![[Feedback]](/icons/banner/feedbackACT.gif){kind=icon}
![[For Subscribers]](/icons/banner/subscriberACT.gif){kind=icon}
![[Archive]](/icons/banner/archiveACT.gif){kind=icon}
![[Search]](/icons/banner/searchACT.gif){kind=icon}
![[Contents]](/icons/banner/tocACT.gif){kind=icon}
|
|
|
|
|
||
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
CARDIOVASCULAR
Discovery Biology, Bristol-Myers Squibb Company, Pennington, New Jersey
Received for publication
December 22, 2006
Accepted
April 6, 2007.
 |
Abstract |
|---|
 Top Abstract Materials and MethodsResultsDiscussionReferences |
|---|
). The structure of the active metabolite and the identity of the platelet receptor for clopidogrel were not known before phase III clinical trials had been completed (CAPRIE Steering Committee, 1996). In fact, the P2Y12 receptor was referred to as the thienopyridine receptor in the original publication that detailed its molecular identification (Hollopeter et al., 2001). The primary activity biomarker for tracking the effects of clopidogrel has been ex vivo inhibition of ADP-induced platelet aggregation (IPA) (Nguyen et al., 2005).
It is well recognized that the standard 75-mg maintenance dose of clopidogrel that reduces clinical cardiovascular events exerts a submaximal effect on ex vivo IPA (Thebault et al., 1999). The antiplatelet effect of clopidogrel in man requires oral bioavailability of parent drug, its bioactivation to the active metabolite, and productive blockade by that metabolite of platelet P2Y12 receptors. Its effects will not prevent ADP-mediated platelet stimulation through P2Y1 receptors. Our goal was to understand whether IPA is in fact a good predictor of clopidogrel's efficacy and bleeding liability in an experimental animal setting. Aspirin was also studied as a comparator to better validate the experimental procedures employed.
The hypothesis that IPA is a robust biomarker of clopidogrel's antithrombotic activity was addressed in dose-response studies conducted in rats that were orally dosed with clopidogrel for 3 days. These rats were anesthetized and subjected to either an acute thrombosis or provoked bleeding procedure or used to measure ex vivo platelet function. Occlusive thrombosis was induced in the carotid artery by topical application of iron chloride, an oxidative injury widely used in experimental rodent models (Leadley et al., 2000). Provoked bleeding times were measured in response to mesenteric vessel puncture, incision of the renal cortex, or cuticle incision. These bleeding models were selected because they had been found to be sensitive to aspirin and represent bleeding responses in a variety of vascular tissues. In addition to IPA, P2Y12 receptor occupancy (Baurand et al., 2000) and phosphorylation of vasodilator-stimulated phosphoprotein (VASP) (Geiger et al., 1999) were studied in the hope of calibrating these sensitive biochemical indicators of clopidogrel activity.
|
Materials and Methods |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
) was synthesized by Sanofi-Aventis Company (Paris, France). Aspirin and MRS-2179 were obtained from Sigma (St. Louis, MO) and ADP from Chrono-Log Corp.(Havertown, PA). The vehicle for clopidogrel and aspirin was water (1 ml/kg). BMS-189664 (Das et al., 2002) was synthesized at Bristol-Myers Squibb (Pennington, NJ), and its vehicle was saline. General Treatment and Surgical Procedure. Each procedure was approved by the Bristol-Myers Squibb Institutional Animal Care and Use Committee and was conducted in accordance with the Guide for the Care and Use of Laboratory Animals as adopted by the U.S. National Institutes of Health. To study clopidogrel, male Sprague-Dawley rats (310-450 g) were dosed once daily for 3 days with a 1 ml/kg oral gavage of either vehicle or clopidogrel (0.3-50 mg/kg). These animals were not fasted. One hour after the final dose, they were anesthetized with Na-pentobarbital (50 mg/kg i.p.), and the trachea was intubated with polyethylene-205 tubing to ensure airway patency. Temperature was maintained with a warming table and/or heat lamp. A polyethylene-50 catheter was inserted into the left carotid artery in those animals selected for biomarker measurement. Other animals were subjected to the arterial thrombosis, venous thrombosis, cuticle bleeding time, mesenteric bleeding time, or renal cortex bleeding time procedure. Only one procedure was conducted on an animal, and none exceeded 90 min in total duration.
To study aspirin, the same surgical preparation was used except that aspirin or vehicle was administered as a 10 mg/kg infusion over 5 min into a polyethylene-50 jugular vein catheter at 10 to 15 min before inducing thrombosis or measuring bleeding time. The 10 mg/kg dose of aspirin was previously shown to cause maximal inhibition of platelet cyclooxygenase, as revealed by a >99% inhibition of thromboxane B2 generation during the clotting of rat blood (Schumacher et al., 1993).
