent agonist for platelet aggregation5 and mediator for a variety of cellular processes.6 7 Furthermore, the action of thrombin on the vascular endothelium is accompanied by the synthesis and release of prostacyclin,8 platelet activating factor,9 plasminogen activator inhibitor,10 and platelet-derived growth factor.11 The multiple physiological actions of thrombin, therefore, make it an ideal target for modulation of thrombus formation. Current antithrombotic therapy relies on the use of heparin and coumarin derivatives that indirectly and incompletely inhibit the procoagulant actions of thrombin.
The recent development of direct-acting thrombin inhibitors has presented challenging opportunities for regulation of altered hemostasis associated with the thrombotic process.12 13 The reported advantages of direct thrombin inhibitors include lack of dependence on plasma cofactors (eg, antithrombin), low potential for plasma protein binding, and most important, the ability to inhibit fibrin-bound thrombin.13
Hirudin, a specific polypeptide inhibitor of thrombin isolated from the leech (Hirudo medicinalis), does not require antithrombin for its anticoagulant activity and is able to inhibit thrombus-bound thrombin.13 A synthetic derivative of hirudin, hirulog, is likewise capable of blocking both the active catalytic and anion-binding sites of the thrombin molecule.13 Unlike hirudin, which has multiple binding sites for thrombin, hirulog demonstrates substrate binding to thrombin, which is more typical of other protease inhibitors.13 d-Phenylalanyl-l-prolyl-l-arginyl chloromethylketone (PPACK) irreversibly inhibits thrombin by alkylating histidine at the active site and is an effective thrombin inhibitor in plasma and in vivo (microvessels) but possesses a relatively rapid loss of activity.14 Argatroban, an arginine derivative, binds thrombin at a region adjacent to the catalytic triad within the active site, thus blocking the active catalytic site of thrombin13 and thereby acting as a potent thrombin inhibitor in vivo, with a moderate affinity for thrombin.15 Despite the development of a number of new pharmacological agents for directly modulating the actions of thrombin, one major disadvantage is the lack of oral bioavailability and efficacy and relatively short durations of action, resulting in the need to give many of the compounds by continuous or repeated parenteral administration.
CVS-1123, a synthetic peptidomimetic (MW, 575), is a kinetically slow, competitive inhibitor of the amidolytic activity of thrombin as well as a potent anticoagulant in plasma in vitro. CVS-1123 has been shown in an anesthetized porcine model to be effective in preventing arterial thrombus formation.16 The primary objective of the present investigation was to assess the oral antithrombotic efficacy of CVS-1123 in a conscious canine model of coronary artery thrombosis by use of an experimental protocol extending to 24 hours. The results of this investigation provide evidence that CVS-1123 is an orally effective agent capable of preventing primary thrombus formation in an experimental model of deep arterial wall injury.
Assessment of Thrombin Inhibition by CVS-1123
The assessment of CVS-1123 as a selective inhibitor of the catalytic activity of thrombin was determined in vitro by evaluation of the concentration required to inhibit the amidolytic activity of the enzyme by 50% (IC50) and comparison of this value with that determined for the following related human serine proteases: rTPA, plasmin, aPC, fXa, and fXIa.
The buffer used for all assays was HBSA (10 mmol/L HEPES, pH 7.5/150 mmol/L sodium chloride/0.1% BSA). The assay for IC50 determinations was conducted by combination, in appropriate wells of a microtiter plate (Corning), of 50 μL HBSA and 50 μL CVS-1123 at a specified concentration (covering a broad concentration range) diluted in HBSA [or HBSA alone for Vo (uninhibited velocity) measurement] and 50 μL of the enzyme diluted in HBSA. After a 30-minute incubation at ambient temperature, 50 μL of the substrate at the concentrations specified below was added to the wells, yielding a final total volume of 200 μL. The initial velocity (Vi) of chromogenic substrate hydrolysis was measured by the change in absorbance at 405 nm on a Thermo Max Kinetic Microplate Reader over a 5-minute period in which <5% of the added substrate was utilized. The concentration of added inhibitor that caused a 50% decrease in the ratio of Vi/Vo was defined as the IC50 value.
