To the Editor:

We read with great interest the paper published in Nature Medicine by Bouman et al.1 showing that a common paraoxonase-1 (PON1) polymorphism (rs662, also known as Q192R) is a major determinant of clopidogrel biological efficacy and stent thrombosis, whereas the common CYP2C19*2 allele (rs4244285-A allele) has no effect in this population. This suggests that previous data about the effect of the rs4244285-A allele on clopidogrel efficacy2,3 were mainly related to linkage disequilibrium with PON1 polymorphisms.

In an attempt to replicate the results of Bouman et al.1, we analyzed, in an ongoing large French registry, the association between PON1 rs662 and CYP2C19 rs4244285 polymorphisms and biological response to clopidogrel in 482 individuals with unstable coronary artery disease in Marseille. Subject characteristics are included in Supplementary Table 1. All patients were treated with 600 mg clopidogrel followed by clopidogrel 150 mg maintenance therapy. Platelet reactivity was assessed using 10 µM ADP (reported as ADP–induced platelet aggregation (ADP-Ag) values), and the specific pharmacological response to clopidogrel was assessed by detecting phosphorylation of vasodilator-stimulated phosphoprotein (VASP; reported as platelet reactivity index VASP (PRI VASP) values) between 12 and 24 h after hospital discharge and at 30 d after hospital discharge. CYP2C19 rs4244285 and PON1 rs662 genotyping was performed using amplification refractory mutation system PCR. PON1 rs662 genotyping was checked with restriction fragment length polymorphism PCR and Sanger sequencing. All methods are detailed in the Supplementary Methods . The study protocol was approved by the University Hospital La Timone Ethics Committee, and patients gave written informed consent for their participation.

Platelet reactivity and clopidogrel response were not affected by PON1 rs662 polymorphism status, either when measured shortly after the loading dose or after steady-state maintenance therapy (Fig. 1 and Supplementary Tables 2 and 3 ). Conversely, a strong influence of the CYP2C19 rs4244285 genotype on both platelet reactivity and clopidogrel response was observed at both time points (Fig. 1 and Supplementary Tables 2 and 3 ), as previously reported by our group and others2,3.

Figure 1: Platelet parameters according to PON1 rs662 and CYP2C19 rs4244285 genotypes.
figure 1

Platelet reactivity was assessed using the 10 µM ADP-Ag assay and the specific pharmacological response to clopidogrel by the PRI VASP assay.

Full size image

The lack of PON1 rs662 genotype influence on clopidogrel efficacy observed in our cohort may have several potential explanations. First, the present study might be underpowered to find a PON1 rs662 polymorphism effect; however, Bouman et al.1 described a significant effect on clopidogrel response with a smaller sample size. Second, the selected unstable patients and high clopidogrel doses used in our study might have affected our results compared with standard clopidogrel doses for stable patients in the Bouman et al.1 study. Moreover, the choice of platelet tests and timing of platelet testing was different, as this was done 12 to 24 h after loading dose and 30 d after the onset of acute coronary syndrome in our study. Bouman et al.1 tested clopidogrel pharmacological response using 20 µM ADP or active metabolite. Here we used 10 µM ADP and also used the most specific pharmacological test available to assess clopidogrel response, the PRI VASP assay. Finally, confounding factors may have influenced our results, as PON1's enzymatic activity is affected by several environmental factors including chronic renal disease, lipids, smoking and inflammation. Indeed, the individuals in our study were in a Mediterranean city, which might be slightly different from the population studied by Bouman et al.1. However, the baseline characteristics of the patients according to PON1 rs662 polymorphism were not different between genotypes in our cohort (data not shown). On the basis of our results, the necessity for PON1 rs662 genotyping in routine clinical practice seems questionable. Further large studies are urgently needed to address the observed discordance between our study and that of Bouman et al.1 and the potential relevance of PON1 genotyping to tailor antiplatelet therapy.

Note: Supplementary information is available on the Nature Medicine website.