Arterial Thrombosis. The right carotid artery was exposed and fitted with an electromagnetic flow probe attached to a model MDL 1401 flowmeter (Skalar, Delft, Netherlands) or a transit time Doppler probe attached to a T206 flowmeter (Transonic Systems Inc., Ithaca, NY). A piece of parafilm "M" (American National Can, Greenwich, CT) was inserted under the vessel and, following baseline flow measurements, a 2- x 5-mm strip of filter paper saturated with a 50% solution of FeCl2 was placed on top of the artery for 10 min. The carotid artery was dissected free 60 min after filter paper application and opened lengthwise to expose the thrombus, which was removed, blotted dry, and weighed on an AE50 balance (Mettler, Toledo, IN). Carotid blood flow was monitored continuously on a TA4000 physiologic recorder (Gould, Cleveland, OH). Total blood flow was determined as an area under the curve (AUC) and normalized as percentage of baseline (0 min) flow over 60 min to provide a measure of integrated blood flow during thrombus formation.
Treatment groups consisted of vehicle (n = 14) and clopidogrel doses of 0.3 (n = 9), 1 (n = 8), 3 (n = 8), 10 (n = 8), 20 (n = 8) or 30 (n = 5) mg/kg. In separate experiments, aspirin (10 mg/kg, n = 6) or vehicle (n = 5) were tested.
Mesenteric Bleeding Time. The abdomen was opened via a midline incision, and the small intestine was exteriorized. The jejunum was exposed, held in place with clamps, and superfused with Ringer's solution maintained at 37°C. Small arteries that branch perpendicular to the mesenteric artery and course over the surface of the jejunum were observed through an SZH10 stereomicroscope (Olympus Corp., Lake Success, NY). These vessels were punctured with a 30-gauge hypodermic needle, and the time in seconds from puncturing until bleeding stopped and remained stopped for 30 s was recorded. The maximal bleeding time recorded was 10 min. Bleeding times were determined in three to five arteries. In this and the other two bleeding models, replicate bleeding times were averaged to single values for statistical comparison, and the number of bleeds that exceeded the maximal cut-off were also recorded.
Treatment groups consisted of vehicle (n = 12) and clopidogrel doses of 1 (n = 7), 3 (n = 5), 10 (n = 7), or 20 (n = 7) mg/kg. Aspirin (10 mg/kg, n = 6) or vehicle (n = 6) were tested in separate experiments.
Cuticle Bleeding Time. Toenails were cut with a single-edged razor blade at the location where the quick meets the nail. The cuticle was immediately superfused with Ringer's solution maintained at 37°C, and the time until bleeding stopped and remained stopped for 30 s was recorded. The maximal bleeding time recorded was 15 min. Three replicate bleeding times were determined on the left hind paw.
Treatment groups consisted of vehicle (n = 12) and clopidogrel doses of 1 (n = 8), 3 (n = 8), 10 (n = 10), 20 (n = 10), or 30 (n = 6) mg/kg. Aspirin (10 mg/kg, n = 6) or vehicle (n = 6) were also tested.
Renal Cortex Bleeding Time. The left kidney was exposed by a midline abdominal incision, and the renal capsule was removed to expose the bare renal cortex. The cortex was superfused with Ringer's solution maintained at 37°C, and controlled incisions were made with a Surgicutt adult template device (International Technidyne Corp., Edison, NJ). This produced a 5-mm-long x 1-mm-deep cut with the tip of a spring-loaded surgical blade (25, sharp tip). While observing under 3x binocular magnification, the time in seconds from injury until bleeding stopped and remained stopped for 30 s was recorded. The maximal bleeding time recorded was 12 min.
Treatment groups consisted of vehicle (n = 5) and clopidogrel doses of 1 (n = 5),3(n = 5), 10 (n = 5), 20 (n = 5), or 50 (n = 6) mg/kg. Aspirin (10 mg/kg, n = 7) or vehicle (n = 6) were also tested.
When the bleeding effect of clopidogrel appeared limited in this model, the incision was increased to 3 mm deep x 10-mm long using the Surgicutt blade adapted to a manual template device. Treatment groups for this deep cut included 20 mg/kg clopidogrel (n = 6) and its vehicle (n = 6), and thrombin inhibitor BMS-189664 (2 mg/kg loading plus 6 mg/kg/h sustaining i.v. infusion, n = 7) and its vehicle (n = 7). BMS-189644 was also tested at the same dose with the standard incision (n = 5). BMS-189664 was found previously effective in the primate version of this model (Das et al., 2002), and the dose was selected to produce a 5.6-fold increase in the activated partial thromboplastin time. This represents a near-maximal systemic anticoagulant effect.
Preparation for Ex Vivo Biomarkers. A 5- to 8-ml blood sample was drawn from the carotid artery or in some cases from the abdominal vena cava into a plastic syringe containing a 1/10 volume of 3.8% sodium citrate. Whole-blood samples were used for VASP assay, whereas platelet-rich plasma (PRP) was prepared for platelet aggregation and receptor binding assays. Whole blood sampled for the receptor binding assay also contained apyrase (0.25 U/ml final concentration; Sigma catalog no. A-6132).