The catalytic activity of thrombin was determined with a chromogenic substrate, CH3SO2-d-hexahydrotyrosine-glycyl-l-arginine-p-nitroaniline (Pefachrome t-PA, Pentapharm Ltd), diluted in HBSA to a final concentration of 250 μmol/L (about 5 times Km). Purified human α-thrombin (Enzyme Research Laboratories, Inc) was diluted in HBSA to a final concentration of 0.50 nmol/L. The catalytic activity of rTPA was determined by use of the substrate Pefachrome t-PA. The substrate was dissolved in deionized water followed by dilution in HBSA before the assay in which the final concentration was 500 μmol/L (about 3 times Km). Human rTPA (Activase) was obtained from Genentech Inc. The enzyme was diluted in HBSA to a final concentration of 1.0 nmol/L. The catalytic activity of plasmin was determined by use of the chromogenic substrate S-2366 (l-pyroglutamyl-l-prolyl-l-arginine-p-nitroaniline, Kabi Diagnostica) diluted in HBSA to a final concentration of 300 μmol/L (about 2.5 times Km). Purified human plasmin (Enzyme Research Laboratories, Inc) was diluted in HBSA (1.0 nmol/L). The catalytic activity of aPC was determined by use of the chromogenic substrate Pefachrome PC (d-carbobenzyloxy-d-lysine-l-prolyl-l-arginine-p-nitroaniline, Pentapharm Ltd) in HBSA at a final concentration of 250 μmol/L (about 3 times Km). Purified human aPC (Hematologic Technologies, Inc) was diluted in HBSA to a final concentration of 1.0 nmol/L. The chromogenic substrate S-2765 (N-α-benzyloxycarbonyl-d-arginyl-l-glycyl-l-arginine-p-nitroaniline, Kabi Diagnostica) was used to assess the catalytic activity of fXa. The substrate was diluted in HBSA to a final concentration of 250 μmol/L (about 3 times Km). Purified human fXa was prepared from fresh-frozen plasma as described by Bock et al.17 The enzyme was diluted into HBSA to a final concentration of 0.25 nmol/L. The catalytic activity of purified human fXIa (Enzyme Research Laboratories, Inc) was determined by use of the chromogenic substrate S-2366. The substrate was diluted in HBSA before the assay to a final concentration of 750 μmol/L (about 2.5 times Km). The enzyme was diluted in HBSA to a final concentration of 0.25 nmol/L.
Guidelines for Animal Research
The procedures followed in this study were in accordance with the guidelines of the University of Michigan (Ann Arbor) University Committee on the Use and Care of Animals. Veterinary care was provided by the University of Michigan Unit for Laboratory Animal Medicine. The University of Michigan is accredited by the American Association of Accreditation of Laboratory Animal Care, and the animal care and use program conforms to the standards set forth in the Guide for Care and Use of Laboratory Animals, Department of Health, Education, and Welfare publication No. NIH 78-23.
Male mongrel dogs were anesthetized with sodium pentobarbital 30 mg/kg IV, intubated, and ventilated with room air delivered under positive pressure at a stroke volume of 30 mL/kg and a frequency of 12 breaths per minute (Harvard Apparatus). Under aseptic technique, a left thoracotomy was performed at the level of the fourth intercostal space. The heart was exposed and suspended in a pericardial cradle. The LCx was isolated by blunt dissection. A ligature stenosis was formed by application of a ligature around an 18-gauge needle and the LCx, followed by removal of the needle. The LCx was instrumented with a Doppler flow probe (model 100, Triton Technology) and an intraluminal electrode. The intracoronary electrode consisted of a 30-gauge, Teflon-insulated, silver-coated copper wire attached to a 25-gauge hypodermic needle tip and inserted through the wall of the LCx so that the uninsulated needle tip was positioned firmly against the endothelial surface of the vessel. The external portion of the stimulating electrode and the flow probe wires were secured to the myocardium by a suture. The thoracotomy incision was closed, and the Doppler flow probe leads and intracoronary electrode wires were exteriorized posteriorly in the midscapular region. Intrathoracic air and blood were evacuated, and the surgical wound was dressed.
Cannulas were placed in the left carotid artery to monitor arterial blood pressure (Statham P23 ID pressure transducer, Gould) and in the jugular vein to administer intravenous fluids and to obtain blood samples. All cannulas were tunneled under the skin and exteriorized on the dorsal surface of the neck. Leads for recording a lead II ECG were placed under the skin. A nylon jacket (Alice King Chatam Medical Arts) was placed on the animal to protect the surgical wound site as well as all external leads. After full recovery from anesthesia, the animals were returned to their quarters and given daily injections of ampicillin suspension (200 mg SC). The animals were allowed 3 to 5 days to recover from the surgical procedure before they were assigned randomly to the experimental protocol.
