Response variability to clopidogrel: is tailored treatment, based on laboratory testing, the right solution?



    1. Dipartimento di Medicina, Chirurgia e Odontoiatria, Università degli Studi di Milano, Unità di Medicina 3, Ospedale San Paolo, Milan, Italy
    Search for more papers by this author

Marco Cattaneo, Dipartimento di Medicina, Chirurgia e Odontoiatria – Università degli Studi di Milan, Unità di Medicina 3, Ospedale San Paolo, Via di Rudinì 8, 20142 Milan, Italy.
Tel.: +39 2 50323095; fax +39 2 50323089.


Summary.  Clopidogrel is an antithrombotic prodrug, whose active metabolite inhibits platelet function by irreversibly binding to the platelet receptor for adenosine diphosphate, P2Y12. Wide inter-individual variability of response to clopidogrel has been reported in several studies: a significant proportion of treated patients (about one-third) exhibit a suboptimal inhibition of platelet function. Genetic and environmental factors that influence the absorption and/or the extent of metabolism of clopidogrel to its active metabolite account for the observed variability of response. Tailored treatment based on the results of laboratory tests of platelet function has been proposed as a solution to this problem, which has important clinical implications. Although it is often considered a desirable evolution of modern medicine, tailored treatment based on laboratory tests is actually an old remedy (of yet unproven efficacy, in the case of antiplatelet therapy) for the problem of response variability to antithrombotic drugs with unpredictable bioavailability. When possible, the use of alternative drugs with more uniform and predictable bioavailability, and with favourable profiles in terms of risk/benefit and cost-benefit ratios should be preferred. Moreover, tailored treatment with laboratory tests must be validated in randomized clinical trials before its implementation can be recommended. We still need to identify and standardize the laboratory test for this purpose, as well as answer basic questions on its clinical utility and cost-effectiveness, before tailoring clopidogrel therapy based on laboratory tests can be recommended in clinical practise.

P2Y12 is one of the two platelet receptors for adenosine diphosphate (ADP), which plays a key role in the pathogenesis of arterial thrombi: it mediates ADP-induced platelet aggregation, the potentiation of platelet secretion induced by strong agonists, the stabilization of thrombin-induced platelet aggregates, shear-induced platelet aggregation and the inhibition of the antiplatelet effects of the natural regulator of platelet function prostacyclin [1]. These characteristics and its selective tissue distribution make it an attractive molecular target for therapeutic intervention. Indeed, P2Y12 is the target of efficacious antithrombotic agents, including the thienopyridines (ticlopidine, clopidogrel and prasugrel) and ticagrelor, which belongs to the new chemical class of cyclopentyl-triazolo-pyrimidines [2].

Inter-individual variability in the pharmacological response to clopidogrel

Clopidogrel is effective in secondary prevention of coronary artery and cerebrovascular events, and, in combination with aspirin, is the mainstay in the prevention of major adverse cardiovascular events (MACE) in patients with acute coronary syndromes [2]. However, its clinical utility is hampered by the high inter-individual variability of inhibition of P2Y12-dependent platelet function, which is mostly caused by the variable bioavailability of its active metabolite [3,4]: about 25% of treated patients display suboptimal response to the drug [3,4].

Like the other thienopyridines, clopidogrel is a prodrug that needs to be metabolized to an active metabolite to exert its pharmacological effect. The formation of the active metabolite of clopidogrel involves a two-step process, which is regulated by isoforms of the hepatic cytochrome P450 (CYP). CYP2C19, CYP1A2 and CYP2B6 are responsible for the first metabolic step, whereas CYP2C19, CYP2C9, CYP2B6 and CYP3A are responsible for the second step [2,3] (Fig. 1). Loss-of-function (e.g. CYP2C19*2 and CYP2C19*3) and gain-of-function (e.g. CYP2C19*17) genotypes are associated with variable degrees of production of the active metabolite and, hence, of the pharmacodynamic response to the drug [2,3]. However, it has been hypothesized that the impaired pharmacodynamic response to clopidogrel in association with CYP2C19*2 may be due to an imbalance in the formation of pro-inflammatory and anti-inflammatory cytokines, which could contribute to altered platelet aggregability, rather than to impairment of the formation of the active metabolite of clopidogrel [5]. The same authors identified paraoxonase-1 (PON-1) as the crucial enzyme for clopidogrel bioactivation [6], but this finding was not confirmed by several subsequent studies [7–13].