There were two separate studies. In the first study ex vivo platelet aggregation responses to 1, 3, and 10 µM ADP were determined in rats treated with either vehicle (n = 9) or clopidogrel at doses of 1 (n = 6), 3 (n = 5), 10 (n = 5), 20 (n = 5), or 30 (n = 5) mg/kg. In the second study, additional biomarkers of VASP phosphorylation and ligand binding were added to the measurement of 10 µM ADP-induced platelet aggregation using an expanded range of clopidogrel doses (0.3, 1, 3, 10, 20, 30, and 50 mg/kg with n = 6 per dose, including six vehicle-treated rats).
Platelet Aggregation. Platelet count in the PRP was determined on a System 9000 cell counter (Serono-Baker Diagnostics, Allentown, PA) and adjusted with platelet-poor plasma to obtain 
4.0 x 105 platelets/µl. Aggregation responses of PRP were measured by standard photometric technique at 37°C in a model 540 or 490 aggregometer (Chrono-Log Corp.). Light transmission through PRP and platelet-poor plasma was set at 0 and 100%, respectively. In the first study, maximal percent aggregation (peak response) observed within 5 min after addition of 1, 3, or 10 µM ADP (Chrono-Log Corp.) was recorded on a model 707 chart recorder (Chrono-Log Corp.). In a second study, the optical signal was digitized and analyzed using aggro/link software version 4.75 (Chrono-Log Corp.) with ADP tested at 10 µM concentration. In addition to the peak aggregation response, AUC, and percent aggregation measured at 5 min after ADP addition (late response) were measured to capture the ability of clopidogrel to enhance the reversibility of platelet aggregation.
Ligand Binding. Approximately 5 to 7 ml of PRP obtained from citrate-anticoagulated blood containing low-level apyrase (0.25 U/ml; Sigma catalog no. A-6132) was incubated with gentle agitation at 37°C with an additional 2.5 U/ml apyrase for 10 min. The sample was then diluted to 15 ml with buffer A (145 mM NaCl, 5.5 mM dextrose, 0.1 mM MgCl2, 5 mM KCl, 15 mM HEPES, pH 7.4, 5 mM EDTA) containing a final concentration of 0.5 µM prostaglandin E1 (Sigma catalog no. P-5515). The diluted PRP sample was then centrifuged at room temperature (1000g for 10 min), supernatant was discarded, and the pelleted platelets were washed once in 15 ml of assay buffer (145 mM NaCl, 0.1 mM MgCl2, 5 mM KCl, 15 mM HEPES, pH 7.4, 5 mM EDTA) containing 0.25 U/ml apyrase and 0.5 µM prostaglandin E1. The sample was recentrifuged (1000g, room temperature, 10 min), the supernatant was again discarded, and the pellet was resuspended in 5 ml of assay buffer. Platelet counts were then determined on a System 9000 cell counter (Serono-Baker Diagnostics) and were generally 0.2 to 0.4 x 106/µl.
Binding reactions were conducted in 96-well filter plates (Millipore Corporation, Billerica, MA; catalog no. MABVN0B50) prewet with 200 µl of assay buffer. Reactions consisted of 50 µl of the platelet preparation (107 platelets/reaction), 12.5 µM MRS-2179 (Sigma catalog no. M-3808) (an amount that had been previously shown to provide complete P2Y1 receptor block under these conditions), and 1 nM 33P-2MeS-ADP (2000 Ci/mol; Perkin-Elmer custom synthesis) in 200 µl of assay buffer. Binding was allowed to progress for 1 h at room temperature, and the platelets were then separated from the reaction by filtration and washed three times with 200 µl of ice-cold phosphate-buffered saline. Plates were allowed to air dry, 100 µl of scintillation fluid was added (Ultima Gold; PerkinElmer Life Sciences, Boston, MA; catalog no. 6013329), and the residual 2MeS-ADP binding was determined by scintillation counting. Total binding (0% receptor occupancy) was determined from vehicle-treated rats and generally corresponded to 10,000 cpm/107 platelets. Nonspecific binding (100% occupancy) was determined from the same vehicle-treated samples containing 12.5 µM cold 2MeS-ADP (Sigma catalog no. M-152) and generally corresponded to 200 cpm/107 platelets.
