OUP user menu

Personalized medicine and antiplatelet therapy: ready for prime time?

Céline Verstuyft , Tabassome Simon , Richard B. Kim
DOI: http://dx.doi.org/10.1093/eurheartj/ehp295 1943-1963 First published online: 28 July 2009
  • Personalized medicine
  • Pharmacogenetics
  • Antiplatelet
  • Aspirin
  • Thienopyridines
  • Cardiovascular disease


The concept of personalized medicine is receiving significant attention due to the greater awareness of the influence of genes to the drug effects. Single nucleotide polymorphisms (SNPs) in the DNA are the most frequent form of sequence variations in the human genome and appear to affect the efficacy and safety of many drugs. The term ‘pharmacogenetics’ was coined over 40 years ago with an ultimate goal of using the genetic makeup of an individual to predict drug response and efficacy.13 We are just at the beginning of a new era in personalized cardiovascular therapies. However there is little doubt that, in the near future, pharmacogenetic testing will become a valuable tool for a drug and dose selection and thus result in a more desirable benefit/risk ratio for drugs prescribed to patients.

Over the past decades, the platelet has emerged as a major pathway involved in cardiovascular diseases. The platelet as a ‘drug target’ has spawned a variety of new drugs that have been shown in large-scale randomized trials to improve patient outcomes in acute coronary syndromes and following percutaneous revascularization procedures.46 Until recently aspirin, centred on the tromboxane pathway, was the only antiplatelet agent considered to be the gold standard for effectiveness in both primary and secondary prevention of atherothrombotic diseases.7 Although it continues to be used as the gold standard antiplatelet therapy, adenosine diphosphate (ADP) receptor antagonists and phosphodiesterase inhibitors in combination therapy appear to exert synergistic effects and provide added benefits among high-risk patients for cardiovascular disease.7,8

Nevertheless an important lesson that has emerged from number of trials is that antiplatelet potency per se does not necessarily guarantee enhanced clinical benefit or tolerability for a given patient.811 This may in part be due to the substantial interindividual variation in platelet response to ADP.911 The mechanism underlying such variation has recently become clearer (Figure 1). Specifically the wide inter-subject variabilities to antiplatelet agents such as clopidogrel, may be genetically mediated and arises from altered drug metabolism or transport.1215 In the current review, we will focus on the key molecular mechanisms involved in the pharmacological action of oral antiplatelet drugs, the environmental and genetic factors that may impact antiplatelet therapies. We will also provide an update on recent advances in personalized medicine of relevance to arterial thrombosis and antiplatelet drugs. Finally, we will provide our perspectives of pharmacogenetic testing for drugs used to treat cardiovascular diseases.

Figure 1

Factors influencing the variability of antiplatelet drug response.

Mechanism of actions and clinical relevance

Current therapeutic strategies for the treatment of arterial thrombosis are based on well-known receptor systems (Figure 2). Collagen and/or thrombin interact with activated platelets and their receptor GPIIb–IIIa to bind fibrinogen and von Willebrand factor and initiate platelet aggregation. Stable aggregation of platelets is amplified by two autocrine factors generated upon platelet stimulation: ADP, released from platelet and Thromboxane A2 (TXA2), generated by the sequential actions of cyclooxygenase-1 (COX-1) and thromboxane synthase from the arachidonic acid released from membrane phospholipids.5

Figure 2

Clopidogrel absorption, metabolism, and aspirin target.


Aspirin was the first and continues to be the most widely used antiplatelet agent. In platelets, the major cyclooxygenase product is thromboxane A2. Aspirin blocks the production of TXA2 by acetylating a serine residue near the active site of platelet COX-1, the enzyme that produces the cyclic endoperoxide precursor of TXA2. Since platelets are not able to synthesize new proteins, the action of aspirin on platelet COX-1 is permanent, and persists for the life of the platelet (7–10 days). Thus, repeated doses of aspirin produce a cumulative effect on platelet function. Complete inactivation of platelet COX-1 has been shown to occur when 160 mg of aspirin is taken daily.16

The efficacy of aspirin has been appreciated for many years and data from the meta-analysis, the Antiplatelet Trialists' Collaboration found an ∼25% relative risk reduction of vascular death, MI, or stroke for antiplatelet therapy, primarily aspirin vs. placebo.17 This data set served as the foundation for the widespread adoption of aspirin as the standard regimen for the secondary prevention of cardiovascular events.

However a number of clinical trials have shown that many patients receiving aspirin still sustain a thrombotic event, and therefore referred as ‘aspirin resistant’. The prevalence of aspirin resistance is thought to range anywhere from 5 to 40%.18 This phenomenon appears to be a true entity of clinical relevance since it cannot be overcome by increasing aspirin dose.19 Despite intensive research relating to aspirin resistance, this topic remains controversial mainly because of the lack of an optimal biomarker and validated assay. A key step to understanding aspirin resistance could be the identification of the relevant genetic determinants that mediate aspirin resistance. Different target protein and genetic polymorphisms such as the PLA1/A2 polymorphism of platelet glycoprotein IIIa have been linked to the response to aspirin therapy2022 as well as an increased risk of thrombotic events.23,24 Moreover increased expression of platelet COX-2 messenger RNA has been linked to aspirin resistance,25,26 although this is controversial.27 Further studies are needed to determine the ultimate clinical relevance of these findings.


The second most widely prescribed antiplatelet agents for chronic therapy are thienopyridines which target the P2Y12 receptor.28 The key mediator of platelet activation is ADP which is released from platelet dense granules by activating stimuli such as thrombin, collagen, and thromboxane A2. Net result of ADP is the alteration of platelet conformation, intracellular calcium increase, adenylyl cyclase down-regulation, protein phosphorylations, activation of the GPIIb–IIIa complex which results in fibrinogen binding, aggregation, and release. Adenosine diphosphate is known to be the fundamental step of platelet activation via the P2Y1 receptor, while binding of ADP to P2Y12 receptor amplifies this response and allows sustained ADP-induced platelet aggregation.29 Consequently, binding of ADP to P2Y12 receptor not only amplifies the aggregation response but also increases granule secretion and platelet procoagulant activity.30 Therefore ADP-mediated activation of P2Y12 represents a critical pathway that results in arterial thrombosis and the accompanying tissue anoxia and inflammatory response. Not surprisingly pharmacological targeting of this receptor has become an important antiplatelet treatment strategy.