On entering the laboratory, dogs were placed in a quiet environment and allowed to rest unrestrained. Blood was withdrawn from the jugular vein for platelet count, ex vivo platelet aggregation studies, and determination of the aPTT. Dogs were randomized to one of two treatment protocols in which the control group received a gelatin capsule containing lactose diluent (placebo) and the drug-treated group was given a gelatin capsule containing CVS-1123. The amount of drug to be administered was preweighed, according to the animal's body weight, and placed in the capsule with sufficient lactose diluent as the excipient.
The study protocol is shown in Fig 1⇓. Arterial blood pressure, the ECG lead II, and the mean and phasic LCx blood flow velocities were recorded on a model 7 polygraph recorder (Grass Instrument). Thirty minutes after administration of drug (CVS-1123, 20 mg/kg PO) or placebo (lactose diluent), a 150-μA continuous, anodal, direct current was applied to the endothelial surface of the LCx through the implanted electrode. The anodal current was delivered from a Grass stimulator (model S88) and isolation unit (model SIU5). The electrical circuit was completed by placement of the cathode in a subcutaneous site. The anodal current to the intimal surface of the coronary artery was maintained for 3 hours. Time to arterial occlusion was monitored, as determined by the Doppler flow signal. The animals were given a second and third dose of their respective treatments at 4 hours and 8 hours. The total dose of CVS-1123 was 60 mg/kg in three divided doses separated by intervals of 4 hours.
All animals were monitored continuously for 8 hours, at which time they were returned to the postoperative recovery facilities until 24-hour measurements were obtained. On completion of the protocol, all animals were euthanatized with sodium pentobarbital, the chest was opened, and the heart was removed. The LCx was dissected free and opened longitudinally. The placement of the intracoronary anodal electrode was verified, and the presence of a deep arterial intimal lesion was identified by gross inspection of the vessel. The thrombus was removed by gentle extraction and weighed on an analytical balance. The heart was sectioned transversely, apex to base, in sections 1.0 cm thick that were incubated without agitation in triphenyltetrazolium chloride for 5 minutes at 37°C. The transverse ventricular sections were weighed and traced onto clear acetate sheets. The red-pigmented tissue containing the precipitated formazan complex represented viable tissue, whereas tissue that remained pallid demarcated irreversibly injured or infarcted myocardium. The demarcated areas were traced onto acetate sheets, which were scanned with a flatbed scanner and digitized with a Macintosh IIci microcomputer (Apple Computer) and appropriate software (MacDraft, Innovative Data Design) that aided in calculating the areas of the respective regions on both faces of each transverse ventricular section. Infarct size was expressed as a percentage of total left ventricular area.
Platelet Studies and Coagulation Measurements
Whole blood (20 mL) was withdrawn from the jugular vein to be used for platelet studies and the determination of the aPTT. The blood was collected in plastic syringes containing 3.7% sodium citrate as the anticoagulant (1:10 citrate/blood, vol/vol) at baseline and at 1, 2, 4, 6, 8, and 24 hours. The platelet count was determined with an H-10 cell counter (Texas International Laboratories). PRP, the supernatant present after centrifugation of anticoagulated whole blood at 1000 rpm for 10 minutes, was used for aggregation studies. PPP was prepared after the PRP was removed by centrifugation of the remaining blood at 3000 rpm for 10 minutes and the bottom cellular layer was discarded. Ex vivo platelet aggregation was assessed by established spectrophotometric methods with a four-channel aggregometer (BioData-PAP-4) by recording of the increase in light transmission through a stirred suspension of PRP (adjusted to 200×103 platelets/μL) maintained at 37°C. Platelet aggregation was induced with arachidonic acid (0.65 and 0.325 mmol/L), ADP (20 and 5 μmol/L), or γ-thrombin (70 nmol/L). A subaggregatory dose of epinephrine (550 nmol/L) was used to prime the platelets before the agonists were introduced. Values are expressed as percent aggregation, which are represented by the fraction of light transmission standardized to PPP samples yielding 100% light transmission.