VASP Assay. The VASP-239 phosphorylation state was determined in citrate-anticoagulated blood samples using the PLT VASP/P2Y12 kit (Biocytex, Marseille, France). Samples were treated according to the protocol included with the kit. In brief, whole blood was incubated with either the PGE1 or PGE1 plus 10 µM ADP solutions provided with the kit for 10 min at room temperature. The exact composition of these reagents was not available from the manufacturer. Reacted blood samples were treated with fixative containing paraformaldehyde. Platelets were permeabilized and treated with either an anti-VASP-P mouse monoclonal antibody (16C2) or the negative isotypic control provided with the kit, and samples were then treated with a polyclonal antibody (anti-mouse IgG-FITC) and a platelet counterstaining reagent-PE. The platelet counterstain did not react efficiently with rat platelets. Reactions were immediately analyzed on a Becton Dickinson (San Jose, CA) FACSCalibur Flow Cytometer. The platelet population was identified by the forward and side scatter distributions, and 5000 platelets were gated and collected. Data from the negative isotypic control was used to correct the mean fluorescent intensities (MFIcs) of PGE1 (MFIcPGE) and PGE1 combined with ADP (MFIcPGE+ADP) samples containing the anti-VASP-P antibody. All blood samples were treated and analyzed within 4 h of sampling. A platelet reactivity index was calculated as outlined in the kit using the corrected mean fluorescence intensities in the presence of either PGE1 alone or PGE1 plus ADP using the following formula: platelet reactivity index = [(MFIcPGE - MFIcPGE+ADP)/MFIcPGE] x 100.
Statistical Analysis and Potency Calculation. Differences in bleeding time between aspirin or BMS-189664 and respective vehicle were determined by Student's t test. Clopidogrel effects on thrombus weight, baseline carotid blood flow, ex vivo biomarkers, and renal cortex bleeding time were determined by analysis of variance with Dunnett's test for comparison with the vehicle group. Due to variability in baseline carotid blood flows, this parameter was added as a covariate in the analysis of variance for integrated carotid blood flow during thrombosis. Contrasts were used for comparisons between clopidogrel doses and in the covariate analysis. Mesenteric and cuticle bleeding time data were analyzed by Kruskal-Wallis nonparametric analysis of variance and the Mann-Whitney U test to detect differences in vehicle and clopidogrel bleeding times and relative change in bleeding time. This nonparametric analysis was applied to mesenteric and cuticle bleeding times because off-scale bleeds in clopidogrel-treated rats rendered the data noncontinuous. The Fisher's exact test was used to compare frequency of occlusion data between treatment groups in arterial thrombosis and the proportion of animals that exceeded maximal recorded bleeding time. Computations were performed using Systat Version 11.00.01 (Systat Software, Inc., San Jose, CA). Data are presented as mean ± S.E.M. p < 0.05 was considered significant.
The clopidogrel dose calculated to produce a half-maximal response in each experimental model (ED50) was determined by this four-parameter logistic fit employing the Levenburg Marquardt algorithm: y = A + ((B - A)/(1 + ((C/x)^D))), where x is clopidogrel dose, y is percentage of maximal response, A and B are minimal and maximal y value set to 0 (A) and 100 (B), C is log ED50, and D is slope factor. The nonlinear curve fitting calculations were performed using XLfit (ID Business Solutions Ltd., Guildford, UK). Significant differences in ED50 potency values were determined by lack of overlap in the 95% confidence intervals.
|
Results |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
|
|
|
|
Aspirin produced a near-threshold antithrombotic effect with a 32 ± 12% reduction in thrombus weight comparing the vehicle (5.6 ± 0.5 mg, n = 5) and aspirin (3.8 ± 0.7 mg, n = 6; p = 0.068) groups. This result was reminiscent of three previous studies where decreases in thrombus weight of 9, 26, and 36% were observed comparing the same aspirin dose with vehicle. None of these thrombus reductions were significant. In view of this repeated and limited response to aspirin, we combined the four studies into a single meta-analysis achieving n = 30 to 33 per treatment group (Fig. 2). In the analysis of this combined study, aspirin treatment decreased thrombus weight by 24 ± 6%, improved average flow during thrombosis by 61 ± 16%, and maintained patency in 33% of the vessels compared with a 100% occlusion rate in vehicle-treated rats. These antithrombotic effects of aspirin were all significant.
|
3-fold (Fig. 3). Renal cortex incisions were practically unaffected by all but the highest 50 mg/kg dose of clopidogrel, where a 40% (i.e., 1.4-fold) increase in bleeding time was observed. The ED50 for prolonging bleeding time in both the mesenteric and cuticle models was 10 mg/kg (Table 1). Higher doses, which provided little additional benefit in preventing arterial thrombosis, resulted in off-scale bleeding upon severance of cuticle blood vessels at 20 mg/kg or mesenteric blood vessels at 30 mg/kg. The off-scale bleeding times were established in each model based on investigator experience that bleeding beyond that point did not stop without mechanical compression and could compromise stability of the animal. Maximal bleeding prolongations were in part dependent on the control bleeding time and averaged 6-fold in cuticle model and 8-fold in the mesenteric model. Bleeding from surgical sites was not observed in any clopidogrel-treated rats. The insensitivity of renal bleeds to clopidogrel was further investigated by increasing the incision depth and width and including a thrombin inhibitor as positive control for this model. Although the 20 mg/kg dose of clopidogrel had no effect on the standard incisions, it increased bleeding time from deep incisions by a factor of 1.44 ± 0.08 (n = 6) from 147 ± 21 to 213 ± 12 s (p < 0.05). The thrombin inhibitor BMS-189664 produced similar relative increases for both standard incisions (4.49- ± 0.69-fold from 70 ± 3 to 313 ± 47 s, n = 5; p < 0.05) and deep incisions (4.03- ± 0.64-fold from 159 ± 10 to 616 ± 74 s, n = 7; p < 0.05).