Clopidogrel and its predecessor ticlopidine are thienopyridine ADP receptor antagonists. These drugs function as irreversible platelet inhibitors, sustaining their activation for the life of the platelets. Note that both are prodrugs, which undergo hepatic metabolism by cytochrome P450 enzymes (CYPs)3A4 and 2C19 before generating the active metabolite, a transient intermediate which inactivates the receptor.28,31 Cytochrome P450 enzymes are important in the biosynthesis and degradation of endogenous compounds such as steroids, lipids, and vitamins and the metabolism of xenobiotics. They reduce or alter the pharmacological activity of most of the currently prescribed drugs and facilitate their elimination. The liver is the major site of CYP metabolism, but the small intestine is also a potentially important organ for drug metabolism and transport.32

Ticlopidine has been shown to be efficacious in conditions such as claudication, unstable angina, coronary artery and peripheral bypass surgery, and cerebrovascular disease.33,34 However, ticlopidine use has been reduced because of rare, but significant, adverse side effects such as neutropenia that require regular monitoring of white blood cell count, and a potentially life-threatening thrombotic thrombocytopenic purpura.35

Clopidogrel requires oxidation mainly dependent on the cytochrome P450 enzymes 2C19 (CYP2C19) and to a lesser extent on isoenzymes CYP2C9, 3A4, 3A5, 2B6.14,3639 Only 15% of the prodrug is available as an active agent; the remaining 85% is hydrolysed into an inactive compound (Figure 2). Although, its half-life is only 8 h, it has an irreversible effect on platelets that lasts 7–10 days. Inhibition of platelet aggregation appeared 2 h after the first dose, became significant after the second dose, and progressed to a steady-state value of 55–57% by day 7.40 It was suggested that the P-glycoprotein(P-gp) transporter also limits the intestinal absorption of clopidogrel, thereby controlling its antiplatelet activity.41,42

The first clear evidence for the efficacy benefit from clopidogrel was shown in CAPRIE trial evaluating patients with atherosclerotic disease.43 Subsequently, several large clinical trials have confirmed in others populations the efficacy of clopidogrel co-administration in reducing cardiovascular events (CURE,44 CREDO,35 PCI-CURE45). The use of clopidogrel has been extended to patients with non-ST-segment elevation ACS (unstable angina and non-ST-segment elevation MI) independent of coronary revascularization,44 and patients with ST-segment elevation MI, including those undergoing PCI.4648

Incidence of side effects, such as gastrointestinal disorders, neutropenia, and thrombotic thrombocytopaenic purpura,49,50 is far lower in comparison to ticlopidine. Moreover, its second major benefit over ticlopidine was its ability to yield antiplatelet effects more rapidly through the administration of a loading dose.51 Occasional resistance to clopidogrel and interpatient variability in drug response has spurred the development of new therapies.

Prasugrel is a third-generation oral thienopyridine that is chemically distinct from clopidogrel. Like clopidogrel, prasugrel is a specific, irreversible antagonist of the platelet P2Y12 ADP receptor. It is rapidly hydrolysed by esterases to an inactive thiolactone, which is then metabolized by hepatic CYPs to the active metabolites. The major hepatic pathway involve CYP3A4 and CYP2B6, and to a lesser extent the CYP2C9 and CYP2C19.52 It is rapidly absorbed and metabolized, with a median time for achieving the maximal concentration of its active metabolite in the circulation of about 30 min.53,54 The mean elimination half-life of active metabolite is 3.7 h, and renal excretion (around 70%) is the major route for elimination.

The major difference between clopidogrel and prasugrel is their bioavailability; in fact a significant portion of the administered dose of clopidogrel is activated rapidly through metabolism, resulting in lower apparent bioavailability of the active metabolite. As a consequence, preclinical studies have shown that prasugrel is an orally active antiplatelet agent that is a more potent inhibitor of platelet aggregation on a milligram per kilogram basis, with a faster onset of action.54,55

Limitations of current antiplatelet therapies

Dual antiplatelet pathway inhibition appears to offer synergistic benefit in preventing thrombus formation,56,57 but all patients do not benefit to the same extent. Up to 15% of the high-risk patients with acute coronary syndrome continue to suffer from ischaemic events, and up to one-third of patients have a marked interindividual variability in the extent of platelet inhibition.58

The prevalence of this phenomenon, referred to a clopidogrel non-responsiveness or resistance, varies widely according to the literature.10,58,59 Table 1 summarizes the studies in which various measures of clopidogrel responsiveness, mainly post-treatment platelet reactivity, have been studied. A recent meta-analysis found an overall prevalence of 21% (95% CI, 17–25%) of laboratory-defined clopidogrel non-responsiveness. The differences in reported prevalences partly depend on the loading dose of clopidogrel and the methods of determining non-responsiveness.60 Interestingly patients labelled as clopidogrel resistant using ex-vivo assays have an increased risk of stent thrombosis and other cardiovascular outcomes,60 but the use of 600 mg clopidogrel loading dose appears to reduce such risks.

View this table:
Table 1

Details of included studies on prevalence of laboratory clopidogrel non-responsiveness