Figure 1.

 Metabolism of clopidogrel. Schematic representation of the main enzymatic pathways involved in the metabolism of clopidogrel. CYP, cytochrome P450 isoenzyme; ES, carboxylesterase; BChE, butyrylcholinesterase.

Variable levels of active metabolite generation and/or of pharmacodynamic response to clopidogrel are also associated with: (i) limited intestinal absorption, which is associated with the homozygous 3435C→T mutation of ATP-binding cassette subfamily B member 1 (ABCB1), a gene encoding for the efflux pump P-glycoprotein, a key protein involved in thienopyridine absorption; (ii) interaction with other drugs, including some proton pump inhibitors (PPIs), calcium channel blockers and lipophilic statins, which are metabolized by CYP2C19 and CYP3A isoenzymes; (iii) stimulation of CYP1A2 activity by tobacco smoking; and (iv) pre-existent variability in platelet response to ADP [2,3]. Other variables that influence the response to clopidogrel include advanced age, high body mass index, diabetes mellitus and renal insufficiency in diabetes mellitus, which are associated with decreased response to the drug (Table 1) [2,3]. Needless to say, non-compliance is to be considered an obvious and frequent cause of poor response to clopidogrel [14].

Table 1.  Main variables affecting the pharmacodynamic response to clopidogrel
  1. ABCB1, ATP-binding cassette subfamily B member 1; CYP, cytochrome P450; ADP, adenosine diphosphate.

Lack of compliance
Reduced absorption (e.g. in carriers of the TT3435 mutation of ABCB1, encoding for P-glycoprotein)
Loss-of-function or gain-of-function mutations of CYP2C19 (and other CYP isoforms)
Interaction of other drugs (proton pump inhibitors, lipophilic statins, calcium channel blockers)
High body mass index
Diabetes mellitus
Renal insufficiency in diabetes mellitus
Pre-existent variability in platelet response to ADP
Increased platelet turnover (theoretical)
Tobacco smoking (heightened response)

Clinical consequences of the inter-individual variability of response to clopidogrel

Several independent studies demonstrated an association between suboptimal generation of the active metabolite of clopidogrel, decreased inhibition of platelet function, presence of enzyme polymorphisms and clinical outcomes [2,3,15,16]. However, no study has yet associated all of these parameters in the same patient population, and some uncertainties still persist. For instance, no clear negative association with clinical outcomes of the co-administration of clopidogrel with drugs that potentially interfere with its metabolism has been documented so far, despite their negative interaction with the pharmacodynamic response to the drug [17,18]. This is true not only for lipophilic statins and calcium channel blockers [3,17], but also for omeprazole and other PPIs interfering with CYP2C19. The evidence of the negative interaction of some PPIs with the pharmacodynamic response to clopidogrel activity [19–22], albeit controversial [23], and the demonstration in observational studies, case-control studies and post-hoc analyses of some randomized clinical trials that the risk of MACE was higher in patients on combined treatment with clopidogrel and a PPI, compared with patients on clopidogrel not in combined treatment with a PPI [24–26], prompted drug regulating authorities to issue a warning for the use of these drugs in combination with clopidogrel [27]. However, the existence of this negative interaction stemming from randomized clinical trials was less evident [24,28–32]. Moreover, the only randomized controlled trial that prospectively tested the interaction between omeprazole and clopidogrel, which was terminated prematurely due to bankruptcy of the sponsor, failed to show that the co-administration of the two drugs increases the incidence of MACE [33]. A third, more recent meta-analysis indicated an obvious discrepancy between the negative results of randomized clinical trials and the positive results of observational studies, concerning the clinical consequences of the negative pharmacological interaction of PPIs with clopidogrel [18].