In each model, the 10 mg/kg dose of aspirin significantly extended bleeding time, but the effect was less than a doubling. The -fold increases in bleeding time were 1.46 ± 0.28 (n = 6) for mesenteric bleeding, 1.42 ± 0.10 (n = 6) for renal cortex bleeding, and 1.70 ± 0.28 (n = 7) for cuticle bleeding (all p < 0.05). There were no off-scale bleeds in any aspirin-treated rat.
Inhibition of Platelet Aggregation. In the first platelet function study, inhibition of ADP-induced aggregation by clopidogrel was dependent upon ADP concentration over the range of 1, 3, and 10 µM ADP (Fig. 4). Increasing ADP to 20 µM did not produce additional aggregation of rat platelets, so 10 µM ADP was set as the top agonist challenge (not shown). Although clopidogrel was most effective against 1 µM ADP, aggregation evoked by this concentration of agonist was highly reversible (Fig. 4, inset), and repeated studies in our lab have characterized this response as poorly reproducible. For the purpose of tracking in vivo effects, a more intermediate ADP concentration provided both reproducibility and sensitivity to clopidogrel. The optimal selection of agonist concentration was best revealed by plotting the relative antithrombotic and hemostatic effects of clopidogrel as a function of ex vivo inhibition of platelet aggregation at each ADP concentration (Fig. 5). By this analysis, the 10 mg/kg optimal antithrombotic dose of clopidogrel was at the midpoint of platelet function inhibition against 3 µM ADP. The corresponding bleeding time effect was a 3-fold increase, or about half-maximal. The 3 µM concentration of ADP, therefore, provided the broadest dynamic range of IPA for tracking both efficacy and bleeding liability. Interestingly, a significant antithrombotic effect was still observed with only 18% inhibition of aggregation to 3 µM ADP, and maximal effects on bleeding time were only achieved with near complete inhibition of this aggregation response. For this reason, a second series of experiments were conducted to evaluate additional biomarkers of platelet activation with the intent of obtaining greater sensitivity in tracking in vivo activity.
|
|
Effects on VASP and Ligand Binding Compared with Platelet Aggregation. In these experiments, simultaneous ex vivo platelet measurements were made of ADP-induced aggregation, ligand binding to the P2Y12 receptor, and phosphorylation of VASP. Clopidogrel dosed from 0.3 to 50 mg/kg produced maximal effects of 
88% inhibition for all indexes of platelet activation (Fig. 6). In the first IPA study, 10 µM ADP had provided the strongest challenge for clopidogrel inhibition; therefore, this stimulus was set as the target for improving IPA. Measuring platelet aggregation by AUC and at 5 min after agonist challenge (late response) were two simple approaches to increase IPA sensitivity. Clopidogrel might be expected to convert stable aggregation to reversible aggregation as is typically seen at lower ADP concentrations (Fig. 4, inset). Both AUC and late aggregation are indicators of reversible aggregation. The resulting ED50 values (Table 1) revealed a minimal sensitivity advantage of the AUC measure but a 2-fold improvement by using late aggregation. The VASP index, which has been characterized as a relevant P2Y12 effector response (Grossmann et al., 2004), offered no advantage over IPA in tracking platelet activation. Receptor occupancy was also a sensitive biomarker of clopidogrel's platelet effects because it matched the ED50 for antithrombotic activity.
The relative antithrombotic and hemostatic activities of clopidogrel were plotted as a function of platelet biomarker inhibition to better visualize the correspondence with in vivo activity (Fig. 7). In this graphical representation, the preferred biomarker would show equal separation between clopidogrel doses over the full dynamic range of the biomarker response and do so with no plateau in the response. Optimally, there would also be good sensitivity to the 1 and 3 mg/kg doses of clopidogrel that produced threshold antithrombotic effects. The result for the VASP assay and the most sensitive IPA measurement (late aggregation) involved overlapping of responses to clopidogrel doses at the high or low ends of the efficacy dose response. The ligand binding assay showed good separation of clopidogrel doses over the full dose response for antithrombotic activity.
|
|
Discussion |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
) is a more likely explanation. It is therefore reasonable to expect that the superior efficacy of P2Y12 receptor antagonists compared with aspirin could be overstated in rats. Nevertheless, the strong effects of P2Y12 antagonists in rats make them ideally suited to investigate the relationship between in vivo effects and ex vivo biomarkers. Unlike humans, the risk/benefit of pushing IPA to full effect can be easily studied in rats.