StudyDesignnClopidogrel dose (mg)Aspirine dose (mg)Functional parameter and/or outcomeDefinition of non-responsiveness aggregation assayDetermination of platelet aggregationEnd-pointFollow-upNon-responsiveness n (%)
Stent thrombosis
 Muller et al. 200310Prospective cohort105LD 600; MD 75100Decrease inhibition of platelet aggregationLTA (5 or 20 µmol/L ADP): <10% reduction /baseline4 h after LDStent thrombosi (ST)14 days5 (5), 12 (11)
 Barragan et al. 200393Prospective cohort1684MD 75×2250Increase P2Y12 reactivity ratio; increase platelet aggregationVASP-P (sodium citrate 0.129 mol/L) monoclonal antibody0, 2, and 4.8 days after PCI (controls)ST30 days17 (1.03)
 Gurbel et al. 2005131Case–control20 cases; 100 controlsLD 300; MD 7581–325Increase P2Y12 reactivity ratio, increase platelet aggregationLTA (5 or 20 µmol/L ADP): after treatment PR > 75th percentile in controlsCases: 218 ± 204 days after LD; Controls: 5–14 days after LDSTCases: 218 ± 204 days after LD; Controls: 5–14 daysNA
 Ajzenberg et al. 2005132Case–control10 cases; 22 controls; 17 healthy volunteersLD 300; MD 7575–250Increase shear-induced platelet aggregation (SIPA) increase P2Y12 reactivity ratioSIPA and LTA monoclonal antibodyCases: within 4.6 ± 3.4 days of SAT; Controls within 3 days after clopidogrelSTNA19 (1.2)
 Buonamici et al. 2007133Prospective cohort804LD 600; MD 75325Increase platelet aggregationLTA (10 µmol/L ADP) 90th percentile of controls (70%)12–108 h from dose 6 days after PCIST6 months25 (3.1)
POST-PCI myonecrosis and ischaemic events
 Matetzky et al. 2004134Prospective cohort60LD 300; MD 75200Increase platelet aggregationLTA (5 µmol/L ADP): first quartile of reductions compared with baseline6 days after LDSTEMI, ACS, PAD ischaemic stroke6 months15 (25.0)
 Gurbel et al. 2005135Prospective cohort192LD 300; MD 7581–325Increase platelet aggregationLTA (20 µmol/L ADP) 4th quartile of aggregation24 h after LDCV death, MI, ACS stroke6 monthsNA
 Cuisset et al. 2006136Randomized controlled trial292LD 600 (n = 146); LD 300 (n = 146); MD 75160Increase platelet aggregationLTA (10 µmol/L ADP) aggregation >70%12 h after LDCV death, SAT, ischaemic stroke, ACS1 month58 (20); 15% LD 600 mg; vs. 25% LD 300 mg
 Lev et al. 2006137Prospective cohort150LD 300; MD 7581–325Increase clopido/aspirin-resistant patientsLTA (5 or 20 µmol/L ADP): <10% reduction/baseline20–24 h after LDCK-MB >5 ng/mL20–24 h36 (24)
 Cuisset el al. 2006138Prospective cohort106LD 300; MD 75160Increase platelet aggregationLTA (10 µmol/L ADP) aggregation 4th quartile of aggregation12 h after LDCV death, ST, stroke, ACS1 month23 (22)
 Hochholzer et al. 2006139Prospective cohort802LD 600; MD 75>100Increase platelet aggregationLTA (20 µmol/L ADP): no definitionAt least 2 h after LDDeath, MI, revascularization1 monthNA
 Geisler et al. 2006140Prospective cohort379LD 600; MD 75100Decrease platelet inhibitionLTA (20 µmol/L ADP) >70%34.8 ± 25.9 h after LDCV death, MI, stroke3 months22 (6)
 Bliden et al. 2007141Prospective cohort100MD 7581 (7 days); 325Augm platelet aggregationLTA (5 µmol/L ADP); thromboelastograph; haemostasisBefore, 3, 18, and 24 h afterwardsCV death, MI, stroke, ischaemia1 year2/22 (9.0)
 Cuisset et al. 2007142Prospective cohort190LD 600; MD 75250Increase platelet aggregationLTA (10 µmol/L ADP) aggregation >70%Before, 12 h, 24 hAMINA54
 Bonello et al. 2007143Prospective cohort144LD 300; MD 75160Increase P2Y12 reactivity ratioVASP (monoclonal antibody 16C2 (2nd through 5th quintiles)After LD 25 ± 3 hMACE6 months21
 Angiolillo et al. 2007144Prospective cohort173MD 75100Increase platelet aggregationLTA (10 µmol/L ADP) aggregation (4th quartile)3–6 and 24 monthsMACE2 years(19.7)
 Frere et al. 2007145Prospective cohort195LD 600; MD 75LD 250; MD 75Increase platelet aggregation; increase P2Y12 reactivity ratio,VASP (monoclonal antibody 16C2 LTA (10 µmol/L ADP) aggregation >70%Before and 18.2 ± 2.2 hCV, death, acute SAT, ACS, and stroke30 days14
 Price et al. 2008146Prospective cohort380LD 600; MD 75LD 325; MD 325Increase P2Y12 reactivity unitsVerifyNow NA P2Y12 (ADP20 µmol/L), PRU12 h afterwardsCV, death, MI, stent6 months10 (2.6)
 Bonello et al. 2008147Prospective control randomized162LD 600 control; LD 600; MD 75MD 160Increase platelet aggregation; increase P2Y12 reactivity ratioVASP (monoclonal antibody 16C2); LTA (ADP+PGE1)24 h after 1 LD; 12 h after 2 LDMACE1 month3/84 (3.6); 0% in VASP-P guided group
 Patti et al. 2008148Prospective cohort160LD 600; MD 75NAIncrease P2Y12VerifyNow NA P2Y12 PRU assayBefore and 8, 24 h afterwardsMACE6 monthsNA
 Marcussi et al. 2009127Prospective cohort683LD 600; MD 75100–325Increase P2Y12VerifyNow P2Y12 assay (ADP10 µmol/L), PRU assay24 h after LDCV death, MI12 monthsNA
 Von Beckerath et al. 2009149Randomized; study66LD 600; MD 75 or 150200% inhibition platelet aggregation; increase P2Y12VerifyNow P2Y12 assay (ADP 5 µmol/L), PRU assay30 days after PCIPlatelet aggregation30 daysNA
 Sibbing et al. 2009150Prospective cohort1608LD 600; MD 150 (3 days); MD 75200Increase platelet aggregationMEA Multiplate Analyser (ADP 6.4 µmol/L)Before and after aspirin doseSAT, death, TIMI major bleeding30 days323 (20)
  • LTA, light transmittance aggregometry; PCI, percutaneous coronary intervention; PRU, platelt reactivity unit; SIPA, shear-induced platelet aggregation; VASP-P, vasodilator-stimulate phosphoprotein phosphorylation; ST, stent thrombosis; LD, loading dose; MD, maintenance dose; NA, not available; MACE, major cardiovascular events; TIMI, Thrombolysis In Myocardial Infarction.

Contemporary basic and clinical pharmacology have evolved to embrace an increasingly sophisticated molecular view of the mechanisms underlying drug action. Variability in drug action may be the result of pharmacokinetic or pharmacodynamic differences. Pharmacokinetic variability refers to variability in delivery of drugs to, or removal from, key molecular sites of action that mediate efficacy and/or toxicity. Pharmacodynamic variability refers to variable drugs effects despite equivalent drug delivery to molecular sites of action. In fact, although the best method of assessing antiplatelet drug response has not been established yet, there is sufficient evidence to support that persistence of enhanced platelet reactivity plays a key role in atherothrombotic complications.8 The mechanisms leading to poor response to clopidogrel have not been fully elucidated and are probably multi-factorial.61 Compliance, cellular, environmental, genetic, and clinical factors such as obesity, diabetes mellitus, nature of coronary injury, and inflammation are known to contribute to variable antiplatelet drug response (Figure 1).62,63

Furthermore, another major limitation inherent to the thienopyridines is attributed to the irreversible antiplatelet effects. Indeed, bleeding events are one of the well-known major side effects for all antithrombotic agents, particularly with antiplatelet therapies. The development of new antiplatelet agents with a reversible mechanism of action, allowing platelet function to return more rapidly to baseline status will likely reduce the risk of bleeding in patients undergoing surgery.28,61

Determinants of antiplatelet therapy: non-genetic factors of variability

The environmental factors, such as diet, drug–drug interaction with drug transporter, protein target function, and CYPs are known to be involved as key determinants of intersubject variation in drug responsiveness32(Figure 1). In fact, it was described that the level of clopidogrel active metabolite concentration needed to inhibit P2Y12 receptor is suboptimal in some patients.63 The limited efficacy of aspirin and clopidogrel suggests the existence of alternative pathways for platelet activation and/or possible drug interactions such as proton-pump inhibitors (PPIs). Indeed similar to clopidogrel, PPIs are sharing the same metabolic pathway extensively metabolizing in the liver.64 The increase in the loading dose, a pharmacokinetic solution that takes into consideration elimination pathways such as certain intestinal transporters, has been suggested as a way for decreasing the risk of drug non-responsiveness.42

Drug transporters are increasingly recognized to be important to drug disposition and response. The oral bioavailability of various drugs is limited by active luminal secretion via adenosine triphosphate binding cassette (ABC) efflux transporters in the intestine—in particular P-gp encoded by the multidrug resistance gene ABCB1 (MDR1). Many substrates of drug metabolizing enzymes, particularly CYP3A4, are also substrates of P-gp; the overlap between CYP3A4 and P-gp substrates may have resulted in part from the coordinated regulation and tissue expression of CYP3A4 and ABCB1 organs such as the liver and the intestine.65 P-gp was found to be a key factor for intestinal absorption of clopidogrel, limiting its bioavailability.41 Moreover, a linear correlation of Cmax values has been shown between clopidogrel and its active metabolite, suggesting that interindividual differences in the activity of metabolizing enzymes (CYP3A4 or 3A5) are not the rate-limiting step for generation of the active metabolite.42

This is interesting with regard to ongoing and future clinical trials. Some studies and a recent meta-analysis support the hypothesis that an increase of clopidogrel loading dose (600 mg/day) could lead to a lower prevalence of clopidogrel non-responsivenes with a more potent and rapid antiplatelet effects than 300 mg dose.60,66,67 Three other studies have confirmed this finding.6870

The results of the large ongoing CURRENT-OASIS 7 trial may help to better define optimal dosing regimens for clopidogrel in acute coronary syndrome patients.71 However, recently in a small number of NSTEACS patients (n = 256), clopidogrel 600 mg LD compared with 300 mg LD was associated with significantly reduced ADP-induced platelet aggregation (49.7 vs. 55.7% with ADP 20 µmol/L) but did not reduce post-PCI myonecrosis or adverse clinical outcomes to 6 months.72 Moreover, the ISAR-CHOICE trial69 showed that an increase of clopidogrel loading dose from 600 to 900 mg was not associated with an additional suppression of platelet function because of limited clopidogrel absorption.