The association between poor clinical outcomes of patients on treatment with clopidogrel and the presence of loss of function mutations of CYP has been demonstrated in observational and intervention studies. However, it must be emphasized that the testing and validation of statistical hypotheses in genetic epidemiology is a task of unprecedented scale [34]. Ioannidis et al. [34] showed that significant between-study heterogeneity is frequent, and that the results of the first studies (usually suggesting a strong genetic defect) correlate only modestly with subsequent research on the same association. Both bias and genuine population diversity might explain why early association studies tend to overestimate the disease protection or predisposition conferred by a genetic polymorphism [34]. As a matter of fact, three meta-analyses, which included the early published studies, demonstrated an increased risk of MACE and, particularly, of stent thrombosis in carriers of either one or two mutated CYP2C19*2 alleles [24,35,36], while two more recent meta-analyses did not indicate a substantial or consistent influence of loss-of-function CYP2C19 gene polymorphisms on the clinical efficacy of clopidogrel [37,38]. In line with these results, a very recent meta-analysis showed that, although there was an association between the CYP2C19 genotype and clopidogrel responsiveness, overall there was no significant association of genotype with MACE [39]. The different results of these latter meta-analyses are likely to be due to the fact that they included studies published after the year 2010 (which showed weaker effects compared with the previous ones) [37,38] and extracted only data of prespecified clinical events that conformed to unbiased and standardized definitions [37]. Moreover, the meta-analysis by Zabalza et al. [38] showed that a significant association between the loss-of-function alleles and a higher risk of cardiovascular outcomes was demonstrated by small-size studies (<500 patients), whereas no significant effect was observed in the pooled analysis of studies with sample sizes >500 patients. The overestimation of the effect size in small genetic association studies is well known and could be related to different factors such as spurious association due to biased publication of positive results in these types of studies [40]. Interestingly, the gain of function genotype CYP2C19*17 was associated with lower incidence of MACE [38].

Finally, the question remains open as to whether the association that was found in some studies is explained by impaired clopidogrel metabolism by carriers of the mutation, or by pleiotropic effects, with negative impact on cardiovascular outcomes, independently of the administration of clopidogrel [41].

In patients treated with clopidogrel, the ABCB1 3435C→T genotype was significantly associated with the risk of cardiovascular death, myocardial infarction or stroke in the TRITON TIMI-38 trial, which compared clopidogrel and prasugrel in patients with acute coronary syndromes undergoing percutaneous coronary intervention (PCI) [42].

In conclusion, there still are contrasting reports in the literature, linking negative interactions with the pharmacodynamic response to the drug and negative interactions with clinical outcomes. These uncertainties are likely to be due to inaccuracies and lack of standardization of the methodological approaches that have been used to detect ‘non-responders’ to clopidogrel. Additionally, uncertainties regarding how to interpret results, stemming from the absence of universally accepted and validated cut-off values of platelet function, may also cause inaccuracy. Indeed, it would appear unquestionable from a logical standpoint that an antithrombotic drug (and any drug in general) that is unable to hit its pharmacological target is also expected to be unable to be fully clinically effective. Therefore, independent of the uncertainties stemming from the published studies, efforts aimed at obtaining good inhibition of P2Y12-dependent platelet function in all treated patients are justified. The most commonly proposed and widely tested approach has been to tailor treatment with clopidogrel to each individual patient, based on the results of platelet function tests, used to monitor the individual pharmacodynamic response to the drug.

Is tailored treatment with clopidogrel based on platelet function testing a desirable solution to the problem of the wide inter-individual variability of response to the drug?