To assess the anticoagulation state of the animals, the aPTT was determined with a Hemochron (Technidyne) with reagents supplied by the manufacturer. Citrated whole blood was used for these determinations.
Quantification of Plasma CVS-1123 Concentrations
Citrated plasma samples were frozen at −20°C and subsequently used for determination of the plasma concentration of CVS-1123. An eight-point standard curve was constructed by spiking 0.1 mL of blank dog plasma with 0.035 to 1.035 μg of CVS-1123 (by peptide content). An internal standard was added to each sample to adjust for recovery. The plasma samples were extracted by use of 4-mm Empore C18 solid-phase extraction columns (3M) and washed with 1 mL H2O, followed by 0.15 mL of 50% methanol/50% H2O. The CVS-1123 and the internal standard were eluted with 0.15 mL of 99.8% methanol/0.02% trifluoroacetic acid. The eluate was dried under vacuum and reconstituted with 0.1 mL of 10% acetonitrile/90% H2O/0.1% trifluoroacetic acid. Samples and standards were injected, followed by separation with an isocratic mobile phase (40% methanol/60% pH 6 potassium phosphate buffer, 60°C) pumped at 0.3 mL/min through a 10-cm×2-mm C18 HPLC column. The addition of 0.6 mol/L KOH (0.15 mL/min) and 0.125% ninhydrin (0.05 mL/min) to the mobile phase was done by postcolumn derivatization using mixing tees and a 1.0-mL reaction coil (0.01-in diameter) at 60°C before detection in a fluorescence detector (excitation wavelength, 390 nm; emission wavelength, 500 nm). The mean relative recovery and RSD for the three standard curves run with these samples were recovery, 103% and %RSD, 11.4% for day 1; recovery, 96% and %RSD, 13.6% for day 2; and recovery, 99% and %RSD, 5.4% for day 3.
Time Course for the Pharmacological Action of CVS-1123
Four dogs that had not been subjected to surgical intervention were fasted overnight before receiving CVS-1123 by the oral route in a dosing regimen identical to that of the study group. Three doses of CVS-1123, 20 mg/kg each, were administered to the conscious, unsedated, fasted animal at 4-hour intervals. Venous blood samples were collected at baseline and at 1, 2, 4, 6, 8, 24, and 48 hours after oral drug administration. Platelet aggregations in response to ADP (20 μmol/L), arachidonic acid (0.65 mmol/L), and γ-thrombin (70 mmol/L) were conducted on PRP prepared from the venous blood samples. Determination of aPTT was made at identical time points before and after oral drug administration.
The data are expressed as mean±SEM. The data were analyzed by one-way ANOVA for group comparisons and for repeated measures followed by a Dunnett post hoc t test to determine the level of significance. Incidence of occlusion was compared by Fisher's exact test. Values were considered to be statistically different at a level of P<.05.
Inhibition of α-Thrombin Amidolytic Activity and Specificity of CVS-1123
The in vitro potency and selectivity of CVS-1123 were determined in a defined biochemical system using purified human enzymes and the corresponding amidolytic substrates for α-thrombin, fXa and fXIa, protein C, plasmin, and rTPA. The potency of CVS-1123 as defined by the IC50 value for inhibition against thrombin was 1.13 nmol/L, with at least a 227-fold selectivity compared with the other serine proteases involved in hemostasis and fibrinolysis (Table 1⇓). The data indicate that CVS-1123 is a potent inhibitor of thrombin activity suitable for evaluation of its role as an inhibitor in the thrombotic process in an in vivo model.
A total of 36 dogs, weighing 13±1 kg, were entered into the study. Ten animals were allotted randomly to the placebo-treated group; 11 animals were treated with CVS-1123 (3×20 mg/kg PO). Randomization was done by selection of coded cards on the day of the study at the time the instrumented animals were brought into the laboratory. Eleven animals were excluded because of technical problems or the presence of heartworms (found on postmortem examination). Four additional dogs were used to assess the pharmacological duration of action of CVS-1123 as determined by a study of ex vivo platelet reactivity.