Therefore, it is probable that there is a threshold, likely attributable to the absorption and clopidogrel metabolite formation rate, which limits additional enhancement of the platelet inhibitory effects beyond a certain dose.

Metabolism and interindividual variability

Differences in drug metabolism are common, often marked and are frequently major contributors to differences in drug response among patients. CYP3A4, CYP2C9, CYP2C19, and CYP1A2 are involved in the formation of the active clopidodrel metabolites.73 CYP3A isoenzymes (CYP3A4 and CYP3A5), which are per se heterogenous, appear to be the primary oxidative pathway for clopidogrel.39,74 CYP3A5, which is polymorphically expressed, may contribute as much as 50% of hepatic CYP3A activity in certain ethnic populations.75,76

Drug–drug interactions resulting in either inhibition or induction of the involved enzymes, especially those in the intestine and liver, can markedly alter oral bioavailability.32 The metabolism of clopidogrel is inhibited by the CYP3A4 inhibitor, ketoconazole and induced by rifampicin.37 Moreover, drug–drug interactions with lipophilic statins and PPIs are thought to alter the pharmacodynamic effects of clopidogrel.


Some studies,7779 but not all,8086 have shown that atorvastatin and simvastatin, which are metabolized by CYP3A4 appear to reduce clopidogrel-induced antiplatelet effects. The discrepancies between the pharmacological findings can be explained at least in part by the study designs. Several studies80,83 have considered all statins instead of evaluating those inhibiting CYP3A. Although many statins (atorvastatin, simvastatin, lovastatin, cerivastatin) are substrate of CYP3A4, the attained therapeutic plasma levels are not sufficient to inhibit CYP3A4. Moreover, the frequent concomitant administration of other CYP3A substrates and inhibitors, modulating clopidogrel activity, were not taken into account in the control groups.87

Finally, these findings were not replicated in larger studies which did not show a clinical or biological interaction between lipophilic statins and clopidogrel.85,86

Proton-pump inhibitors

Recent guidelines published by the American Heart Association, the American College of Gastroenterology, and the American College of Cardiology advocate PPI therapy for patients receiving ASA after myocardial infarction, especially those 60 years or older.88 Proton-pump inhibitors are thus often prescribed prophylactically at the initiation of clopidogrel therapy although the rationale for this co-prescription is not fully validated.

Proton-pump inhibitors can alter the extent of drug absorption through modifying intragastric pH.89 Similar to clopidogrel, they share the same metabolic pathway in terms of hepatic metabolism.64,90

As shown in Table 2, PPIs are not only substrates,90 but also inhibitors of CYP2C19;64 therefore, those poor metabolizer (PM) patients with CYP2C19 loss-of-function alleles may not only have impaired formation of clopidogrel active metabolite but also the highest concentrations of omeprazole, a potential double hit. Recent mechanistic studies have shown that omeprazole, the most potent CYP2C19 inhibitors in clinical use, reduced the inhibitory effect of clopidogrel on platelet aggregation.91,92 Gilard et al. used the vasodilator-stimulated phosphoprotein phosphorylation (VASP) test as the index of platelet reactivity to clopidogrel and defined poor responders according to Barragan et al.93 criteria, in patients receiving omeprazole, 60.9% of patients were considered as poor clopidogrel responders compared with 26.7% in the placebo group (odds ratio 4.31, 95% CI 2.0–9.2). However, this interesting finding might be biased since the authors did not evaluate the percentage of clopidogrel non-responders before inclusion and did not exclude them. Because the primary hypothesis of the study was that omeprazole–clopidogrel drug–drug interaction is via a CYP2C19 competitive or non-competitive inhibitory mechanism, patient carriers of CYP2C19 loss-of-function alleles should have been excluded. Nevertheless, this study remains to date the only randomized placebo-controlled trial evaluating this important drug–drug interaction. Interestingly, some small studies have suggested that the PPI–clopidogrel interaction is not a class effect. Concomitant treatment with lansoprazole, pantoprazole, and esomeprazole did not alter the pharmacokinetic or pharmacodynamics of clopidogrel while omeprazole and rabeprazole appeared to interact94,95 (Table 3).

View this table:
Table 2

Common drug substrates and clinically important inhibitors of CYP2C19

CYP2C19 substratesCYP2C19 inhibitorsCYP2C19 inducers
Proton-pump inhibitors: omeprazole, esomeprazole, lansoprazole, rabeprazole, and pantoprazoleOmeprazole, esomeprazole, lansoprazole, rabeprazoleRifampicin
Antiprotease: Nelfinavir
Antiplatelet: clopidogrel, ticlopidineTiclopidine, clopidogrel
Anticonvulsivant: phenytoin, diazepamCarbamazepine
Anticancer: cyclophosphamide, tamoxifene
Antidepressants: amitriptyline, citalopram, clomipramine, sertralineFluvoxamine
View this table:
Table 3