The standard for assessing clopidogrel activity ex vivo remains IPA, using the peak or maximal platelet aggregation to quantify the response (Cattaneo, 2006). Choice of ADP concentration has a direct impact on clopidogrel sensitivity. In rats, reversible aggregation to 1 µM ADP was too variable for reliable use. Full aggregation to 10 µM ADP was the least sensitive to inhibition, whereas 3 µM ADP offered a balance of reproducibility and sensitivity. The 10 mg/kg clopidogrel dose inhibited peak aggregation to 3 µM ADP by 49%, decreased thrombus weight by 60%, and provided maximal preservation of blood flow. Higher clopidogrel doses offered marginal improvements in antithrombotic effect despite having a much greater impact on IPA. Higher doses also revealed a bleeding liability. Mesenteric and cuticle bleeds were half-maximal at 10 mg/kg, whereas higher doses could elicit maximal bleeds. Using an intermediate IPA to 3 µM ADP, we identified the 10 mg/kg clopidogrel of having the most favorable balance between efficacy and safety.
However, low IPA measured in response to 3 µM ADP was also meaningful. The subthreshold 18% IPA achieved with the 3 mg/kg clopidogrel dose decreased thrombus weight by 45% and doubled the bleeding times. This suggests that standard IPA with 3 µM ADP was insensitive for identifying threshold antithrombotic activity. Standard IPA also did not produce a graded response in tracking in vivo activity over the full range of clopidogrel doses, opening the opportunity to improve or replace IPA with more sensitive measures.
We first sought an IPA measure having an improved potency against the strong 10 µM ADP stimulus. It has been observed in human PRP that clopidogrel reduces the stability of platelet aggregates, leading to augmented reversal of optical aggregation (Labarthe et al., 2005). Destabilization of aggregation results in diminished late aggregation, which can be measured as percent light transmission at 5 min after agonist challenge or as an AUC of the light transmission record. The result was a 2-fold decrease in the ED50 for late aggregation compared with peak aggregation. It is tempting to speculate that destabilization of platelet aggregates contributes to the antithrombotic and bleeding effects of P2Y12 antagonists. In this regard, clopidogrel was found to destabilize thrombi rather than limit initial thrombus formation in an in vitro perfusion system using human blood (Phillips et al., 2005).
Activation of the Gi-coupled P2Y12 receptor by ADP leads to both reduced cAMP formation and the associated dephosphorylation of VASP. Platelet fibrinogen receptors required for platelet aggregation are negatively regulated by phosphorylated VASP, and P2Y12 activation augments aggregation by releasing this inhibition. Clopidogrel inhibits platelet aggregation in part by limiting the ADP-mediated dephosphorylation of VASP in response to cAMP-elevating stimuli (especially PGE1). This formed the basis for the original VASP assay (Geiger et al., 1999), which was subsequently modified to track VASP phosphorylation by flow cytometry (Aleil et al., 2005). The VASP assay has been used in small clinical studies to quantify responsiveness of patients to clopidogrel therapy (Grossmann et al., 2004). However, in our experiments, clopidogrel effects on the VASP index were no better and slightly less sensitive to clopidogrel than standard IPA. It is possible that the coupling between receptor and effector responses, such as VASP phosphorylation, differs between rodents and man. In comparing the proportionate effect on VASP and receptor binding (Fig. 6), it would appear that in rats, blockade of 90% P2Y12 receptors is required to produce a 50% effect on the VASP index.
The irreversible binding of clopidogrel to P2Y12 receptors offers the opportunity for an ex vivo binding assay based on the readily available radioligand ([33P]2MeS-ADP). Because this antagonist binds to both P2Y1 and P2Y12 receptors (Baurand et al., 2000), an excess of P2Y1 antagonist was added to confer selectivity to P2Y12 receptors. In rats, clopidogrel (20 mg/kg) achieved 96% inhibition of [33P]2MeS-ADP-specific binding under these conditions. Ligand binding was a sensitive measure of platelet inhibition by clopidogrel with an ED50 equal to the antithrombotic ED50, but it was only marginally superior to the late aggregation response. Although binding showed significant activity over the entire clopidogrel dose-response range for both antithrombotic and bleeding effects, as an ex vivo biomarker, it is only applicable to irreversible antagonists like clopidogrel. Its ability to replace modified IPA assays is therefore limited and would be best suited to comparing compounds of similar mechanism.