Since their introduction to clinical practise, antiplatelet drugs have been administered to patients at standard doses, without monitoring their pharmacological effects by means of laboratory tests. Although tailored treatment with clopidogrel based on platelet function testing is often considered a desirable evolution of modern medicine (which, ideally, should be personalized based on individual needs), it is actually an old remedy (of yet unproven efficacy, in the case of antiplatelet therapy) for the problem of response variability to antithrombotic drugs. The aim of personalized medicine is to optimize the risk/benefit and the cost/benefit ratios of healthcare by identifying ‘a priori’ for each individual patient, the best possible management. This should include the choice between treatment with the most effective and safe drug and no treatment at all (many patients would not develop clinical events even if left untreated). In contrast, tailored treatment, based on the results of laboratory tests, simply aims to adjust ‘a posteriori’ the dose of a drug with unpredictable and variable bioavailability. Treatment with vitamin K antagonists (VKAs) and with unfractionated heparin has always been tailored to the individual patient, based on laboratory monitoring, because the bioavailability of these drugs is unpredictable and highly variable [43]. However, laboratory monitoring is expensive, increases the workload of healthcare personnel, may be inaccurate and imprecise, is uncomfortable for patients, and may decrease their adherence to treatment [44]. For the above reasons, treatment with anticoagulant drugs is progressively disposing of laboratory monitoring due to the introduction of new drugs with very good and predictable bioavailability, such as low-molecular-weight heparins, which do not need laboratory monitoring, and have progressively replaced unfractionated heparin [45]. Therefore, it appears that antiplatelet therapy is heading in the opposite direction when compared with anticoagulant therapy.

In conclusion, given its many drawbacks, laboratory monitoring of clopidogrel therapy should be taken into consideration as a means of solving the problem of the high inter-individual variability of pharmacological response to the drug. However, this should only be considered if easier solutions are unavailable, such as the use of alternative drugs with more uniform and predictable bioavailability, and with favourable profiles in terms of risk/benefit and cost/benefit ratios.

If tailored treatment based on platelet function testing is the only solution to control for the inter-individual variability in response to clopidogrel, should this be adopted into clinical practise?

For tailored treatment with clopidogrel to be adopted into clinical practise, it should undergo the same validation process that was undertaken for laboratory monitoring of VKAs. This included: (i) identification of the most sensitive and specific laboratory test to identify accurately those patients who exhibit an abnormal response to the drug; (ii) identification of cut-off values for both the risk of thrombosis and the risk of bleeding (‘therapeutic window’); (iii) standardization of both the pre-analytical and the analytical conditions of the laboratory test; and (iv) identification of efficacious, safe and cost-effective treatments for patients whose values fall outside the ‘therapeutic window’.

Identification of the most sensitive and specific laboratory test to accurately identify patients that exhibit an abnormal response to the drug.

Various in vitro techniques have been used to measure the inhibition of platelet function by clopidogrel and, in some instances, to predict the risk of MACE [4,46] (Table 2). Although this approach is justified and rational, it should be considered that the relative importance of the ADP/P2Y12 pathway in platelet activation varies considerably among different subjects, as well as with the type of laboratory test used [4,14]. Therefore, the finding of high, residual platelet reactivity in vitro in patients on clopidogrel may not necessarily imply that these patients are poor responders to treatment, unless platelet function is measured by laboratory tests that are highly specific for the ADP/P2Y12 pathway [4].

Table 2.  Main laboratory tests that have been used to measure the degree of inhibition of platelet function by clopidogrel
  1. For details about these tests, see references [4,46].