Proton-pump inhibitors–antiplatelet agents drug–drug interaction studies

StudyDesignSubjectsnProton-pump inhibitorAntiplateletFunctional parameter and/or outcomeFollow-upResult
Small et al. 200894Prospective studyHealthy volunteers26Lansoprazole 30 mg (6 days)Clopidogrel 300 mg; prasugrel 60 mgInhibition platelet aggregation IPA7 daysLansoprazole + clopido: decrease IPA. Lansoprazole + prasugrel: no decrease
Gilard et al. 200692Observational studyPatients at high-risk coronary angioplasty.105OmeprazoleClopidogrel (dose NA); + aspirinVASP phosphorylation test Day 22 daysHigher VASP values in PPI users when compared with PPIs non-users
Gilard et al. 200891Prospective double blind, placebo, controlled, randomized OCLA studyUndergoing artery stent implantation124Omeprazole 20 mg/day or placeboClopidogrel (LD: 300 mg+MD: 75 mg/day)+ aspirinVASP phosphorylation test; Day 1; Day 77 daysOmeprazole decrease clopido inhibitory affect on platelet P2Y12
Pezalla et al. 200898Case–control studyAcute coronary syndrome1010All PPIsN.AIncidence of Acute MI1 yearAcute MI rates higher in the high PPI exposure group
Sibbing et al. 200999Cross-sectional observational study previous coronary stenCAD with previous PCI (median 7 months)1000PPI group n = 268; omeprazole n = 64; pantoprazole n = 162; esomeprazole n = 42Clopidogrel (MD: 75 mg/day) + aspirinAggregometry test (Multiplate analyser)7 monthsOmeprazole associated with an attenuated platelet response; no effect with pantoprazole and esomeprazole
Siller et al. 200995Non-randomized studyCAD undergoing PCI300No PPI n = 74; PPI group n = 226; pantoprazole n = 152; esomeprazole n = 74Clopidogrel (LD: 600 mg+MD: 75 mg/day) + aspirin (100 mg/day)VASP; Aggregometry test (Multiplate analyser) MI3 monthsNo effect
Juurlink et al. 200996Case–control retrospective studyFollowing acute myocardial infarction13636; controls 2057All PPIsClopidogrel (dose NA)Risk of reinfarction1 yearPPIs, other than pantoprazole, were associated with reduced beneficial effects of clopidogrel and an increased risk of reinfarction
Simon et al. 200912Cohort prospectiveFollowing acute myocardial infarction2208PPI group n = 1606; omeprazole n = 1147Clopidogrel (LD: 300 mg, MD: 75 mg/day) ± aspirinRecurrence of events1 yearNo effect
Chen et al. 2009151Randomized cross over trialHealthy volunteers CYP2C19 genotype12Omeprazole 40 mgClopidogrel [LD: 300 mg, MD: 75 mg/day (3d)]Pharmacokinetic4 daysAUC of omeprazole increased by 30.02% in EMs. No change in PMs.
Ho et al., 200997Cohort observationalAcute coronary syndrome8205; PPI group 5244All PPIsClopidogrel (dose NA)All-cause mortality, rehospitalization for ACSMedian 521 daysUse of PPIs associated with an attenuation of the clopidogrel efficacy
  • LD, loading dose; MD, maintenance dose; NA, not available; PPI, proton-pump inhibitor.

Five recent studies in large populations addressed the issue of clopidogrel–PPI interactions by examining their impact on the incidence of clinical events.12,9699 In a retrospective claims-based analysis, Pezalla et al.98 found a link between the PPIs use and the incidence of MI among patients aged below 65 years receiving clopidogrel. In the French FAST-MI registry, the use of PPIs had no impact on the clinical response of clopidogrel among the subgroup of 2208 AMI genotyped patients receiving clopidogrel.12 In contrast, a significant association was found between incidence of recurrent myocardial infarction within 90 days after discharge and current use of PPI (adjusted OR 1.27, 95% CI 1.03–1.57) in a Canadian nested case–control study.96 Treatment with pantoprazole, which does not potently inhibit CYP2C19, was not associated with recurrent infarction, whereas treatment with other PPI (omeprazole, lansoprazole and rabeprazole) was associated with reinfarction. However, neither major cardiac risk factors nor the use of over-the-counter medications, particularly aspirin, were taken into account in the multivariate analysis. The use of PPI was also associated with a higher risk for recurrent ACS (OR, 1.86; 95% CI 1.57–2.20) in a retrospective study of 8205 ACS patients receiving clopidogrel.97 The association was observed with both omeprazole (OR, 1.24, 95% CI 1.08–1.41) and rabeprazole (OR, 2.8, 95% CI 1.96–4.09). Unfortunately, the interaction with other PPIs (i.e. lansoprazole and pantoprazole) was not explored given the small numbers of patients. Finally, a possible ‘class effect’ for PPIs was outlined recently by Stanek100 who reported the findings, as a late-breaking clinical trial at the SCAI 2009 Scientific Sessions (unpublished data). They evaluated major cardiovascular events (MACE) among 16 700 patients, members of the Medco Health Solutions pharmacy, who received clopidogrel after a PCI. All PPIs were associated with a higher risk of MACE in clopidogrel users [hazard ratio: 1.51 (95% CI 1.39–1.64); P < 0.0001] (MACE rate: 25.1% for omeprazole, 24.9% for esomeprazole, 29.2% for pantoprazole, and 24.3%. lansoprazole) when compared with non-PPI users (17.9%). Further studies are needed to replicate these findings and determine the precise clinical impact of the drug–drug interaction in terms of benefit/risk considering the high rate of the co-prescription in North America96,97 and European countries.12,99 Moreover, it is noteworthy to underline that the clinical relevance for this co-prescription effect should be viewed with caution as the findings are from a single randomized clinical trial.101 In the latter, 123 patients with Helicobacter pylori infection and ulcer complications after using low-dose aspirin continuously for more than 1 month were randomized. The recurrence of ulcer complications during the 1 year follow-up was 14.8% compared with 1.6% in the placebo and lansoprazole groups, respectively (adjusted hazard ratio, 9.6; 95% CI 1.2–76.1). Therefore, prospective larger-scale studies are needed for evaluating the effectiveness of PPIs used concomitantly with clopidogrel and their potential class effects in terms of clinical outcomes Their design should include proper pharmacokinetics/pharmacodynamics investigations of different PPIs with clopidogrel and exclude or analyse separately those patients with CYP2C19 loss-of-function polymorphisms.

Determinants of antiplatelet therapy: genetic factors of variability


Aspirin covalently modifies both COX-1 and COX-2, although its affinity for COX-1 is 50 to 100 times greater than for COX-2. Importantly up to 40% of patients with cardiovascular disease do not comply with aspirin therapy.102 Incomplete platelet response to aspirin, likely reflects a composite of multiple processes. However, the mechanisms of aspirin resistance remain uncertain.103,104

From a pharmacological perspective, COX-1 is the key target for aspirin and non-selective non-steroidal anti-inflammatory drugs (NSAIDs). Genetic polymorphisms in enzymes involved in arachidonic acid metabolism (including COX-1), platelet glycoprotein, and collagen receptors have been identified. A clinical study in healthy volunteers showed that COX-1 genetic polymorphism (A682-G), which might affect enzyme expression, is present in 10% of the population.105 In patients taking aspirin for secondary prevention of CAD, genetic variability in COX-1 appears to have some impact on AA-induced platelet aggregation and thromboxane generation.106


As outlined earlier, there is growing evidence that a subtherapeutic response to clopidogrel may relate to altered pharmacokinetic parameters such as intestinal absorption and liver metabolic activation, both of which are affected by genetic polymorphisms. The impact of ABCB1 genetic polymorphism on clopidogrel clinical response was found recently in FAST-MI study.12 Patients with the ABCB1 3435TT genotype had a higher rate of cardiovascular events at 1 year than those with the ABCB1 wild-type genotype (adjusted HR, 1.72; 95% CI 1.20–2.47).12 Regardless of the exact link between the ABCB1 C3435T genetic polymorphism and P-glycoprotein expression, these results are consistent with a prior study showing lower plasma concentrations of clopidogrel and its active metabolite in patients carrying the ABCB1 3435TT genotype.41 However, as ABCB1 genetic polymorphism was not an independent predictor of outcomes in the large population of patients undergoing PCI, these results should be considered with caution until additional studies replicate the findings.