Regardless of the biomarker selected, a proportionately greater effect on efficacy over bleeding time prolongation was observed at all but the highest levels of biomarker inhibition. In general, inhibition of arterial thrombosis was well matched to P2Y12 ligand binding, whereas increased cuticle bleeding time tracked well with either VASP index or inhibition of peak (maximal percentage) platelet aggregation (Fig. 8).
|
). Off-scale bleeding in mesenteric or cuticle models required clopidogrel doses (10 and 20 mg/kg) that had maximal effects on the most sensitive biomarkers. Other bleed models may be less sensitive. Depending on methodology, a 20 mg/kg dose of clopidogrel produced either off-scale bleeds (Herbert et al., 1993) or no effect (Daykin et al., 2006) on rat tail bleeding time. In the renal cortex bleeding time, only the highest clopidogrel dose (50 mg/kg) had a significant effect in matching the threshold activity of aspirin. When the depth of the renal incision was increased, the 20 mg/kg dose of clopidogrel produced a significant yet modest 44% increase in bleeding time. We have also found renal cortex bleeds insensitive to the combination of a P2Y1 antagonist (MRS-2500) and clopidogrel at maximal IPA doses, whereas anticoagulants (heparin, thrombin inhibitors, and factor Xa inhibitors) have produced bleeding time increases of 2- to 5-fold (W. A. Schumacher, unpublished data). In the current study, the thrombin inhibitor BMS-189664 was highly effective in prolonging bleeding in response to renal injury, suggesting that thrombin generation is a most important contributor to hemostasis in the rat renal cortex. The discrepancy in tail bleeds and the limited effects of clopidogrel on renal bleeds suggest that the microcirculation of different tissues vary in sensitivity to P2Y12 antagonism. Although the applicability of these results in other species remains to be investigated, it is clear that vascular bed selection is crucial when assessing hemostatic liability.
In summary, the study of clopidogrel in these rat models supports paying careful attention to biomarker selection when attempting to predict in vivo activity. These findings strongly suggest that methods more sensitive than standard IPA may need to be developed to detect functional antiplatelet activity of clopidogrel associated with less than 50% P2Y12 blockade. In addition to measurements of P2Y12 receptor binding and postreceptor signaling, alternatives may include modifications of IPA with either simultaneous blockade of P2Y1 receptors or amplification of the cAMP effect (Malinin et al., 2007). Regardless of the approach used, any effect on an ex vivo platelet function biomarker becomes validated on the basis of its consistent and predictable association with in vivo efficacy and bleeding safety. A very high standard of evidence is required for such a biomarker to guide medical treatment decisions, and establishing the target range for therapeutic biomarker responses will require extensive testing in large clinical event trials.
|
Footnotes |
|---|
ABBREVIATIONS: IPA, inhibition of platelet aggregation; VASP, vasodilator-stimulated phosphoprotein; MRS-2179, 2'-deoxy-N6-methyladenosine-3',5'-diphosphate; AUC, area under the curve; PRP, platelet-rich plasma; PGE, prostaglandin E; MFIc, corrected mean fluorescent intensity; BMS-18664, N-[[1-(aminoiminomethyl)-4-piperidinyl]methyl]-1-[N-(methylsulfonyl)-D-phenylalanyl]-L-prolinamide; 33P-2MeS-DP, 33P-2-methylthio-adenosine diphosphate.
Address correspondence to: Dr. William A. Schumacher, Bristol-Myers Squibb Company, 311 Pennington-Rocky Hill Road, Pennington, NJ 08534. E-mail: william.schumacher{at}bms.com
|
References |
|---|
|
TopAbstractMaterials and MethodsResultsDiscussionReferences |
|---|
Aleil B, Ravanat C, Cazenave JP, Rochoux G, Heitz A, and Gachet C (2005) Flow cytometric analysis of intraplatelet VASP phosphorylation for the detection of clopidogrel resistance in patients with ischemic cardiovascular diseases. J Thromb Haemost 3: 85-92.[CrossRef][Medline]
Baurand A, Eckly A, Bari N, Léon C, Hechler B, Cazenave J-P, and Gache C (2000) Desensitization of the platelet aggregation response to ADP: differential down-regulation of the P2Y1 and P2cyc receptors. Thromb Haemost 84: 484-491.[Medline]
CAPRIE Steering Committee (1996) A randomised, blinded, trial of clopidogrel versus aspirin in patients at risk of ischaemic events. Lancet 348: 1329-1339.[CrossRef][Medline]
Cattaneo M (2006) ADP receptors: inhibitory strategies for antiplatelet therapy. Drug News Perspect 19: 253-259.[CrossRef][Medline]
Das J, Kimball S, Hall SE, Han W, Iwanowicz E, Lin J, Moquin RV, Reid JA, Sack JS, Malley MF, et al. (2002) Molecular design and structure-activity relationships leading to the potent, selective, and orally active thrombin active site inhibitor BMS-189664. Bioorg Med Chem Lett 12: 45-49.[CrossRef][Medline]
Daykin HJ, Sturgeon SA, Jones C, and Wright CE (2006) Arterial antithrombotic effects of aspirin, heparin, enoxaparin and clopidogrel alone, or in combination, in the rat. Thromb Res 118: 755-762.[CrossRef][Medline]
Geiger J, Brich J, Honig-Liedl P, Eigenthaler M, Schanzenbacher P, Herbert JM, and Walter U (1999) Specific impairment of human platelet P2Y(AC) ADP receptor-mediated signaling by the antiplatelet drug clopidogrel. Arterioscler Thromb Vasc Biol 19: 2007-2011.