Light transmission aggregometry
Impedance aggregometry
Platelet VASP®
ADP-stimulated formation of platelet-leukocyte hetero-aggregates, expression of activated αIIbβ3, or expression of P-selectin on the platelet surface (flow cytometry)
Thromboelastography Platelet Mapping System
IMPACT cone-and-plate(let) analyzer

ADP-induced platelet aggregation, measured with light transmission aggregometry or by impedance aggregometry may overestimate the prevalence of poor responders to P2Y12 inhibitors, because ADP induces platelet aggregation by interacting with both its platelet receptors, P2Y1 and P2Y12, the relative contribution of which to the overall platelet response is variable among different subjects [2]. The platelet aggregation-based assay VerifyNow (Accumetrics, San Diego, CA, USA) P2Y12 and the flow cytometry-based assay Platelet VASP® (Diagnostica Stago, Paris, France) (which measures the P2Y12–dependent inhibition by ADP of phosphorylation of vasodilator-stimulated phosphoprotein, VASP), are more specific assays for measuring the effects of thienopyridines and other drugs inhibiting the platelet P2Y12 receptor [4].

Many studies compared the performance of different laboratory tests for evaluating the pharmacological effects of clopidogrel. All studies demonstrated that the degree of agreement among different tests is unacceptably low [47–54]. For example, the same patients could be considered poor responders by one test and good responders by another test, and vice-versa. Therefore, as no single laboratory test has been shown to be more accurate than any other in measuring the degree of inhibition of ADP/P2Y12-dependent platelet function by clopidogrel, the equivalent of the prothrombin time (PT, the gold standard for monitoring VKA therapy) for treatment with clopidogrel has not yet been identified.

Identification of cut-off values for both the risk of thrombosis and the risk of bleeding (‘therapeutic window’)

Inhibition of the hemostatic system by anticoagulant or antiplatelet drugs decreases the risk of thrombosis, but also inevitably increases the risk of bleeding. Therefore, when monitoring an antithrombotic treatment, it is necessary to identify not only a cut-off value for the degree of inhibition of hemostasis below which protection from MACE is suboptimal, but also the cut-off value above which the risk of bleeding is high. Indeed, reduction of the risk of bleeding is of paramount clinical importance as severe bleeding associated with antithrombotic therapy has severe consequences. This is not only because such events may be fatal, disabling and expose the patients to the risks that are associated with blood transfusion, they are also associated with poor prognosis of the patients, whose risk of death is increased during follow-up [55–57]. The cut-off values of this therapeutic window for the degree of anticoagulation that can be achieved with VKAs have been validated and are universally accepted by all laboratories worldwide. In contrast, only recently has this important clinical issue started to be addressed for some tests of platelet function used for tailoring clopidogrel therapy [3,15,58]. However, the cut-off values for both the risk of thrombosis and the risk of bleeding are far from being universally accepted (see later).

Standardization of both the pre-analytical and the analytical conditions of the laboratory test

Standardization of laboratory tests is essential if one is to achieve comparable results in different laboratories. This is an absolute requirement for the implementation of laboratory monitoring in clinical practise. An example of the importance of laboratory standardization is provided by the evolution of VKA monitoring with the PT. Monitoring problems arose, especially in the USA, from the introduction of some poorly responsive commercial tissue extracts used as the thromboplastin reagent in the PT [43]. More oral anticoagulant drug was then needed to prolong the test to the required therapeutic targets, with a resultant increase in bleeding [43]. It was not until 1983 that the problem was resolved thanks to the PT standardization scheme of the World Health Organisation, using the international normalized ratio (INR), which allowed the widespread adoption of a ‘low-dose warfarin’ regimen, leading to improved effectiveness and safety of VKAs [43].

Many pre-analytical and analytical variables affect the results of light transmission aggregometry, the standardization of which poses several problems [59,60]. As a consequence, the results obtained within one laboratory cannot be compared with those obtained in a different laboratory. Therefore, any attempt to define a universal therapeutic window of platelet aggregation to monitor clopidogrel treatment would be pointless.

Analytical variables may be more easily standardized for other laboratory techniques, especially for those that rely on commercially available kits that need no, or very limited, handling of blood samples, such as the VerifyNow-P2Y12 and the Platelet VASP® assay. However, for these techniques the following pre-analytical variables also need to be standardized.