To become active, clopidogrel requires oxidation dependent on CYP as described previously. Although in vitro studies have shown that CYP3A4 was the major oxidative pathway for clopidogrel, CYP2C19 is now believed to be the major pathway in the bioactivation of clopidogrel as confirmed recently with pharmacodynamic or/and pharmacokinetic studies in healthy volunteers.14,73,107

CYP3A4 and CYP3A5 genetic polymorphisms

Most CYP3A4 variants are SNPs of low allelic frequencies, and many are population specific.108 However, because of their low allelic frequencies, their contribution to the interindividual variability of CYP3A4 expression is limited,109 although they may play a role in the atypical response to drugs such as clopidogrel.32 The impact of CYP3A5 genetic polymorphism on clopidogrel metabolism was controversial until recently. Suh et al.110 reported a higher frequency of atherothrombotic events within 6 months of coronary angioplasty in patients with the CYP3A5 non-expressor genotype (CYP3A5*3) receiving clopidogrel therapy. While others studies found no association between CYP3A5 genetic polymorphism and the antiplatelet effect of clopidogrel ex vivo both in patients111,112 and in healthy subjects.14,107,113 We confirmed the lack of association between CYP3A5 genetic polymorphism and major clinical outcomes at 1 year follow-up in the large-scale FAST-MI cohort12 (Table 4).

View this table:
Table 4

Genetic polymorphisms associated with platelet or antiplatelet drug responsiveness

StudyDesignSubjects or patientsnAntiplateletGene or allelic variantsFunctional parameter and/or outcomeEffect outcomeFollow-upResults
Fontana et al. 2003152Prospective studyHealthy98No drugP2Y12; GPIIb/IIIaADP-induced platelet aggregationPharmacodynamic7 daysADP-induced platelet aggregation is associated with a haplotype of P2Y12 receptor
Fontana et al. 2003124Case–controlPAD184No drugP2Y12αIIIbβ3PLA1/A2 α2β1NARisk of PADNDRole of H2 haplotype in atherosclerosis
Lau et al. 2004153Prospective studyHealthy25Clopidogrel; LD: 450 mgADP-induced platelet aggregation (before and 5 days after stent)Pharmacodynamic5 daysInterindividual variability in platelet inhibition which correlates with CYP3A4 activity: contribution to the clopido resistance
Healthy10Clopidogrel; MD: 75 mg (6 days) + rifampicin 300 mg × 2/day (4 days)CYP3A4 activity measured by Erythromycin Breath TestADP induced platelet aggregation (before and 4 h after LD)4 h
CAD32Clopidogrel LD: 300 mg; MD: 75 mg/dayADP-induced platelet aggregation.30 days
Hetherington et al. 2005123Prospective studySubject with no history of CAD200No drugP2Y1; P2Y12ADP-induced platelet aggregationPharmacodynamicNAP2Y1 variant associated with platelet reactivity to ADP
Angiolillo et al. 2006109Prospective studyPatients stable CAD82Aspirin + clopidogrel; MD: 75 mg/day); clopidogrel; LD: 300 mgCYP3A4ADP induced platelet aggregation; 2 h, 4 h after intake; ADP-induced platelet aggregation before 4 h, 24 h after LDPharmacodynamicCYP3A4 IVS10+12G>A modulates platelet activation
Naive patients scheduled coronary stenting45
Suh et al. 2006110Prospective cohortHealthy volunteers Koreans32Clopidogrel; LD: 300 mg; MD: 75 mg (6 days)CYP3A5*3ADP-induced platelet aggregationItraconazole interaction6 dayCYP3A5 expressor: change in platelet aggregation greater
Patients coronary angioplasty with stent348Pharmacodynamic atherothrombotic events6 monthsAtherotrombotic events occurred more frequently within 6 months after stent among CYP3A5 non-expressor
Hulot et al. 2006107Prospective studyHealthy volunteers28Clopidogrel; MD: 75 mg/day (7 day)CYP2C19; CYP2B6*5; CYP1A2*1F; CYP3A5*3Platelet aggregation (5, 10 µmol/L ADP); VASP phosphorylation testPharmacodynamic14 dayCYP2C19*2 is associated with a decrease in platelet responsiveness
Fontana et al. 2007154Prospective studyHealthy volunteers94Clopidogrel; LD: 300 mg; MD: 75 mg/day (7 day)CY2C19; CYP3A4 (IVS10+12G>A)ADP-induced platelet aggregation (20 µmol/L ADP)Pharmacodynamic8 dayNo association between CYP3A4 (IVS10+12G>A) and responsiveness; Association with CYP2C19*2
Giusti et al. 2007155Prospective studyPatients acute coronary syndrome1419Clopidogrel; LD 600 mg+500 mg aspirin IV followed by 75 mg clopido +100 mg aspirin /dayCYP2C19; CYP3A4/5; P2Y12; GpIa; GpIIIa; GpIb-alpha; GpVI; P-selectin; COX1/2Platelet aggregation (PRP: 2, 10 µmol/L ADP and AA); residual platelet; reactivityPharmacodynamic24 h after PCICYP2C19*2 associated with a higher platelet aggregability and RPR in high-risk vascular
Brandt et al. 2007112Prospective studyHealthy volunteers74Clopidogrel 300 mgCYP2C19; CYP1A2; CYP2B6; CYP3A4/5LTA (20 µmol/L ADP) 4 h after dosePharmacodynamic; pharmacokinetic1 dayLoss-of-function alleles CYP2C19 and CYP2C9 decreased metabolite of Clopidogrel but not prasugrel. Decrease pharmacodynamics response for Clopidogrel
71Prasugrel 60 mg
Kim et al. 2008113Prospective studyHealthy volunteers35Clopidogrel; LD 300 mg; MD 75 mg (6 day); metabolite SR26334CYP3A5ADP induced platelet aggregation (8 day just before the daily MD); pharmacokinetic (24 h after LD)Pharmacodynamic; pharmacokinetic8 dayCYP3A5 did not substantialially affect pharmacokinetic and pharmacodynamics effect of clopidogrel
Trenk et al. 2008118Prospective cohortPCI797Clopidogrel; LD 600 mg; MD 75 mg; aspirin 100 mg/day for at least 5 daysCYP2C19*2RPA (5 µmol/L ADP)Clinical outcome: death, non-fatal MI1 yearCarriers of at least one CYP2C19*2 allele are more prone to high RPA on poor clinical outcome after PCI
Geisler et al. 200815Prospective cohortCAD237Clopidogrel; LD 600 mgCYP2C19*2; CYP2C19*3; CYP3A4; CYP3A5RPA (20 µmol/L ADP) 6 h after LDPharmacodynamic6 hRisk for higher RPA increased with one CYP2C9*2 allele (OR: 3.71) and 2 variant (OR:10.72)
Taubert et al. 200841ProspectivePatients CAD percutaneous coronary intervention60Clopidogrel 300 mg and 600 mgMDR1 C3435TPharmacokineticClopido absorption and thereby active metabolite formation are diminished by Pgp influenced by MDR1 genotype
Mega et al. 200914Prospective studyHealthy volunteers162Clopidogrel; LD 300 mg or 600 mg; MD 75 mgCYP2C19; CYP1A2; CYP2B6; CYP3A4/5LTA (20 µmol/L ADP) 4 h after dosePharmacodynamic pharmacokinetic15 monthsReduced function CYP2C19 allele: lower levels of active metabolite; diminished platelet inhibition; higher rate of CV events, including stent thrombosis
ACS with PCI1477Clopidogrel; LD 300 mg; MD 75 mgCV events TIMI major and minor bleedingClinical outcome
Simon et al. 200912Prospective cohortPatients after AMI2208Clopidogrel; LD 300 mg; MD 75 mgCYP2C19; CYP3A5; P2Y12; ITGB3; MDR1 C3435TCV eventsClinical outcome1 yearCarriers of at least one CYP2C19*2 allele are higher risk bad outcome; TT 3435 bad outcome
Sibbing et al. 2009116ProspectivePatients CAD undergoing coronary stent2485Clopidogrel; LD 600 mg; MD 75 mgCYP2C19*2Stent thrombosis (ST)Clinical outcome: cumulative incidence of definite ST30 daysCYP2C19*2 associated with an increased risk of ST following coronary stent placement
Collet et al., 200913Prospective studyPatients (<45y) after AMI259Clopidogrel MD 75 mgCYP2C19CV eventsClinical outcome1.07 yearCYP2C19*2 major determinant in young patients
Mega et al. 2009125Prospective studyHealthy volunteers238Prasugrel; LD 60 mg; MD 10 mgCYP2C19; CYP1A2; CYP2B6; CYP3A4/5LTA (20 µmol/L ADP) 4 h after dosePharmacodynamic; Pharmacokinetic15 monthsNo effect on Pharmacodynamic pharmacokinetic response or clinical CV events rates in carriers vs. non-carriers of at least one loss function allele for any CYP
Patients; acute coronay syndrome; TRITON TIMI 381466CV events; TIMI major, and minor bleedingClinical outcome
  • LTA, light transmittance aggregometry; PCI, percutaneous coronary intervention; RPR, residual platelet reactivity; VASP-P, vasodilator-stimulate phosphoprotein phosphorylation; LD, loading dose; MD, maintenance dose; NA, not available; MACE, major cardiovascular events; CV, cardiovascular events; TIMI, Thrombolysis In Myocardial Infarction.