Grossmann R, Sokolova O, Schnurr A, Bonz A, Porsche C, Obergfell A, Lengenfelder B, Walter U, and Eigenthaler M (2004) Variable extent of clopidogrel responsiveness in patients after coronary stenting. Thromb Haemost 92: 1201-1206.[Medline]
Herbert JM, Bernat A, and Maffrand JP (1993) Aprotinin reduces clopidogrel-induced prolongation of the bleeding time in the rat. Thromb Res 71: 433-441.[CrossRef][Medline]
Hollopeter G, Jantzen HM, Vincent D, Li G, England L, Ramakrishnan V, Yang RB, Nurden P, Nurden A, Julius D, et al. (2001) Identification of the platelet ADP receptor targeted by antithrombotic drugs. Nature 409: 202-207.[CrossRef][Medline]
Labarthe B, Théroux P, Mand A, and Ghitescu M (2005) Matching the evaluation of the clinical efficacy of clopidogrel to platelet function tests relevant to the biological properties of the drug. J Am Coll Cardiol 46: 638-645.
Leadley RJ Jr, Chi L, Rebello SS, and Gagnon A (2000) Contribution of in vivo models of thrombosis to the discovery and development of novel antithrombotic agents. J Pharmacol Toxicol Methods 43: 101-116.[CrossRef][Medline]
Malinin A, Pokov A, Spergling M, Defranco A, Schwartz K, Schwartz D, Mahmud E, Atar D, and Serebruany V (2007) Monitoring platelet inhibition after clopidogrel with the VerifyNow-P2Y12(R) rapid analyzer: The VERIfy Thrombosis risk AS-sessment (VERITAS) study. Thromb Res 119: 277-284.[CrossRef][Medline]
Nguyen TA, Diodati JG, and Pharand C (2005) Resistance to clopidogrel: a review of the evidence. J Am Coll Cardiol 45: 1157-1164.
Phillips DR, Conley PB, Sinha U, and Andre P (2005) Therapeutic approaches in arterial thrombosis. J Thromb Haemost 3: 1577-1589.[CrossRef][Medline]
Savi P and Herbert JM (2005) Clopidogrel and ticlopidine: P2Y12 adenosine diphosphate-receptor antagonists for the prevention of atherothrombosis. Semin Thromb Hemost 31: 174-183.[CrossRef][Medline]
Schumacher WA, Heran CL, Youssef S, Megill JR, Michel I, and Durham SK (1993) Comparison of a thromboxane receptor antagonist and aspirin in experimental arterial thrombosis. Haemostasis 23: 219-228.[Medline]
Schumacher WA, Steinbacher TE, Megill JR, and Durham SK (1996) A ferret model of electrical-induction of arterial thrombosis that is sensitive to aspirin. J Pharmacol Toxicol Methods 35: 3-10.[CrossRef][Medline]
Thebault JJ, Kieffer G, Lowe GD, Nimmo WS, and Cariou R (1999) Repeated dose pharmacodynamics of clopidogrel in healthy subjects. Semin Thromb Hemost 25 (Suppl 2): 9-14.
van Gestel MA, Heemskerk JW, Slaaf DW, Heijnen VV, Reneman RS, and Oude Egbrink MG (2003) In vivo blockade of platelet ADP receptor P2Y12 reduces embolus and thrombus formation but not thrombus stability. Arterioscler Thromb Vasc Biol 23: 518-523.
This article has been cited by other articles:
|
|
 |
|
  M. Gilard, J.-C. Cornily, and J. Boschat Reply J. Am. Coll. Cardiol., September 16, 2008; 52(12): 1039 - 1039. [Full Text] [PDF]
|
|
|
|
|
|
L. Bonello, L. Camoin-Jau, S. Arques, P. Barragan, F. Dignat-George, and F. Paganelli Reply J. Am. Coll. Cardiol., August 26, 2008; 52(9): 791 - 792. [Full Text] [PDF]
|
|
|
|
|
|
L. Bonello, L. Camoin-Jau, S. Arques, C. Boyer, D. Panagides, O. Wittenberg, M.-C. Simeoni, P. Barragan, F. Dignat-George, and F. Paganelli Adjusted Clopidogrel Loading Doses According to Vasodilator-Stimulated Phosphoprotein Phosphorylation Index Decrease Rate of Major Adverse Cardiovascular Events in Patients With Clopidogrel Resistance: A Multicenter Randomized Prospective Study J. Am. Coll. Cardiol., April 8, 2008; 51(14): 1404 - 1411. [Abstract] [Full Text] [PDF]
|
|
| |||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
 |
 |
|
 |
 |
 |