aPTT and Ex Vivo Platelet Aggregation
The aPTT did not differ between groups at baseline (Table 2⇓). In the presence of CVS-1123, the aPTT was prolonged significantly from baseline values. The aPTT increased ≈10-fold from a baseline value of 24±0 to 238±29 seconds (P<.05) 1 hour after the oral administration of CVS-1123. The aPTT returned to baseline by the end of the protocol at 24 hours. Animals in the placebo-treated group (n=10) did not exhibit a change from baseline values for aPTT throughout the course of the study. Ex vivo platelet aggregations in response to arachidonic acid (0.65 mmol/L), ADP (20 μmol/L), and γ-thrombin (70 nmol/L) are shown in Table 2⇓. There were no differences in the aggregation status between groups throughout the protocol to either arachidonic acid or ADP. Responses induced by 0.325 mmol/L arachidonic acid and 5 μmol/L ADP were identical to those to 0.65 mmol/L arachidonic acid and 20 μmol/L ADP, respectively (data not shown). However, ex vivo platelet aggregation in response to γ-thrombin was inhibited significantly within 1 hour after the oral administration of CVS-1123. The ex vivo platelet responses to γ-thrombin were attenuated throughout the experimental protocol compared with baseline values. Most notable was the reduction in platelet reactivity to γ-thrombin up to 16 hours after the last oral dose of CVS-1123. Whole blood cell counts were not altered in either group throughout the experimental protocol (Table 3⇓).
Heart Rate and Mean Arterial Pressure
The administration of placebo or CVS-1123 did not directly affect heart rate (Table 4⇓). The progressive decrease in blood pressure during the first 6 hours of the protocol in both groups of animals was the result of acclimatization to the laboratory environment during the 8-hour observation period as the animals rested quietly. It should be noted that there was a statistically significant reduction in mean arterial pressure in the CVS-1123–treated animals at 4, 6, and 8 hours compared with baseline values. However, we are unable to explain the cause of this reduction in arterial pressure. The arterial pressure measured in both groups of animals returned to baseline values at the 24-hour time point.
Coronary Artery Blood Flow Velocity
The consequence of electrolytic vessel wall injury and subsequent thrombus formation on LCx blood flow velocity is shown in Fig 2⇓. Each of the animals in the placebo group (n=10) exhibited a progressive reduction in LCx blood flow velocity, culminating in total occlusion, as evidenced by the loss of the Doppler flow signal. The mean time to LCx occlusion in the placebo-treated group was 114±28 minutes. Absence of the Doppler flow signal was noted when the animals were monitored 24 hours into the experimental protocol, which confirmed that neither spontaneous lysis nor dislodgment of the occlusive thrombus had occurred.
Of the 11 dogs treated with CVS-1123, 8 maintained patent LCxs throughout the entire protocol. In the three CVS-1123–treated animals that did undergo coronary occlusion, time to occlusion was 145±40 minutes, compared with placebo-treated animals, 114±28 minutes (n=10) (P=.59).
Posterolateral Infarct Size and Ventricular Fibrillation
Left ventricular infarct size, expressed as a percentage of the total left ventricle, was determined in the 8 placebo-treated and 3 CVS-1123–treated animals that failed to maintain a Doppler flow velocity signal, indicating thrombotic occlusion of the LCx. Left ventricular infarct size in the placebo-treated group was 21±5% (n=8) of the left ventricle, compared with 21±9% (n=3) in the CVS-1123–treated group. Thrombus weight determined at 24 hours was reduced significantly in animals that had been treated with CVS-1123 compared with placebo treatment: 1.8±1.5 (n=11) versus 24±5 mg (n=10), respectively (P<.05).
The patency status of individual dogs is depicted in Fig 3⇓. Coronary vessels in 9 of 10 dogs in the placebo control group occluded within 4 hours from initiation of the injury current. In the placebo-treated group, 2 of the 10 animals died of ventricular fibrillation and 2 died of low-output failure after LCx occlusion. The remaining animals survived for the entire experimental protocol despite evidence of a significant area of infarction in the region of distribution of the LCx determined on postmortem examination. In the CVS-1123 treatment group, only 1 of the 11 animals experienced ventricular fibrillation; the remaining 2 animals in which the LCx occluded survived for 24 hours. Histochemical evidence of myocardial infarction, as determined by the reduction of triphenyltetrazolium chloride, was not detectable in any of the 8 animals in which patency of the LCx was maintained.