Many studies have shown that platelet function in patients receiving clopidogrel is a rather unstable parameter, with up to ∼50% of treated patients switching between the category of ‘poor responders’ and that of ‘good responders’ at different times after the start of therapy [15,61–63]. This variability in results of platelet function tests casts doubts about their accuracy and points to the need for repeated testing during the follow-up of the patients. The consequence of this is increasing cost and complexity of patient management. In addition, there is no clear information on when platelet function testing provides the best prediction of clinical events. Campo et al. [15] recently showed that platelet reactivity at day 30 after the start of treatment best predicts both ischemic and bleeding events. However, one wonders whether earlier information might be more clinically useful.

The time elapsed since the last dose of clopidogrel and blood sampling is also expected to affect the results of any platelet function test, especially in patients with coronary artery disease whose increased platelet turnover accelerates the entry of newly formed, non-inhibited platelets into the circulation.

Finally, the time of day at which blood is sampled may significantly influence results as platelet function follows a circadian rhythm, being highest at mid-morning [64]. Consistent with this hypothesis, Kozinsky et al. [65] showed that, among patients under maintenance treatment with 75 mg day−1 clopidogrel, which was administered at 08.00, the highest prevalence of ‘poor responders’ (20.3%) was observed at 10.00 compared with 8.5% at 06.00 and 10.2% both at 14.00 and 19.00 (P < 0.02).

Identification of efficacious, safe and cost-effective treatments for patients whose values fall outside the ‘therapeutic window’

Preliminary experiments aimed at tailoring clopidogrel treatment based on the results of laboratory tests gave results that were rather unsatisfactory. Bonello et al. [66,67] identified patients with acute coronary syndromes scheduled for PCI, who were considered resistant to a loading dose of 600 mg clopidogrel, based on the results of the Platelet VASP® assay. These patients were randomized to undergo either Platelet VASP®-guided additional loading doses of 600 mg clopidogrel until they reached adequate inhibition of P2Y12 function, or to have no further treatment (control group). Some patients in the Platelet VASP®-guided group achieved this goal after repeated clopidogrel doses, but about 10% of these were still resistant after a total of 2400 mg clopidogrel (32 pills) [66,67]. Although the rate of MACE [66] or stent thrombosis [67] at 30 days was lower in the Platelet VASP®-guided group than in the control group, it appears that the tailored treatment approach that was used in this study was far from ideal as it was cumbersome, time consuming, expensive and, most importantly, ineffective in many patients, despite multiple interventions to correct the dose of the drug. In addition, one wonders whether resistant patients who, after many loading doses of clopidogrel, did eventually display an adequate response to the drug, maintained a satisfactory inhibition of platelet function when given the much lower daily maintenance dose of clopidogrel that was administered thereafter. This concern has been further substantiated by another study that showed how difficult it may be to override clopidogrel resistance through increasing the maintenance doses of the drug. Some patients remained resistant to maintenance daily doses of 300 mg clopidogrel, which could not be continued due to the occurrence of severe side-effects (e.g. stomach discomfort and joint pain) [68]. It was only after the administration of regular maintenance doses of prasugrel that these patients exhibited adequate inhibition of P2Y12-dependent platelet function [68]. In conclusion, the Bonello et al. studies represent important proof-of-concept data as they demonstrate that the improvement of the pharmacological response to clopidogrel may improve its clinical efficacy. However, the management strategy that was employed in these studies can hardly be transferred into clinical practise.

Until the results of large-scale trials of personalized antiplatelet therapy are available, the routine use of platelet function measurements in the care of patients with cardiovascular disease cannot be recommended. This is based on the considerations given above, and is in compliance with the rules of Evidence Based Medicine, guidelines of Scientific Societies and a recent consensus paper [3,69,70].

