August 31st, 2010

The CURE for Clopidogrel Genotyping?

CardioExchange welcomes Guillaume Paré to discuss his team’s work on the  utility of clopidogrel genotyping. The researchers genotyped for CYP2C19 alleles associated with loss-of-function or gain-of-function of clopidogrel in some 5,000 patients with ACS or A-fib from two large randomized trials. In both studies, clopidogrel had similar efficacy over placebo regardless of whether patients had loss-of-function alleles. In contrast, clopidogrel-treated patients with gain-of-function alleles derived more benefit over placebo recipients than did clopidogrel recipients with standard genotype. The findings were presented at the European Society of Cardiology Congress and published online simultaneously in the New England Journal of Medicine.

In contrast to previous research, your study found that CYP2C19 loss-of-function alleles had no effect on cardiovascular risk among clopidogrel recipients with ACS or A-fib. Why the discordance between what might be expected from the basic science studies and the results of the trial?

Guillaume Paré: Our understanding of the clopidogrel dose-response curve is limited with regards to outcomes, both ischemic and hemorrhagic.  While carriers of loss-of-function alleles are expected to have approximately 30% less active metabolite than noncarriers, it is not entirely clear how such a decrease will impact on clinical outcomes.  It is quite reasonable to assume that in certain individuals modest changes in active metabolite concentration will not materially alter the benefit of clopidogrel, for example if a dose-response plateau is reached.

What are you now recommending to practitioners? To the clinician who is doing genotyping or who was considering doing it?

Guillaume Paré: Clinical decisions must be made using all available information. Our data show that there is no effect of loss-of-function alleles in conservatively managed ACS and AF patients and provide a convincing argument against genotyping in these populations, but these data may not be applicable to other patient populations. At this point, we can only speculate on reasons behind differences between our study and previous reports.  Whether the effect of loss-of-function alleles on active metabolite concentration varies from one population to the other, or the concentration of active metabolite necessary for clinical benefit varies, these important questions will have to be resolved to define which patient populations are likely to benefit from genetic testing.

Clearly, more data will be necessary to better understand the pharmacological effects underlying these results.  Ultimately, randomized clinical trials including genetic information will be needed (and are underway) to determine the best treatment option for individual patients.

What’s the potential significance—if any—to your finding that ACS patients with the gain-of-function alleles had better outcomes over ACS patients with wild-type or loss-of-function alleles?

Guillaume Paré: Taken together with recent data from Sibbing and colleagues demonstrating increased bleeding in gain-of-function carriers, our data confirm the importance of the gain-of-function allele in clopidogrel-treated patients.  Nevertheless, there is still much to learn about this association and other populations will have to be investigated to fully characterize its effect.  If anything, this finding emphasizes the complexity of pharmacogenetic associations.  There are undoubtedly factors that need to be identified to explain why this allele seems to be associated with outcomes in certain populations but not others.

One Response to “The CURE for Clopidogrel Genotyping?”

  1. A point has been made that the critical pharmacogenomics
    (clopidogrel-CYP2C19 interaction) is with PCI/stenting. Let us examine
    the PLATO data in which nearly two-thirds of ACS patients underwent
    PCI with stenting. Amongst clopidogrel-treated patients, there were
    149 MACE events observed in 1388 patients with loss-of-function
    CYP2C19 genotype compared with 332 events in 3516 patients without
    loss-of-function CYP2C19 genotype The calculated sensitivity and
    specificity for CYP2C19 genetic test result is 31% and 72%,
    respectively. The corresponding positive predictive value (PPV) is 11% and the negative predictive value (NPV) is 91%. The PPV is 2% and NPV 98% for stent thrombosis related to the fact there were only 56 total events observed. These predictive accuracies are consistent with the 11% PPV and 94% NPV observed in TRITON and 10% PPV and 92% NPV for CYP2C19 genotype observed in the Hulot et al meta-analysis published recently in JACC 2010.

    In the ACTIVE A trial, amongst clopidogrel-treated patients, there
    were 22 stroke, non-CNS systemic embolism, MI, and vascular death
    events observed in 102 poor or intermediate metabolizers compared with
    34 events in 199 extensive metabolizers. The calculated sensitivity
    and specificity for CYP2C19 genetic test result is 39% and 67%,
    respectively. The corresponding PPV) is 22% and the NPV) is 83%. In contrast, the PPV in CURE for CV death, Mi or stroke is 8% and NPV is 89%. Thus, the predictive performance of CYP2C19 is actually better in ACTIVE A (an atrial fibrillation study) than in CURE (an ACS study).

    Bottom line, the predictive accuracies are of not sufficient magnitude
    to warrant routine genotyping at the present time.