CYP2C19 genetic polymorphisms

Almost 25 genetic variants in CYP2C19 has been found www.cypalleles.ki.se, although only two (CYP2C19*2 and *3) account for more than 95% of cases of PM phenotypes. There are substantial differences in the prevalence of CYP2C19 polymorphisms among various population groups, as described in Table 3. Two to 3% of Caucasians and 4% of Africans have the PM phenotype, whereas 10–25% of Southeast Asians exhibit the PM phenotype.114 Recently, a new allele (CYP2C19*17) was described, and noted to be associated with an increased activity in vivo as measured by omeprazole and mephenytoin as probe drugs. The variant is fairly common among Caucasians and Ethiopians (18%)115 (Table 5).

View this table:
Table 5

Allele frequencies of CYP2C19*2 and *3 polymorphisms in various ethnic populations

PopulationSubject, n*1*2*3Study
 Caucasians, Germany3288415.90.3Aynacioglu et al.156
 Caucasians, Italy36088.911.10Scordo et al.157
 Caucasians, Turkey4048415.90.15Aynacioglu et al.156
 Caucasians, European-American21087130Ozawa et al.;158 Goldstein et al.159
 Caucasians, European-American54686.412.70.9Luo et al.160
 African American21675250Goldstein et al.159
 African American4728118.20.8Luo et al.160
 Bolivian77892.27.80.1Bravo-Villalta et al.161
 Ethiopian11486.413.60Persson et al.162
 Mexican Americans69290.29.70.1Luo et al.160
 Palestinian20091.35.83Sameer et al.163
 Saudi Arabian19485150Ozawa et al.158
 Native Canadian Indians11580.919.10Nowak et al.164
 Burmese12766304Tassaneeyakul et al.165
 Chinese2750.045.54.5Yamada et al.166
 Chinese Han40069.7324.673.27Chen et al.167
 Filipinos10454397Goldstein et al.159
 Iranian40086140Zand et al.168
 Indian-North20070300Lamba et al.169
 Indian-Tamilian11260382Adithan et al.170
 Japanese3061.827.410.8Takakubo et al.171
 Japanese106672310Ozawa et al.158
 Korean20667.520.911.6Herrlin et al.173
 Korean37764.228.37.6Lee et al.172
 Thai77468293Tassaneeyakul et al.165
 Southeast Asians16063.131.25.7Luo et al.160
 Vietnamese16568.826.44.9Lee et al.172
 Vietnamese90622414Yamada SJ et al.166

In healthy subjects, carriers of the defective CYP2C19 allele, are more likely to have an impaired antiplatelet activity.14,107 Moreover, they have significantly lower levels of the active clopidogrel metabolite and diminished platelet inhibition.14 The impact of CYP2C19 loss-of-function alleles on clinical outcomes has been recently evaluated in several studies.1214,116118 In the FAST-MI study,12 we found that patients carrying any two CYP2C19 loss-of-function alleles (*2, *3, *4, or *5) had a higher rate of death, recurrent MI or stroke, than patients with none (21.5 vs. 13.3%; adjusted hazard ratio, 1.98; 95% CI 1.10–3.58). Among the 1535 patients who underwent PCI during hospitalization, the rate of cardiovascular events among patients with two CYP2C19 loss-of-function alleles was 3.58 times the rate among those with none (95% CI 1.71–7.51). In contrast, patients with one CYP2C19 loss-of-function allele did not have an increased risk when compared with those who had no CYP2C19 variant alleles. Accounting for the presence of CYP2C19*17 had no significant effect on these risks. In FAST-MI registry, the loading dose of clopidogrel was 300 mg and the mean daily dose was 75 mg/day. In a German cohort of patients undergoing coronary stent placement after pre-treatment with 600 mg of clopidogrel, the risk of stent thrombosis at 30 days was increased in CYP2C19*2 allele carriers (*1/*2 or *2/*2) with the highest risk in patients with the CYP2C19 *2/*2 genotype.118 Among clopidogrel-treated subjects in TRITON–TIMI 38,14 carriers of one or both CYP2C19 loss-of-function alleles had increased risk of cardiovascular events when compared with non-carriers (HR = 1.53; 95% CI = 1.07–2.19) and stent thrombosis (2.6 vs. 0.8%; HR = 3.09; 95% CI 1.19–8.00). Unfortunately, the authors did not evaluate separately the impact of one or two variants alleles on outcome. This is an important issue considering the percentage of patients involved. Further studies are needed before drawing a definite conclusion of the range of patients at high risk of events.