Our inability to personalize antiplatelet treatment has been recently underscored by the negative results of GRAVITAS, the first, large randomized prospective trial testing the efficacy and safety of tailored clopidogrel treatment in patients undergoing PCI [71]. GRAVITAS enrolled 5429 patients with coronary artery disease (predominantly, albeit not uniquely, with stable disease), who were treated with a loading dose of clopidogrel (600 mg) before undergoing PCI with stent implantation. Platelet function was measured by the VerifyNow P2Y12 test 12–24 h after PCI, which identified 2214 patients with high on-treatment platelet reactivity (Platelet Reactivity Units [PRU] ≥ 230). These patients were randomized to standard treatment (daily maintenance dose of 75 mg clopidogrel) or to high-dose clopidogrel (additional loading dose, plus daily maintenance dose of 150 mg) for 6 months. The primary efficacy end-point was the incidence of cardiovascular death, acute myocardial infarction or stent thrombosis, the safety end-point was the incidence of severe or moderate bleeding, based on the GUSTO definition, while the pharmacodynamic end-point was the prevalence of persistent high on-treatment platelet reactivity during the follow-up of the study. The results of the GRAVITAS trial showed that treatment with high-dose clopidogrel of patients with high on-treatment platelet reactivity after a loading dose of the drug did not reduce the incidence of MACE compared with the standard dose (2.3% in both groups, P = 0.97), nor did it increase the incidence of bleeding (1.4% and 2.3%, P = 0.1) [71]. The prevalence of high on-treatment platelet reactivity during follow-up decreased significantly more in the high-dose group (−62%), when compared with the standard dose group (−49%, P < 0.001). Moreover, the incidence of the primary efficacy end-point among patients treated with standard dose clopidogrel was not significantly different from that of a group of 586 patients who were randomly chosen among the 3215 who showed a satisfactory inhibition of platelet function after the loading dose of clopidogrel, and were treated with a standard dose of the drug [71]. These last findings emphasize the dubious association between the results of platelet function tests in patients having treatment with clopidogrel and clinical outcomes, as mentioned earlier in this review.

In conclusion, (i) the negative results of the GRAVITAS trial confirm and emphasize that, in compliance with the rules of Evidence Based Medicine, no treatment with proven efficacy and safety should be replaced by new treatments, even if theoretically more rational, prior to demonstration of their efficacy, safety and favourable cost-benefit ratio; (ii) the adoption of the personalized treatment strategy that has been tested in the GRAVITAS trial would cause (and has already caused, in the institutions that have adopted it) an unjustified expenditure of resources without translating into any patient benefit.

As part of the debate that followed the publication of the results of the GRAVITAS trial, many criticisms were raised about its design, which included the type of patients enrolled, the choice of platelet function test, the timing and/or the frequency of the test, and the type of pharmacological intervention [72,]. Even the choice of the cut-off value of PRU has been criticised, despite the fact that it was chosen based on the results of previous observational studies [71]. This is a further demonstration that the cut-off value for high on-treatment platelet reactivity for clopidogrel has not yet been clearly validated and universally accepted. A post-hoc analysis of the results of the GRAVITAS trial showed that the choice of a different cut-off value (PRU ≥ 208), according to the indication of a more recent observational study [73], would have allowed a more accurate identification of poor responders to clopidogrel and, possibly, the therapeutic success that was missed by the GRAVITAS trial [74]. However, the same authors that identified PRU ≥ 208 as the best cut-off value [73] indicated that PRU ≥ 235 best predicts the risk of cardiovascular events in a later publication [15]. A very recent meta-analysis identified PRU ≥ 230 as the best cut-off value [16], adding to the confusion on this very important issue. Moreover, the results of RECLOSE 2-ACS, a recently published large, prospective observational study [75], refute that a more accurate identification of poor responders to clopidogrel would be sufficient to ensure the success of treating poor responders with a high dose of clopidogrel. Indeed, despite the fact that the laboratory test (platelet aggregation induced by ADP, studied with light transmission aggregometry) could predict the risk of MACE of poor responders, the improvement of the pharmacological response to high doses (up to 300 mg daily maintenance dose) of clopidogrel (or to ticlopidine, 500–1000 mg daily) was not associated with a reduction of the incidence of MACE [75]. The criticisms that have been raised about the GRAVITAS trial, the amendments that have subsequently been proposed by its authors and the therapeutic failure of RECLOSE 2-ACS further emphasize our uncertainties and, as a consequence, the prematurity and incorrectness of tailoring clopidogrel treatment based on laboratory tests in clinical practise. Other trials of tailored clopidogrel treatment based on laboratory monitoring are ongoing and their results are expected in the near future [3].