Plasma Concentrations of CVS-1123
Plasma concentrations of CVS-1123 were determined by reverse-phase HPLC on blood specimens collected at preselected times throughout the investigation. Fig 4⇓ depicts the mean profile of CVS-1123 in the plasma of treated animals. Two hours after the initial oral dose (t=2 hours), the plasma concentration was 13±1 μg/mL. Two hours after the second dose of CVS-1123 (t=6 hours), the maximum measured concentration of the drug in the plasma was 15±1 μg/mL. Plasma samples obtained 24 hours after the initial dose (16 hours after the third oral dose) had a concentration of CVS-1123 that averaged 4.4±0.5 μg/mL.
Endothelial damage initiates a response to injury involving platelet activation characterized by localized vasoconstriction18 and cyclic flow variations in stenosed and endothelium-injured vessels, culminating in occlusive thrombus formation.19 Exposure of subendothelial collagen fibrils serves as an agonist for platelet activation and release of additional agonists including ADP, thromboxane, and serotonin.20 Concomitant with the activation and adhesion of platelets, there is initiation of the coagulation cascade and generation of thrombin at the site of vessel wall injury. Thrombin is the catalyst for conversion of fibrinogen to fibrin. In addition, thrombin is a potent agonist for platelet activation.21 The response to the agonistic potential of a deep vascular wall lesion differs from that occurring in response to normal hemostasis in injured but otherwise healthy vessels. Pathological or occlusive arterial thrombi that form in response to deep arterial wall injury may have the morphological features of a hemostatic plug but involve more platelet adhesion, activation, recruitment, and consolidation.21 Development of pathological arterial thrombi is the result of multiple factors required for induction, assembly, and stabilization of the occlusive mass at the site of vessel wall injury. The difficulty in preventing occlusive arterial lesions may be related to the multifactorial nature of the arterial thrombotic process.22
In this investigation, we examined the in vivo antithrombotic activity of an orally administered, slow-onset, direct thrombin inhibitor, CVS-1123.16 CVS-1123 is relatively specific for the inhibition of thrombin activity with good selectivity versus other serine proteases involved in hemostasis or fibrinolysis, including fXa, aPC, plasmin, and rTPA (Table 1⇑). The high specificity for thrombin inhibition against plasmin and rTPA suggests that CVS-1123 may be compatible with fibrinolytic agents, in contrast to other peptide inhibitors of thrombin, which directly counteract the action of one or more fibrinolytic agents.23
The experimental model used to evaluate the oral efficacy of CVS-1123 consisted of the chronically instrumented, conscious dog, in which deep vessel wall injury of the LCx occurs in response to the local application of a 150-μA anodal current.24 In the absence of effective antithrombotic therapy, an occlusive, platelet-rich arterial thrombus develops at the site of vessel injury. Formation of an occlusive lesion is prevented by inhibition of the platelet glycoprotein IIb/IIIa receptor25 26 or, once formed, is susceptible to lysis by systemic thrombolytic therapy.22 The experimental model permits one to examine a potential therapeutic intervention in the anesthetized or conscious animal. Furthermore, the model provides the opportunity to administer the test agent by several routes and to continue the observation for an extended duration with or without repeated drug administration, as demonstrated by a variety of investigators.22 25 26 27 28 29 30
Currently, the only clinically available antithrombotic agents are heparin, low-molecular-weight heparin, and coumarins, each of which have known limitations.4 31 Even though heparin is commonly used as an antithrombin, it is relatively ineffective in preventing arterial thrombosis.32 Thrombocytopenia, bleeding complications, and repeated laboratory monitoring associated with the use of heparin also detract from its complete acceptance for clinical use. The principal antithrombotic action of heparin requires the presence of cofactors, ie, antithrombin III. Only after heparin is complexed with antithrombin III is it able to express its antithrombotic action. However, a heparin–antithrombin III complex is not able to inactivate thrombin that is bound to a thrombus.33 In contrast, clot-bound thrombin is susceptible to inactivation by antithrombin III–independent inhibitors, because the sites of their interaction are not masked by thrombin binding to fibrin.33 The suggestion has been offered that antithrombin III–independent inhibitors may be more effective than heparin in preventing thrombus formation,33 thereby providing the impetus to search for direct thrombin inhibitors.
CVS-1123 inhibited the aggregation of washed human platelets only when α-thrombin was used as an agonist. There was no effect of the compound on ADP- or collagen-induced platelet aggregation.16 Similar observations were made in the present study, in which CVS-1123 did not prevent ex vivo platelet aggregation in response to either ADP or arachidonic acid. However, γ-thrombin–induced ex vivo platelet aggregation was inhibited within 1 hour of oral administration of CVS-1123. A significant degree of inhibition of platelet reactivity to γ-thrombin was observed 16 hours after the last oral dose of CVS-1123 and at a time when the drug remained present in the circulation.