Clopidogrel targets: GIIbIIa, P2Y12

Marked variations reported in the concentration of ADP required to produce irreversible aggregation have been reported suggesting a possible genetic determinant of the ADP effect on aggregation. The effect of ADP on platelets is mediated by two P2Y receptors, designated P2Y1 and P2Y12. Both are heterotrimeric G-protein coupled receptors: P2Y1 to Gq and P2Y12 to Gi. Stimulation at P2Y1 leads to intracellular calcium mobilization and change in platelet shape,119 whereas stimulation at P2Y12 leads to inhibition of adenyl cyclase120 and activation of phosphoinositide 3 kinase.121 The net effect is the modulation affinity of the glycoprotein Iib–IIIa (GPIIb–IIIa).122

Among different genetic polymorphisms observed in Caucasians with no history of coronary heart disease and no antiplatelet medication, P2Y1A1622G polymorphism was found to have a significant association with platelet response to ADP, as defined by the binding of fibrinogen to activate GPIIb–IIa.123 For P2Y12, some genetic polymorphisms defined as the haplotype H2 has been found to be strongly associated with increased ADP-induced platelet aggregation in healthy volunteers.124 Most studies have evaluated the impact of the pharmacological parameters of clopidogrel on biological platelet function. Their clinical impact was not confirmed in FAST-MI registry, the single study to date evaluating this hypothesis in AMI patients.12 These clinical data outline again the fact that results should be viewed with caution and considered exploratory findings that need to be replicated.

Prasugrel is a novel and potent thienopyridine that targets the same P2Y12 ADP receptor as clopidogrel. Unlike clopidogrel, conversion of prasugrel to its active metabolite involves rapid hydrolysis by esterases followed by a single CYP-dependent step. Prasugrel is absorbed rapidly after dosing with concentrations of its active metabolite peaking ∼30 min after dosing. On a molar basis, the active metabolites of clopidogrel and prasugrel are equipotent platelet inhibitors.94 Interestingly, the pathway leading to the conversion of prasugrel and clopidogrel to their respective active metabolites differs. Prasugrel is rapidly hydrolysed by esterases to an inactive thiolactone, which is then metabolized by CYPs to the active metabolite. The responsible enzymes appear to be CYP3A4 and CYP2B6 and to a lesser extent, CYP2C9 and CYP2C19.52

CYP2C19 genetic polymorphisms do not affect prasugrel pharmacodynamics and pharmacokinetic parameters in healthy subjects112 (Table 4). Moreover, similar rates of cardiovascular events were observed in TRITON-TIMI 38 trial among ACS patients who were carriers and non-carriers of a CYP2C19 loss-of-function allele, treated with prasugrel.125

Surprisingly, in a study with healthy subjects, lansoprazole slightly reduced the plasma level of prasugrel active metabolite without affecting the inhibition of platelet aggregation.94 A single loading dose of prasugrel 60 mg associated with or without lansoprazole 30 mg was used in this study, with a 7-day run-in period of IPP prior to receiving the loading dose (Table 3). Therefore this result should be taken cautiously and needs confirmation with longer exposure and follow-up.

Other novel antiplatelets with promising and less dependent on hepatic metabolism, are still in development with currently ongoing clinical trials.126

Perspective of personalized medicine in antiplatelet therapies

Observational studies dating back the late 1940s onwards have unravelled the key factors that influence risk for CVD such as cigarette smoking, cholesterol levels, diabetes, and blood pressure. In addition, progress in the treatment of cardiovascular disease relates in part to greater knowledge of platelet function and the benefits of antiplatelets drugs.

However the extent of variability in response to antiplatelet drugs is proving to be a clinical problem. This is further compromised by the lack of an assay with a sufficient accuracy and predictive value in terms of platelet aggregation and clinical outcome. The promising P2Y12 assay (VerifyNow, Accumetrics Inc.) has a positive predicted value of 12% to detect ACS patients at risk of 12 month cardiovascular events (Table 1).127 Thus the majority of patients with a positive test will not experience an ischaemic event. The results of the ongoing studies, such as GRAVITAS128 will help to examine whether tailored clopidogrel therapy, using a point-of-care platelet function assay, may reduce major adverse cardiovascular events after PCI.

Interestingly, genetic variations in the pathways which govern drug metabolizing enzymes are proving to be quite relevant to clopidogrel antiplatelet therapy. Indeed genetic testing could be a new tool for identifying patients at higher risk of events.

Identification of patients at ‘higher or lower risk of poor clopidogrel responsiveness’ defined as carriers or non-carriers of CYP2C19 loss-of-function alleles may help to better optimize the choice of the antiplatelet drug. As an example, for the treatment of peptic ulcer disease, clinical pharmacologists have already begun modelling the economic utility of CYP2C19 genotyping prior to prescribing PPIs. Considering a maximum treatment duration of 3 months and an estimated genotyping cost of 10 USD per allele, investigators projected a cost saving of >5000 USD per 100 Asian patients genotyped. Due to ethnic variation in allele frequency, cost saving was lower in other populations129 but remained significant in patients of European descent.130 Therefore it is probable by extrapolation that in ACS patients who undergo a PCI targeting antiplatelet treatment by genotyping would probably be a cost-effective strategy. The higher benefit/risk ratio of prasugrel seems to be particularly relevant in those patients at ‘higher risk of clopidogrel poor response’ but is not conclusive for those patients who do not carry any CYP2C19 loss-of-function allele. In TRITON-TIMI 38, the rate of cardiovascular events was 9.8% for CYP2C19 non-carriers in the prasugrel group125 and 8.5% in those receiving clopidogrel during the trial follow-up.14 Thus although there was no planned head-to-head comparison with regard to genotype data, current available results suggest that clopidogrel (300 mg LD and 75 mg thereafter) may remain the drug of choice in terms of benefit/risk and benefit/cost ratios among those homozygous CYP2C19 wild-type patients representing the majority of treated patients. In contrast, in those patients carrying the two loss-of-function variant alleles of CYP2C19, prasugrel may be preferred over clopidogrel.

However, the positive predictive value of CYP2C19 loss-of-function genetic variants is not optimal particularly among heterozygous subjects. Further studies are necessary for evaluating whether combining laboratory assay and genotyping may enhance the predictability of clopidogrel non-responsiveness among heterozygous patients.

The comparison of the effects of prasugrel and clopidogrel among heterozygous CYP2C19 loss-of-function patients were not shown in TRITON-TIMI 38,14,125 whereas in FAST-MI, this population receiving clopidogrel were not at higher risk of events compared with homozygous wild-type patients.12 Therefore, among heterozygous patients, the use of prasugrel or a higher dose of clopidogrel should be discussed on an individual basis with regard to the benefit/bleeding-risk ratio. Larger prospective randomized clinical trials are needed to confirm these hypotheses.


Great hope has been expressed towards the development of personalized medical care strategies in terms of appropriate diagnosis, treatment, and CVD prevention. The issue of validated point-of-care testing and their ability to predict clinical outcomes remains unresolved for antiplatelet drugs. Recent research findings highlight the role of genetic variation as an important variable for optimizing the response to antiplatelet drugs such as clopidogrel. The goal of personalized medicine is to utilize in part the person's genetic makeup for selecting the best drug and dose. In addition, this approach should also include the impact of important non-genetic factors, such as the clinical status of the patient, the environmental factors including diet, and drug–drug interactions.


This work was supported in part by the Academic Medical Organization of Southwestern Ontario AHSC AFP Innovation Fund (RBK) and Post Doctorat 2007 Leem Research Fund, France (CV).

Conflict of interest: T.S. has received consulting fees from Bayer-Schering, Pfizer, sanofi-aventis, and Eli Lilly, lecture fees from Bayer-Schering, and grant support from Pfizer and Servier.