Alternative approaches to the problem of response variability to clopidogrel

As an alternative to laboratory monitoring with platelet function tests, the identification of carriers of loss-of function mutations of CYP through genotyping can be considered. In March 2010, the US Food and Drug Administration (FDA) added a ‘boxed warning’ to the label of clopidogrel including a reference to patients who do not effectively metabolize the drug and, therefore, may not receive the full benefits on the basis of their genetic characteristics [76]. Despite this warning, indications of how to manage these patients were not given. More recently the American College of Cardiology Foundation and the American Heart Association have published a consensus document addressing this FDA warning, which states that the role of genetic testing and the clinical implications and consequences of this testing remain to be determined [68]. Therefore, whilst waiting for the results of some ongoing studies [77], genetic testing should not be performed in clinical practise. Based on the findings of observational studies, it is difficult to foresee that this approach will be very successful as it has been demonstrated that CYP2C19*2 accounts for only 5–12% of the variability of response to clopidogel [78–80]. In addition, the sensitivity of CYP2C19 genotyping for poor responders to clopidogrel is as low as about 45% [78,81].

Other potential strategies of personalized treatment (serial testing, combined patient genotyping and serial testing, and use of the new P2Y12 inhibitors instead of high-dose clopidogrel in low responders) might prove effective [71,72]. However, (i) serial testing with or without genotyping will increase the overall cost of treatment, possibly offsetting the advantage of using the cheaper drug clopidogrel instead of the new, more expensive P2Y12 inhibitors; (ii) tailored antiplatelet treatment should not be considered an achievement of modern medicine to be pursued by any means, but, rather, a potential solution to the problem of hyporesponsiveness to clopidogrel (it is quite obvious that not having to face the problem would be preferable); and (iii) the use of the new P2Y12 antagonists prasugrel or ticagrelor instead of clopidogrel would markedly lessen the problem of hyporesponsiveness, because they effectively inhibit platelet function in the vast majority of patients [22], although some degree of variability of response is observed also with these drugs [82–84], as could be easily predicted. Both prasugrel and ticagrelor increase the incidence of bleeding, but this can be mainly ascribed to the fact that most patients treated with these drugs display a good inhibition of platelet function and are therefore protected from thrombosis and exposed to the risk of bleeding (as opposed to only about 70% of patients treated with clopidogrel) [22]. Accordingly, it has been clearly demonstrated that patients displaying good response to clopidogrel are at higher risk of bleeding than those who are non-responsive [22]. In other words, if all patients were to respond well to clopidogrel (which is the aim of tailored treatment with clopidogrel) they would most likely have a lower incidence of MACE, but also a higher incidence of bleeding, like that seen for prasugrel or ticagrelor-treated patients. The rate of bleeding complications is related to the degree of inhibition of platelet function, rather than to the type of drug used. The use of the new P2Y12 inhibitors in all patients, without testing, might prove more effective and cost-effective than personalized treatment. This hypothesis should be tested in controlled studies. While we await the results of additional controlled studies, personalized treatment should not yet be implemented in clinical practise.

Disclosure of Conflict of Interests

Honoraria and research support by Eli-Lilly/Daiichi-Sankyo and AstraZeneca.