Rote et al16 evaluated the in vivo antithrombotic efficacy of CVS-1123 in models of arterial thrombosis in pigs and rats. Before the application of current to the intimal surface of the left anterior descending coronary artery, the pigs received either saline, heparin, or CVS-1123 (10 mg/kg intraduodenal). The left anterior descending coronary artery of saline-treated pigs occluded in 30±7 minutes, compared with CVS-1123–treated pigs, in which time to occlusion was 137±20 minutes. In our investigation, we demonstrate that CVS-1123 prevented thrombotic occlusion in 8 of 11 dogs, compared with 0 of 10 dogs in the placebo-treated group. This implicates the pivotal role of thrombin in the present model of arterial thrombosis.
This particular model of electrolytic deep arterial wall injury was developed in our laboratory24 and has been used to explore mechanisms of thrombosis as well as rethrombosis.25 This model can be used in both conscious and anesthetized preparations. The design of this particular study included a conscious dog, since the primary aim of the investigation was to determine the oral efficacy of CVS-1123. A particular advantage of CVS-1123 includes its ability to affect aPTT values without altering ex vivo platelet aggregation in response to either arachidonic acid or ADP. Platelet aggregation to each agonist was unaltered in animals treated with CVS-1123 compared with the aggregation profiles from placebo-treated animals. As expected, CVS-1123 significantly reduced the ex vivo platelet aggregation when γ-thrombin was used as the agonist, a result that is in keeping with the action of CVS-1123 as an active-site inhibitor of thrombin.
We did not observe any untoward effects (spontaneous bleeding from the recent surgical wound site, vomiting, diarrhea, agitation, etc) in any of the dogs treated with CVS-1123. Although only one dosing regimen was used in this investigation, selection of the dose was based on preliminary studies conducted in the anesthetized animal, thereby obviating the need to engage in costly dose-ranging studies in the chronically instrumented animal. Results from the present study suggest that CVS-1123 (3×20 mg/kg PO) is an orally effective antithrombotic agent that may offer a therapeutic alternative to current antithrombins that require parenteral administration. It is unknown at this time whether lower plasma concentrations of CVS-1123 are capable of preventing primary arterial thrombus formation or whether CVS-1123 would be effective in preventing arterial rethrombosis after thrombolytic therapy. It should be noted that the plasma concentration achieved after oral administration of CVS-1123 (15±1 μg/mL) exceeded the concentration shown to inhibit fXa amidolytic activity (Table 1⇑). Therefore, we cannot say with certainty that the antithrombotic effects of CVS-1123 observed in this model were exclusively a result of direct thrombin inhibition. We can conclude, however, that an orally effective, direct-acting anticoagulant, such as CVS-1123, may serve as an alternative antithrombotic agent and is deserving of further evaluation.
Selected Abbreviations and Acronyms
|aPC||=||activated protein C|
|aPTT||=||activated partial thromboplastin time|
|HPLC||=||high-performance liquid chromatography|
|LCx||=||left circumflex coronary artery|
|RSD||=||relative standard deviation|
|rTPA||=||recombinant tissue plasminogen activator|
This study was supported by National Institutes of Health, National Heart, Lung, and Blood Institute, grant HL-19782-16. The authors wish to thank the following persons from Corvas International for their respective contributions to this work: Susanne Anderson and Peter Bergum for performing the in vitro inhibition studies and Robert Ardecky, Pam Leon, and Yu Ge for providing the CVS-1123.
- Received December 20, 1995.
- Revision received April 1, 1996.
- Accepted April 9, 1996.
- Copyright © 1996 by American Heart Association
Badimon L, Badimon JJ, Lassila R, Heras M, Chesebro JH, Fuster V. Thrombin regulation of platelet interaction with damaged vessel wall and isolated collagen type I at arterial flow conditions in a porcine model: effects of hirudins, heparin and calcium chelation. Blood. 1991;78:423-434.
Coughlin SR. Thrombin receptor function and cardiovascular disease. Trends Cardiovasc Med. 1994;4:77-83.
Coughlin SR, Vu T-KH, Hung DT, Wheaton VI. Characterization of the cloned platelet thrombin receptor: issues and opportunities. J Clin Invest. 1992;89:351-355.
Weksler BB, Ley CW, Jaffe EA. Stimulation of endothelial cell prostacyclin by thrombin, trypsin, and the ionophore A23187. J Clin Invest. 1982;62:923-930.
Prescott SM, Zimmerman GA, McIntyre TM. Human endothelial cells in culture produce platelet-activating factor when stimulated by thrombin. Proc Natl Acad Sci U S A. 1984;81:3534-3538.
Daniel TO, Gibbs VC, Milfay DF, Garavoy M, Williams LT. Thrombin stimulates c-sis gene expression in microvascular endothelial cells. J Biol Chem. 1986;261:9579-9582.
Lefkovits J, Topol EJ. Direct thrombin inhibitors in cardiovascular medicine. Circulation. 1994;90:1522-1536.
Ikoma H, Ohtsu K, Tamao Y, Kikumoto R, Okamoto S. Effect of a potent thrombin inhibitor, MCI-9038, on novel experimental arterial thrombosis. Blood Vessels. 1982;13:72-77.
Rote WR, Dempsey EM, Oldeschulte GL, Ardecky RJ, Tamura SY, Nutt RF, Bergum PW, Tran HS, Anderson SM, Vlasuk GP. Evaluation of a novel orally active direct inhibitor of thrombin in animal models of thrombosis. Circulation. 1994;90(pt 2):I-344. Abstract.
Marcus AL. Platelets and their disorders. In: Ratnoff OD, Forbes CD, eds. Disorders of Hemostasis. 2nd ed. New York, NY: WB Saunders Co; 1991:75-140.
Yao SK, Ober JC, Gonenne A, Clubb FJ, Krishnaswami A, Ferguson JJ, Anderson HV, Gorecki M, Buja LM, Willerson JT. Active oxygen species play a role in mediating platelet aggregation and cyclic flow variations in severely stenosed and endothelium-injured coronary arteries. Circ Res. 1993;73:952-967.
Kroll MH, Schafer AI. Biochemical mechanisms of platelet activation. Blood. 1989;74:1181-1195.
Marcus AJ, Safier LB. Thromboregulation: multicellular modulation of platelet reactivity in hemostasis and thrombosis. FASEB J. 1993;7:516-522.
Rote WE, Nedelman MA, Mu DX, Manley PJ, Weisman H, Cunningham MR, Roncinske RA, Frelinger AL III, Lucchesi BR. Chimeric 7E3 prevents carotid artery thrombosis in cynomolgus monkeys. Stroke. 1994;25:1223-1232.
Nicolini FA, Lee P, Rios G, Kottke-Marchant K, Topol EJ. Combination of platelet fibrinogen receptor antagonist and direct thrombin inhibitor at low doses markedly improves thrombolysis. Circulation. 1994;89:1802-1809.
Shebuski RJ, Stabilito IJ, Sitko GR, Polokoff MH. Acceleration of recombinant tissue-type plasminogen activator–induced thrombolysis and prevention of reocclusion by the combination of heparin and the Arg-Gly-Asp–containing peptide bitistatin in a canine model of coronary thrombosis. Circulation. 1990;82:169-177.
Lynch JJ Jr, Sitko GR, Mellott MJ, Nutt EM, Lehman ED, Friedman PA, Dunwiddie CT, Vlasuk GP. Maintenance of canine coronary artery patency following thrombolysis with front loaded plus low dose maintenance conjunctive therapy: a comparison of factor Xa versus thrombin inhibition. Cardiovasc Res. 1994;28:78-85.
Scharfstein JS, Loscalzo J. Molecular approaches to antithrombotic therapy. Hosp Pract. 1992;27:77-86.
Yasuda T, Gold HK, Fallon JT, Leinbach RC, Guerrero JL, Scudder LE, Kanke M, Shealy D, Ross MJ, Collen D, Coller BS. Monoclonal antibody against the platelet glycoprotein (GP) IIb/IIIa receptor prevents coronary artery reocclusion after reperfusion with recombinant tissue-type plasminogen activator in dogs. J Clin Invest. 1988;81:1284-1291.
Weitz JI, Hudoba M, Massel D, Maraganore J, Hirsh J. Clot-bound thrombin is protected from inhibition by heparin-antithrombin III-independent inhibitors. J Clin Invest. 1990;86:385-391.